Microsoft Tripled Revenue Under Ballmer as Enterprise Fueled Survival

Acquired

Despite high-profile consumer mobile failures and antitrust fallout, Ballmer's aggressive enterprise bundling and multi-year corporate agreements built the financial engine that preserved the company's monopoly power and funded its future cloud dominance.

AI-assisted summary of Microsoft Volume II: The Complete History and Strategy of the Ballmer Years (Audio) — original source. Check the original for full context.

Source published: 2024-07-22

Detailed notes

## Background This episode of Acquired was published on YouTube by the Acquired channel on July 22, 2024. According to the page description, it explores and analyzes the twenty-year period between 1999 and 2019 to uncover what actually happened between Microsoft's two eras as the world's most valuable company, tracking key milestones like the browser wars, Xbox, mobile, and the rise of Azure.

## TL;DR This comprehensive analytical walkthrough debunks the traditional 'Rise and Fall' narrative of Microsoft during the Steve Ballmer era, illustrating how the company transformed into a dominant enterprise technology powerhouse despite high-profile consumer stumbles. The narrative details critical operational arcs, including the DOJ antitrust trial, the systemic architectural delays of Windows Vista and Windows 8, and the strategic failure to match Google's 'less than free' Android mobile model. Ultimately, the program demonstrates how Ballmer insulated and funded early cloud infrastructure like Project Red Dog, creating an enterprise engine and data center footprint that laid the groundwork for Azure's ultimate multi-trillion dollar triumph under Satya Nadella.

## Atmosphere & style The presentation feels deeply researched, authoritative, and revisionist, driven by a collaborative, insider-heavy narrative style that balances dense historical strategy with engaging corporate folklore. There is a palpable sense of executive tension and high-stakes drama, moving from the defensive anxieties of Bill Gates's grueling antitrust deposition to the high-energy, morale-boosting resilience of Steve Ballmer. The conversational dynamic between the hosts evokes an analytical yet appreciative vibe, shifting from a clinical post-mortem of consumer failures to a profound realization of the invisible genius behind Microsoft's long-term enterprise agreements.

## Key takeaways - [00:32](https://www.youtube.com/watch?v=6oER1bdQa_s&t=32s) The conventional 'Rise and Fall' narrative of Microsoft under Steve Ballmer is inaccurate; the company successfully grew annual revenue from $6 billion in 1995 to $80 billion in 2014 by mastering enterprise software sales. - [11:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=663s) Bill Gates systematically shifted Microsoft's entire corporate focus toward the internet after recognizing its explosive growth, abandoning his previous 'Information Superhighway' and interactive TV frameworks. - [40:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2414s) Microsoft utilized a aggressive bundling strategy, offering Internet Explorer for free within Windows 95 to dry up Netscape's revenue and secure total browser market dominance by 2000. - [52:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3177s) The landmark 1998 United States v. Microsoft Corporation antitrust trial exposed aggressive behaviors and culminated in a 2000 breakup order that deeply damaged Microsoft's valuation and internal culture. - [01:30:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5403s) Brad Smith led a strategic pivot to settle outstanding regulatory disputes, turning Microsoft from a regulatory pariah into a trusted software provider for governments worldwide. - [01:32:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5531s) Microsoft built an enterprise juggernaut fueled by Enterprise Agreements (EAs), a flat-fee annuity model that bundled Windows, Office, and server products to create extreme customer stickiness. - [01:51:47](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6707s) In 2007, Microsoft's business model relied heavily on corporate stability, with 55% of its $51 billion total revenue derived directly from multi-year or single-year Enterprise Agreements. - [02:07:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7674s) Xbox operated as a financial sinkhole for years, but it kept Microsoft relevant to younger consumers and developed the online subscription DNA that helped birth Azure. - [02:27:02](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8822s) The engineering disarray of the Windows Vista development cycle acted as a human capital black hole, causing elite systems talent to migrate from Redmond to Amazon, laying the groundwork for AWS. - [03:28:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12480s) The commercial failure of Windows 8, which attempted to forcefully combine a touch-centric tablet UI with a traditional desktop, marked the definitive death of Microsoft as a dominant consumer company. - [03:38:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13128s) Steve Ballmer protected and funded 'Project Red Dog' (Azure) from internal on-premise Windows Server political antibodies, providing the vital corporate cover needed to incubate its cloud ecosystem. - [04:00:04](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14404s) Google successfully counter-positioned Microsoft in mobile by giving Android away to hardware OEMs for free, leveraging its core search advertising engine to subsidize the platform. - [04:08:17](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14897s) During the Ballmer era (2000-2014), annual revenue tripled and operating income grew 3x, but market capitalization shrank by 45% due to severe compression of its P/E multiple from dot-com bubble heights. - [04:26:26](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15986s) Satya Nadella revitalized Microsoft by returning to the classic 'Embrace and extend' playbook, accepting user behavior shifts toward open source and iPhones to capture new cloud revenue.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=0s) Introduction: A New Chapter for Microsoft

Ben Gilbert and David Rosenthal kick off the season 14 finale of *Acquired*, diving into the highly anticipated "Microsoft Volume 2." While the first volume dealt with ancient history, this episode marks a transition into the era listeners actually grew up with—the age of Windows XP, Xbox, the internet, search, mobile, and the Zune. Ben admits he initially expected this research to reveal a simple narrative of a fall defined by cultural decay, failed consumer products, and antitrust battles. However, after months of unpacking the history, they realized the framing is far from straightforward. Even the 1998 antitrust suit against the Department of Justice holds a lot more nuance than popular memory suggests.

The hosts also break some massive news for their community: on September 10th, *Acquired* will host its largest live event ever at the Chase Center in San Francisco, partnered with JP Morgan Payments. The evening will culminate in an onstage interview with a massive special guest: Mark Zuckerberg, who is not only a central figure in past tech acquisitions but is currently defining the next decade of computing through Meta's AI push. Before diving into the history, Ben delivers the standard disclaimer that the show is for informational and entertainment purposes only, joking that while they don't offer investment advice, anyone owning index funds technically holds a stake in the companies being discussed.

[00:32](https://www.youtube.com/watch?v=6oER1bdQa_s&t=32s) Debunking the Simple "Rise and Fall" Narrative

David introduces the conventional narrative surrounding Microsoft between 1995 and Satya Nadella taking the reins in 2014. He points to the opening paragraph of Nadella’s 2017 book, *Hit Refresh*, where Nadella writes about joining Microsoft in 1992 out of a desire to change the world, only to watch innovation later become strangled by bureaucracy and teamwork replaced by internal politics. Nadella even references the famous tech organization chart cartoon by Manu Cornet, which depicts Microsoft's internal divisions pointing guns at one another. The standard summary has always been: Microsoft was winning, then it sucked for a long time, and now it is winning again thanks to Satya.

However, after speaking with dozens of insiders, Ben and David discovered this trope doesn't hold up. The sheer scale of their primary research was unprecedented for the show. They interviewed four to five times as many people as they usually do for a single episode, compiling a thank-you list of over 20 people to unearth what actually happened beneath the public surface.

- Microsoft expanded its annual revenue from $6 billion in 1995 to $80 billion in 2014, fundamentally cracking the code on enterprise software sales despite public missteps. - Satya Nadella joined Microsoft in 1992, witnessing firsthand the shift from mission-driven culture to hyper-politicized bureaucracy.

[05:06](https://www.youtube.com/watch?v=6oER1bdQa_s&t=306s) The Missing Piece of Windows 95

The narrative picks up exactly where Volume 1 left off: the August 1995 consumer launch of Windows 95. It was an unabashed, perfect celebration of software complete with Jay Leno, the Rolling Stones, and the introduction of the Start button. Yet, there was one glaring omission from the operating system that consumers would soon desperately want: a built-in internet browser.

Ben notes that they intentionally omitted the internet from their previous coverage because it belongs entirely to this next chapter. Looking back, the true legacy of Windows 95 wasn't just its sleek desktop UI, but the fact that it became the foundation upon which humanity first accessed the modern web. At the time, things were shifting at a blistering pace dictated by "internet time"—where tech cycles occurred in weeks rather than years.

[06:33](https://www.youtube.com/watch?v=6oER1bdQa_s&t=393s) The Era of Walled Gardens and Interactive TV

Rewinding to the early 1990s, the concept of going online meant dialing into proprietary, walled-garden services like CompuServe, Prodigy, or the undisputed giant, AOL. These services were completely detached from our modern understanding of the open web; they bundled dial-up access with curated content. Ben emphasizes how restrictive this ecosystem was. If an independent creator wanted to host a website or publish content, they couldn't just spin up a server. Instead, they had to possess massive negotiating leverage and pitch an executive at AOL just to secure a spot on the network. Highlighting how strange this corporate landscape was, David reveals that CompuServe was actually owned by the tax preparation company H&R Block.

Aiming to secure its own piece of the online pie, Microsoft began eyeing AOL for a potential acquisition in 1993. When AOL founder Steve Case refused to sell, Microsoft co-founder Paul Allen independently bought a massive stake in the company anyway, creating a corporate headache. Realizing an acquisition was off the table, Microsoft pivoted to build a direct competitor code-named "Project Marvel." This proprietary network was designed to compete with AOL, though it would eventually fail and see its name repurposed for the internet-based media entity known as MSN.

Concurrently, Wall Street and tech elites believed these walled gardens were merely a temporary bridge to a grander, utopian infrastructure called the "Information Superhighway." The prevailing wisdom dictated that this future would manifest as interactive television controlled by pay-TV providers like Comcast, Time Warner Cable, and DirecTV. This assumption wasn't crazy at the time; television was the dominant consumer medium, whereas personal computers were not yet universally adopted entertainment hubs.

- To build a interactive media network, cable companies already possessed the screens, the input mechanisms, the content, and the physical wires leading into American homes. - Building an entirely separate infrastructure from scratch for PCs seemed far less practical to investors than leveraging existing television endpoints. - In the summer of 1993, rumors swirled around a massive joint venture between Microsoft and major cable providers, colloquially dubbed "Cablesoft."

[11:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=663s) The Tiger Project and Bill Gates's Foresight Pivot

Under the envisioned Cablesoft framework, cable companies would control the pipes and customer relationships, while Microsoft would provide the software engine. This effort became code-named "Tiger," an ambitious project to develop operating software for both consumer set-top boxes and the massive back-end media servers required to stream content. To complete an "Unholy Alliance," Microsoft partnered with Silicon Graphics (SGI) to build the powerful hardware needed to handle heavy visual data. SGI was a legendary Silicon Valley powerhouse responsible for the groundbreaking CGI in *Jurassic Park*, led by its iconic chairman, Jim Clark.

As Wall Street hyped this up as a trillion-dollar opportunity, Bill Gates was caught in a media whirlwind. Paparazzi captured "spy shots" of Gates meeting with media moguls like John Malone of TCI and Gerald Levin of Time Warner. Gates even began spending time with Hollywood super-agent Michael Ovitz to figure out how Microsoft could break into content creation, a thread that eventually birthed the MSNBC cable network. Up to this point, Microsoft had thrown its weight behind Marvel and the Information Superhighway—neither of which involved the open internet or a web browser. It was a classic example of Microsoft's historical strategy: place multiple massive bets near the horizon to ensure they are in the mix, even if the exact destination is unclear.

This strategic blur is perfectly captured in Bill Gates's publishing timeline. Gates wrote a book to evangelize his vision of the future, but due to the slow nature of traditional publishing, *The Road Ahead* didn't hit shelves until November 1995. By then, the Netscape IPO had already taken the world by storm and Windows 95 had shipped. David reveals he keeps both the hard cover and revised soft cover editions on his desk to illustrate a staggering corporate pivot. The 1995 hard cover focuses entirely on the Information Superhighway and Gates's concept of "information at your fingertips." When the soft cover released in 1996, Microsoft had effectively performed a "control-find" on the manuscript, replacing references to the Information Superhighway with the internet and the web browser.

- In the hard cover edition of *The Road Ahead*, the index lists only three distinct sections discussing the internet. - In the soft cover version published just a year later, the index entries for the internet require an entire page, complete with detailed sub-indexes.

[14:32](https://www.youtube.com/watch?v=6oER1bdQa_s&t=872s) The J. Allard Memo and "Embrace and Extend"

While corporate leadership was focused on interactive TV, a young Windows networking engineer named James "J." Allard saw the future clearly. In January 1994, Allard penned a legendary internal memo to Bill Gates and senior executives titled "Windows: The Next Killer Application for the Internet." Allard pointed out a nascent piece of software spreading like wildfire across academia: a web browser called Mosaic, built out of the National Center for Supercomputing Applications at the University of Illinois. Remarkably, the software was coded by a student programmer named Marc Andreessen in his spare time.

Allard's memo argued that Mosaic's explosive, exponential growth curve signaled that the open internet was the real future. He urged Microsoft to "Embrace and extend" the internet directly into the Windows operating system. This memo marks the birth of the famous corporate mantra. While critics and political adversaries would later twist the phrase into "embrace, extend, and extinguish," Ben notes that the original line—"Embrace, extend, innovate"—was an incredibly sound business playbook. It allowed Microsoft to eliminate product-market-fit risk by identifying a user behavior that people already loved, incorporating it into their dominant OS, and leveraging their existing OEM business model to monetize it. Allard laid out this entire winning internet blueprint months before Netscape as a corporate entity was even founded.

The momentum built rapidly. In February 1994, Steven Sinowski—then serving as Gates’s technical assistant and shadow—went on a recruiting trip to his alma mater, Cornell University. A massive snowstorm trapped Sinowski on the Ithaca campus for several extra days. Hunkered down, he noticed that the stranded students weren't using proprietary networks; they were entirely captivated by the open internet. Though Sinowski was an academic who understood the internet's roots as a tool for scientists to swap data, seeing ordinary students treat it as a vibrant, everyday community sparked a massive realization of what was to come.

[17:41](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1061s) The "Cornell is Wired!" Memo and Microsoft’s Internet Awakening

In the early 1990s, academia viewed the internet primarily as a research network with some "cool entertainment use cases," but absolutely no commercial or business interest. It was seen strictly as a medium for academic communication. This perception was shattered when Steven Sinowski visited the Cornell campus. He was floored to discover that students were not using the network for papers or schoolwork; instead, they were using it for flirting, registering for classes, and messaging each other.

Excited by this grassroots digital culture, Sinowski penned a company-wide memo to Bill Gates and the leadership team titled "Cornell is Wired!" This sparked a major realization within Microsoft, where milestones are historically driven by such internal proclamations. Reflecting on Sinowski's insights, Gates later remarked, "When I heard Steven talk about what was happening at Cornell, I began to take the internet quite seriously."

* Sinowski’s memo highlighted non-academic student usage: messaging, flirting, and class registration. * Microsoft organized an "internet offsite" on April 5, 1994, to investigate the phenomenon. * The offsite occurred exactly one day after Netscape was incorporated (April 4, 1994). * Gates embraced the internet as an "exponential phenomenon," aligning it with Microsoft's foundational belief in Moore's Law.

[19:58](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1198s) Jim Clark, Mark Andreessen, and the Settop Box Misstep

In February 1994, Silicon Valley legend Jim Clark resigned from Silicon Graphics Inc. (SGI), the company he founded, out of frustration with a board that refused to aggressively pursue the "information superhighway." On his very last day at SGI, Clark cold-emailed Mark Andreessen, the young programmer behind the celebrated Mosaic browser at the University of Illinois. Clark expressed his admiration for Mosaic's growing adoption and suggested a collaboration.

The duo teamed up and founded a company called Electric Media on April 4, 1994. Ironically, despite Andreessen's deep understanding of the web's exponential potential—having already introduced the `<img>` tag to render images inline—their initial business plan ignored the open internet. Instead, they focused on building set-top boxes for the information superhighway, leveraging Clark's relationships with Nintendo to try and transform the upcoming N64 console into a living-room gateway.

* Jim Clark emailed Mark Andreessen in February 1994, stating a collaboration would be "of interest to me." * Electric Media (soon renamed Netscape) initially focused on the information superhighway rather than the open web. * The information superhighway concept eventually failed, existing only as small-scale cable company tests involving roughly 300 homes. * The hosts note that this misstep reveals how the human brain is simply "not wired to understand compounding" growth.

[23:51](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1431s) The Rise of Netscape and the Battle Over the Mosaic Name

By late spring 1994, Mosaic reached 1 million active users, catching the attention of Bill Gates. Realizing the massive traction of the open web, Clark and Andreessen scrapped their N64 set-top box plans to pivot fully to the browser. They raised venture capital from Kleiner Perkins, adding legendary investor John Doerr to their board. By October 1994, the newly christened Mosaic Communications Corporation released its first browser, Mosaic Navigator, for free, planning to monetize by charging corporations for web server software.

However, using the name "Mosaic" triggered an intense legal battle. The University of Illinois and the NCSA realized they owned a highly valuable codebase in the original Mosaic. Even though Andreessen's team wrote their commercial browser entirely from scratch, the university licensed the original code to a local company named Spyglass. Spyglass repeatedly blew up commercial deals for Andreessen’s startup by threatening to sue their prospective server software clients, culminating in a lawsuit, a settlement, and a permanent corporate rebranding to Netscape in the fall of 1994.

* Netscape's original business model: give away the consumer client browser for free, charge for server software. * Andreessen's browser used completely new code, but adopted the "Mosaic" moniker to draft off his past academic success. * Andreessen famously and facetiously told the press: "You go to school, you do your research, you leave and then they try and sue your business. Had I known this would happen, I would have gone to Stanford."

[27:26](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1646s) The Scramble for a Browser and AOL's $30 Million Preemption

By late 1993 and early 1994, mainstream online service providers like CompuServe, Prodigy, and AOL began interconnecting with the internet, allowing users to surf the open web alongside proprietary networks. These platforms urgently needed to license web browsers. Netscape refused to license its Navigator client because it conflicted with its business model, forcing online services to look elsewhere. Spyglass began licensing the original academic Mosaic code, while a small startup called Booklink coded its own browser from scratch to enter the licensing market.

In May 1994, hot off their internet retreat, Bill Gates and Steven Sinowski met with Booklink to negotiate a $2 million licensing deal to integrate the browser into Windows. In a sudden twist, AOL swooped in and purchased Booklink outright for $30 million, leaving Microsoft completely empty-handed. With Booklink off the market and Netscape refusing to license, Microsoft turned to Spyglass, paying $2 million to license the source code of the original academic Mosaic to serve as the foundation for Internet Explorer.

* Online services sought browsers to transition users from proprietary platforms to the open web. * Microsoft was in the middle of negotiating a $2 million deal with Booklink when AOL preempted them. * Microsoft purchased its license from Spyglass for $2 million to acquire the NCSA Mosaic source code.

[29:39](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1779s) The Messy Reality and Architecture of Internet Explorer

While public narrative and the early Internet Explorer "About" boxes credit Spyglass and NCSA Mosaic as the bedrock of IE, the internal reality at Microsoft was far more complex and disorganized. In late 1994, a Windows 95 team led by Thomas Reardon was spun off to plan for an internet-native operating system upgrade, provisionally called "Windows 97." Rather than seeing a browser as a standalone application, this team envisioned integrating HTTP protocols directly into the OS shell and building a reusable HTML-rendering UI component that any Windows application could frame in.

Microsoft actually offered this built-in HTTP stack and HTML engine to Netscape, inviting them to use the off-the-shelf Windows code rather than rewriting it, an offer Netscape famously declined. Consequently, Internet Explorer became the primary application utilizing these deeply embedded Windows components. Furthermore, the Spyglass code Microsoft bought turned out to be heavily altered and low-quality, forcing Microsoft engineers to spend significant time undoing the Spyglass modifications. The acquisition was highly reminiscent of Microsoft's early purchase of QDOS to create MS-DOS—a mere starting point that required massive development to become a commercial product.

* Early IE versions contained a text box noting it was "based on NCSA Mosaic distributed under a licensing agreement with Spyglass Inc." * The Windows 97 team envisioned a future where the entire Windows shell operated like an HTML-rendered server. * Internet Explorer was fast-tracked for Windows 95 and shipped as part of the optional "Plus Pack." * The structural integration of internet protocols into the Windows OS would later become a critical focal point in Microsoft's antitrust litigation.

[35:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2115s) The "Internet Tidal Wave" Memo

As preparations for the launch of Windows 95 intensified in the spring of 1995, Bill Gates formalized Microsoft's radical strategic shift. He authored an overarching company memo intended for wide circulation and publication. This document, titled the "Internet Tidal Wave," signaled that the internet had officially become the absolute top priority across every single division of the technology giant.

> "Perhaps you have already seen memos from me or others here about the importance of the internet. I have gone through several stages of increasing my views of its importance. Now I assign the internet the highest level of importance. In this memo I want to make clear that our focus on the..."

[35:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2148s) The Existential Threat of Netscape

The internet is explicitly described as the most important single development in technology since the introduction of the IBM PC in 1981, carrying even greater significance than the arrival of the graphical user interface. This massive technological shift sets the stage for the August 1995 launch of Windows 95, scheduled for the 24th. However, just a couple of weeks prior on August 9th, Netscape goes public with a staggering market capitalization of $3 billion. Netscape’s explosive growth—surging from 1 million to 15 million users in a single year—signals an instant, undeniable product-market fit. In an ecosystem where tech observers are constantly questioning if a new trend will "become a thing," there was absolutely never a question about the internet from 1994 onward.

During the IPO press cycle, Netscape co-founder Marc Andreessen famously boasts that Netscape will soon "reduce Windows to a poorly debugged set of device drivers." This provocative quote carries a profound strategic implication: while Windows was currently the dominant platform where independent software vendors wrote applications, the static web pages of 1995 were merely a stepping stone toward rich web applications utilizing JavaScript, CSS, Java, and Flash. Andreessen's vision is that users will eventually boot up an operating system simply to connect to their screen, mouse, and keyboard, and then perform all of their actual work and computing inside the browser itself.

This possibility terrifies Microsoft, which independently arrives at the exact same conclusion. If the web becomes the application platform of the future, Microsoft's incredibly lucrative business model—built on developers needing their operating system to reach users, and users needing it to run software—could completely evaporate. In his famous "Internet Tidal Wave" memo, which the Department of Justice later seized as an exhibit, Bill Gates recognizes this immediate threat, writing that Netscape's 70% browser usage share allows them to determine network extensions and commoditize the underlying operating system via a multi-platform strategy. To defend its ecosystem, it becomes existentially critical for Microsoft to crush Netscape by building Internet Explorer directly into Windows. Ironically, Netscape's brilliant "Trojan horse" strategy of distributing the platform of the future as a Windows application relies entirely on Microsoft’s massive 80% to 90% market share to reach consumers in the first place.

- Netscape goes public on August 9, 1995, with a $3 billion market capitalization. - Netscape's user base expands from 1 million to 15 million users within a single year. - Bill Gates notes in his memo that Netscape holds a dominant 70% browser usage share.

[40:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2414s) The Rise and Monopoly of Internet Explorer

When Windows 95 launches just a few weeks after the Netscape IPO, Internet Explorer is notably absent from the retail box version. Instead, consumers have to purchase it for $50 as part of the "Plus Pack." While Microsoft makes money on these software sales, a paid add-on does nothing to dent the free version of Netscape Navigator dominating the market. Because Microsoft's true strategic goal is to "cut off the air supply" of Netscape and achieve instant ubiquity, they realize they cannot care about direct revenue; they must distribute the browser freely to prevent the internet from kneecapping their core business. Meanwhile, Netscape’s hyper-growth continues through the autumn of 1995, with its stock tripling over September, October, and November to reach an insane $10 billion market cap. This massive valuation is driven entirely by the public's belief that Netscape holds the platform of the future, rather than its current financial performance or server software revenue. The massive "hype train" that previously backed the concept of the "Information Super Highway" completely ports over to Netscape.

The turning point occurs on December 7, 1995, when Bill Gates blindsides the industry by announcing that Internet Explorer will be completely free and bundled into every single copy of Windows 95 moving forward. On that day alone, Netscape stock plummets by about a third, marking a high-water mark from which the company never recovers. The swift shift underscores a brutal business reality: if your primary distribution channel decides to compete with you and makes it a priority, your business can end in a minute.

Microsoft aggressively executes this bundling strategy. By the end of 1996, they secure deals with the major legacy online services—AOL, CompuServe, and Prodigy—to ditch competing browsers and bundle Internet Explorer instead. The resulting market share shift forms a perfectly rounded hill over the next decade. Internet Explorer goes from 0% market share in 1995 to passing 20% in 1996, 40% in 1997, 60% in 1998, and effectively achieving 100% worldwide browser market share by the year 2000, before eventually sliding back down to 0% by 2010. Netscape's revenue dried up because controlling the client (the browser) was critical to selling their lucrative server software; once developers and users shifted away from Netscape's browser, the incentive to buy their server tools vanished. The company-killing 80% stock drawdown leaves employees undercompensated, completely destroying the company's momentum. Throughout this entire conflict, Microsoft never actually changes its underlying business model—they simply vanquish a dangerous potential future while continuing to cash checks from selling standard Windows licenses through OEMs and retail.

- On December 7, 1995, Bill Gates announces Internet Explorer is free and bundled with Windows, causing Netscape stock to drop by roughly 33%. - Internet Explorer market share milestones: >20% in 1996, >40% in 1997, >60% in 1998, and virtually 100% by 2000. - Netscape is ultimately acquired by AOL in November 1998 in an all-stock deal valued at just over $4 billion.

[44:47](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2687s) Saving Apple and the Peak of Microsoft's Power

A fascinating and often misunderstood piece of color from this era unfolds in August 1997 at the famous Macworld expo, where Steve Jobs returns to lead Apple. Bill Gates famously calls into the event via a giant, legendary satellite feed. The resulting partnership consists of four core deal points. First, Microsoft makes a $150 million investment in Apple, which at the time was worth only $2 billion, effectively giving Microsoft an 8% stake in its rival. Second, Microsoft makes a five-year commitment to continue shipping Microsoft Office for Mac. This second point is truly existential for Apple; the company was in such a precarious financial position that if Microsoft had walked away from Office for Mac, consumers would have stopped buying Mac computers altogether, putting Apple completely out of business.

The third point of the deal formally ends all ongoing patent disputes and copyright lawsuits between the two companies. The fourth point, which is frequently overlooked, mandates that Internet Explorer replace Netscape as the default browser on the Mac—a arrangement that stands from 1997 until Apple introduces Safari in 2003. Steve Jobs cleverly leverages Microsoft's mounting legal anxieties to close the deal. Knowing that the US Department of Justice is actively investigating Microsoft for antitrust violations, Jobs approaches them with a compelling pitch: instead of facing further IP lawsuits, Microsoft can appear collaborative with a major ecosystem player by investing in Apple. Microsoft agrees, prioritizing their aggressive goal to get Internet Explorer everywhere, resulting in a massive cross-licensing win-win that ultimately saves Apple from bankruptcy.

By November 1998, when AOL buys Netscape for just over $4 billion in an all-stock deal (a little over a year before the infamous Time Warner merger), Microsoft reaches what is described as the absolute peak of its power as a consumer technology company. Their market dominance is unparalleled: Apple is existentially reliant on them, Netscape is thoroughly crushed, and the modern tech titans do not yet exist—Google is still three years away from being founded, and Facebook is nine years away. While formidable competitors like Sun Microsystems and Oracle exist in the enterprise space, Microsoft holds absolute, unchecked monopoly power over the consumer technology landscape. This unprecedented concentration of market power raises a natural, lingering question: shouldn't completely wiping out an existential threat by cutting off their air supply be illegal?

- Microsoft's 1997 investment in Apple totals $150 million, representing an 8% stake in the $2 billion company. - Internet Explorer serves as the default browser on Mac computers from 1997 until 2003. - The late 1990s represent a pre-modern tech landscape where Google is 3 years away from inception and Facebook is 9 years away.

[49:51](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2991s) Mid-Episode Sponsor: J.P. Morgan Payments

The host smoothly transitions from discussing historical software platforms to thanking the episode's sponsor, J.P. Morgan Payments, highlighting what it takes to build a truly great modern developer platform. Just as Microsoft and Netscape battled over development ecosystems in the 1990s, creating an empowering developer platform today requires a deep cultural focus on user enablement and a long-term investment horizon. J.P. Morgan is proving to be far more than a traditional global bank by heavily prioritizing its developer ecosystem, pouring $7 billion annually into technology and R&D, and fully embracing a "dev-first mentality" over the past year with its API-powered payment solutions.

This investment reflects a broader shift in corporate decision-making. Traditional finance executives are no longer the sole gatekeepers; developers, engineers, and product managers now hold a major seat at the table when evaluating payments infrastructure. Modern, digital-first industries like ride-sharing, e-commerce, and travel rely entirely on seamless, native payment flows built directly into the user experience. J.P. Morgan’s new payments developer portal serves as a self-serve, One-Stop Shop where engineers can build customized business solutions on top of a highly trusted, massive-scale platform.

- J.P. Morgan invests $7 billion annually into technology and research and development. - The company's financial infrastructure moves $10 trillion per day globally. - J.P. Morgan processes greater than 50% of all e-commerce transactions in the United States. - The upcoming product roadmap includes payment APIs spanning treasury, commerce, embedded finance, and account validation.

[52:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3177s) Antitrust: The United States v. Microsoft

The narrative arrives at the landmark 1998 antitrust trial: *United States v. Microsoft Corporation*. Reflecting on the dramatic conclusion of the Browser Wars, the hosts note that their previous casual jokes about Microsoft "cutting off Netscape's air supply" cut straight to the core of the legal battle. The fundamental question of the trial is whether Microsoft's aggressive behavior should be deemed illegal. Such a massive concentration of market power had arguably never been held by a single corporate entity in American history, setting up an inevitable clash with the United States Department of Justice.

[53:27](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3207s) The Core of the Microsoft Antitrust Saga

The debate over Microsoft's practices raises complex legal and emotional questions. While competitors generated a growing public perception of Microsoft as a "bully," the internal reality for the vast majority of Microsoft employees was entirely different. Most engineers and developers were not focused on business or competitive strategy; their goal was simply to "ship great software and make things that people love to use." This creates an interesting tension between the intentions of the creators and the reality of antitrust laws, which look past subjective intent to analyze actual market behavior.

From a consumer standpoint, the issue hinges on the concept of "consumer harm" and the consumer welfare standard. After all, a typical user might wonder why they should be unhappy about receiving a world-class web browser bundled for free inside their operating system. This core question—whether Internet Explorer was a standalone product being illegally tied to Windows or simply a modern operating system feature—became the battlefield for the US government's case.

[55:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3300s) FTC Deadlocks and the DOJ Stepped In

The legal battle did not actually begin with the Department of Justice (DOJ) in 1998, but rather with the Federal Trade Commission (FTC) all the way back in 1990. The FTC opened an investigation centering on Microsoft's "per-process licensing" practices. By July 1993, the FTC Commissioners met to vote on whether Microsoft was a monopoly deserving of penalties, resulting in a 2-2 deadlock. Legally, this tie meant no action would be taken, handing Microsoft a massive victory that theoretically should have closed the federal government's antitrust case.

However, the very next month, in August 1993, the DOJ took the highly unprecedented step of picking up the exact same case. Having one department of the federal government decline to prosecute only for another department to step in weeks later meant Microsoft was effectively standing trial for the same accused crimes twice. Despite internal opposition from FTC commissioners who refused to turn over their investigative notes to the DOJ, the justice department pushed forward. Rather than enduring a lengthy trial, Microsoft chose to settle in July 1994 by entering into a famous consent decree.

- **1990:** FTC opens its initial antitrust investigation into Microsoft. - **July 1993:** FTC commissioners deadlock 2-2; no penalties are applied to Microsoft. - **August 1993:** DOJ takes over the case in an unprecedented double-investigation move. - **July 1994:** Microsoft signs a consent decree agreeing not to tie independent application sales (like Office) to the sale of Windows.

[57:59](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3479s) Product vs. Feature: The Consent Decree Loophole

The 1994 consent decree drew a sharp line: Microsoft could not tie the sale of independent application products to the sale of Windows, meaning they couldn't force original equipment manufacturers (OEMs) to buy Office just because they bought the operating system. Crucially, however, the decree stated that Microsoft remained free to integrate additional *features* into the Windows operating system. This created a massive gray area as technology evolved: what is considered a standalone application today can easily become a standard core feature years down the line.

From Microsoft's perspective, Bill Gates and engineers like Jay Allard always intended for an internet browser to be a core feature of Windows. While they were clearly motivated by a desire to maintain their monopoly market share and "cut off the air supply" of rising competitors like Netscape, there was also a highly sympathetic engineering argument that an operating system in the late 1990s simply required built-in internet functionality. Today, it is impossible to imagine purchasing a device whose core operating system does not include a built-in web browser.

[01:00:12](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3612s) The 1997 Contempt Motion and Petty Compliance

In October 1997, the DOJ escalated the battle by filing a motion in Federal District Court, claiming Microsoft's bundling of Internet Explorer directly violated the 1994 consent decree. The DOJ demanded that Microsoft immediately stop shipping IE with Windows. Microsoft pushed back hard, arguing that the browser was not a separate piece of software but a deeply integrated part of the Windows shell—pointing out that users could even paste web addresses directly into Windows Explorer to render pages.

Federal Judge Thomas Penfield Jackson sided with the government and ordered Microsoft to provide a version of Windows to OEMs that completely excluded IE. Microsoft complied with the literal order but did so in a way that exposed the software dependencies: disabling the browser code caused the rest of the operating system to break. This move angered Judge Jackson, who felt Microsoft was complying with the letter of the law while completely "thumbing its nose" at the spirit of his ruling. Because the broken version of Windows was unusable, PC makers never actually shipped it to customers.

- **October 1997:** DOJ files a court motion accusing Microsoft of violating the consent decree through product tying. - **Judge Jackson's Order:** Microsoft is ordered to supply an IE-free version of Windows to OEMs. - **The Result:** Microsoft strips the code, the operating system breaks, and the DOJ attempts to hold Microsoft in contempt of court.

[01:02:33](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3753s) The Launch of the Massive 1898 Antitrust Case

The petty back-and-forth in Judge Jackson's court quickly led to a Microsoft appeal. In early May 1998, the appellate court handed Microsoft a temporary victory, ruling that the company could continue bundling IE—or any other features—as long as the integration provided a clear benefit to consumers. Because US antitrust courts prioritize consumer welfare (such as keeping prices low and adding value), Microsoft appeared to be in the clear.

The victory lasted exactly one week. On May 18, 1998, the DOJ retaliated by launching a massive, wide-ranging antitrust lawsuit against Microsoft for violating the Sherman Antitrust Act. This new suit went far beyond product tying; it fundamentally questioned whether Microsoft was a monopoly and whether it was systematically abusing that power across its entire business to suppress competition. Inside Microsoft, this felt like a bitter betrayal and a case of "triple jeopardy" by their own government. Believing they were being targeted by heavy behind-the-scenes lobbying from Netscape and other rivals, Microsoft's leadership dug in their heels and resolved to fight the new lawsuit with an incredibly aggressive legal strategy.

- **Early May 1998:** Appellate court rules Microsoft can bundle features that benefit consumer welfare. - **May 18, 1998:** DOJ retaliates with a massive, comprehensive lawsuit under the Sherman Antitrust Act. - **Market Share Fact:** At the time of the trial, Microsoft controlled over 90% of PC operating system sales.

[01:06:22](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3982s) Bill Gates' Disastrous 20-Hour Deposition

As the trial loomed in the fall of 1998, Judge Jackson issued a critical pre-trial order: all depositions were to be submitted to the court strictly in written transcript form. Operating under the assumption that no video would ever be shown in court or seen by the public, Bill Gates and his legal team tailored a hyper-defensive strategy for his deposition. On August 27, 1998, Gates sat down for a grueling 20-hour interview over multiple days with the DOJ's master prosecutor, David Boies. Gates' strategy was to never give an inch, pause for long periods, and avoid saying anything that could be parsed cleanly on a written page.

The resulting footage—12 hours of which are publicly available on YouTube—backfired catastrophically. Because Gates believed he was only writing for a transcript, he displayed pure disdain for the prosecutor, coming across as "rude, pedantic, and disrespectful." Gates repeatedly rat-holed on trivialities, such as refusing to answer questions about specific "memorandums" because they were technically sent as emails. At one point, Gates famously smirked at David Boies and asked him how he would define the word "definition."

On October 9, shortly after the deposition concluded, Judge Jackson issued a surprise reversal, ruling that videotaped depositions *could* be played aloud in court. Prosecutor David Boies weaponized this shift masterfully. Throughout the trial, whenever the defense built up sympathy for Microsoft, Boies would strategically play a clip of Gates' evasive, smirking deposition. This tactical dripping of video segments completely destroyed any judicial sympathy, kept the packed press box supplied with daily juicy headlines, and permanently damaged public opinion of both Bill Gates and Microsoft.

[01:10:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4254s) The Decision Not to Appeal the Deposition Tape Rule

The speakers address a question from a listener named David about whether Microsoft formally appealed Judge Jackson’s procedural change that allowed videotaped depositions rather than standard written transcripts. In preparing for the episode, a book titled *World War 3.0*—which focuses exclusively on the historic antitrust trial—revealed that Microsoft deeply feared Judge Jackson was fundamentally hostile toward them. They were especially aggrieved by his October 9th ruling allowing the video recordings, suspecting that the Department of Justice (DOJ) had somehow backchanneled or influenced the judge to alter his original pre-trial order.

While Judge Jackson categorically denied any improper influence, he couldn't quite recall his exact reasoning for granting the order. Microsoft chose to grumble about the decision only in their most guarded, private moments. They were utterly terrified of further agitating or offending the judge, convinced he was already heavily biased in favor of the government. Ultimately, Microsoft made no formal challenge to the rule change. The speakers observe that the company likely realized they had a long road ahead with Jackson and didn't want to dig themselves into a deeper hole, though they also note that Microsoft likely failed to grasp just how profoundly disastrous those video tapes would eventually be for Bill Gates and the company's public image.

---

[01:12:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4344s) The Finding of Fact and the Landmark Breakup Order

The long, grueling legal process began to culminate in November 1999 when Judge Jackson issued his official "finding of fact," formally declaring Microsoft a monopoly in the PC operating systems business. The speakers emphasize an essential nuance of antitrust law: it is not inherently illegal to be a monopoly; it is only illegal to abuse that monopoly power. However, Jackson's finding signaled to everyone that the other shoe was about to drop. The core of his finding rested on market network effects—specifically, the massive installed base of Windows users and the vast ecosystem of compatible applications made it prohibitively expensive for any competitor to build a viable alternative.

The true shockwave hit on June 7, 2000, when Judge Jackson issued his final judgment. Ruling that Microsoft had definitively abused its monopoly power, he ordered the ultimate structural remedy: Microsoft was to be broken up into two completely separate corporate entities—an operating system company and an applications company. The hosts stress that this jaw-dropping piece of history is often entirely forgotten by everyone except tech old-timers. This was not a mere legal slap on the wrist; the United States government ordered the dismantling of Microsoft, drawing direct parallels to the historic Standard Oil breakup nearly a century prior.

- **The Separation Mandate:** The court ordered that Bill Gates and Steve Ballmer must work at separate companies; they were barred from staying together. - **Divestment Rule:** Whichever company Gates or Ballmer chose, they were legally required to fully divest all their shares from the other entity to completely eliminate conflicts of interest. - **Market Cap Destruction:** Following the initial 1999 finding of fact and the subsequent year of turmoil, Microsoft’s market cap plummeted 55%, cratering from $600 billion down to $270 billion.

---

[01:16:12](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4572s) Judge Jackson’s Secret Meetings and the Reversal on Appeal

The hosts draw a parallel to modern tech giants, asking the audience to imagine the market chaos if a modern judge ordered Apple to split iOS entirely away from its physical devices. Amidst the chaos of the ruling, a stunning technicality came to light in June 2000: it was discovered that Judge Jackson had been secretly meeting with reporters in his private chambers, giving them embargoed stories to drop the moment his rulings went live. Because this extrajudicial conduct is strictly forbidden, the appeals court promptly stripped Jackson of his duties and removed him from the case.

For a wild 15 to 16 months, the official law of the land dictated that Microsoft prepare for a corporate split. A new judge was installed to readjudicate the remedy portion of the case and push the parties toward a settlement. The speakers point out two massive macroeconomic and political catalysts that completely shifted the momentum: the tragedy of September 11, 2001, which galvanized the government to pull everyone into a room and settle the exhausting case, and the political transition from the Clinton Administration to the Bush Administration.

---

[01:17:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4666s) The 2001 Settlement and Its True Cultural Toll

On November 2, 2001—just weeks after the launch of Windows XP—the DOJ and Microsoft proposed a final settlement. The speakers laugh bitterly at the sheer timeline of the case: the antitrust lawsuit was originally filed during the Windows 95 era, persisted through Windows 98, and was only settled as Windows XP rolled out, all while Internet Explorer remained bundled with the OS the entire time. By the time of the settlement, Internet Explorer commanded roughly a 90% market share. This left critics feeling the legal process was an absolute failure because the anticompetitive damage was already done, while Microsoft sympathizers remained furious that the government had hamstrung a hyper-innovative company trying to navigate the existential shift to the internet.

The final settlement, formally approved in 2002, completely reversed the catastrophic breakup order. It upheld the monopoly ruling but replaced the split with a five-year consent decree containing specific behavioral restrictions. However, the hosts argue that the true, staggering cost of the trial was not monetary or product-related—after all, Windows XP went on to be a massive success, selling over half a billion copies and unifying the OS under the NT architecture. Instead, the true damage was emotional and cultural. Microsoft employees went from being celebrated as the "best and brightest in America" to being publicly vilified as "evil," creating a depressing, heavy cloud over the entire campus.

- **Consent Decree Terms:** Microsoft was banned from entering exclusive contracts with PC manufacturers that locked out rival software. - **Interoperability Requirement:** Windows had to remain fully interoperable with third-party software, requiring Microsoft to explicitly document its APIs for outside developers. - **The Competitor Hotline:** The settlement established an independent technical committee to handle complaints from competitors, which the hosts jokingly describe as essentially creating a "call line." - **Collateral Lawsuits:** The DOJ trial triggered a domino effect of private civil suits from rivals like Sun Microsystems, AOL, and RealNetworks, costing Microsoft billions in follow-on settlements. - **Two Decades of Scrutiny:** State attorneys general and European regulators dragged out independent litigation for years; the EU version wasn't settled until 2009, and the final US consent decree didn't expire until May 2011, meaning Microsoft spent 21 years under active antitrust litigation.

---

[01:24:52](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5092s) The Permanent Distraction of Bill Gates

Reflecting on the philosophy of antitrust enforcement, the hosts question whether everyday consumers were ever actually harmed by Microsoft's actions, noting that US antitrust frameworks traditionally protect consumer welfare rather than competitors. While Microsoft technically emerged from the battle unbroken, the operational toll was severe. Software development was permanently bogged down by massive protocol documentation requirements, saddling engineering managers with structural "drag" that slowed iteration times and allowed the company to fall behind emerging tech trends.

The ultimate casualty of the entire 21-year war, the hosts argue, was Microsoft's single greatest competitive advantage: Bill Gates himself. Going through such an intensely invasive, deeply personal legal battle fundamentally changed him. The immense distraction shattered the company's focus at a time when technology required a clear, firing-on-all-cylinders North Star strategy. An insider from that era later characterized the period as a collective "mental breakdown for the whole company," providing the chaotic backdrop for a historic guard-changing event.

- **The Executive Transition:** In July 1998, as the DOJ suit intensified, Steve Ballmer was promoted to President, solidifying his spot as the clear number two. - **Gates Steps Down:** On January 13, 2000, sandwiched tightly between the finding of fact and the final breakup order, Bill Gates officially resigned as CEO. - **The New Architecture Role:** Gates handed the chief executive reins to Ballmer, pivoting into a newly minted role as Chief Software Architect while retaining his title as Chairman.

[01:28:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5311s) Steve Ballmer as the Emotional Rock

The conversation highlights a frequently overlooked aspect of Microsoft's history: how Steve Ballmer served as the company’s "emotional rock" during its most turbulent times. While the public often remembers Ballmer for his high-energy, sweat-drenched antics on stage—screaming "developers, developers, developers" or jumping around declaring "I love this company"—the speakers contextualize these moments. This legendary, seemingly crazy dancing occurred in September of 2000, precisely when Microsoft employees faced the terrifying prospect that the government was going to break the company apart.

Ballmer's primary mission was to keep morale high and ensure everyone kept moving forward. The hosts note that nearly everyone they interviewed from that era expressed disbelief at how he pulled it off, emphasizing how much his leadership meant to the rank and file. It is described as "actually shocking" that Microsoft managed to retain so much of its core talent during a grueling 15-month window where the public and the tech industry assumed a corporate split was inevitable. Learning this context completely reframes the modern perception of Ballmer's intense stage persona as a calculated, deeply necessary effort to hold the company together emotionally.

- **Ballmer's 3-Point Agenda as CEO:** 1. Hold the company together emotionally ("job number one"). 2. Clean up the massive antitrust mess. 3. Keep the company growing, winning, and innovating.

[01:30:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5403s) Brad Smith and the Pivot to Peace

Upon taking the helm, one of Ballmer’s very first and most definitive moves was promoting Brad Smith to General Counsel—a figure who remains a pillar at Microsoft as its president. Ballmer gave Smith a simple, direct mandate: "go make peace." The hosts share an incredible anecdote about Smith's internal job interview with the Microsoft Board of Directors. For his presentation, Smith brought a PowerPoint deck consisting of exactly one slide containing a single sentence: "It's time to make peace."

Smith's philosophy was to identify which regulatory settlements the company could reasonably live with and systematically settle every single outstanding dispute. He recognized that for Microsoft to survive, the company had to stop nursing its wounds. It no longer mattered if the employees felt the antitrust lawsuit was inherently unfair or that the legal proceedings were a "sham of a process." They had to accept a new reality, which required a completely different psychological approach. The speakers marvel at how Ballmer, an institutional member of the "Old Guard," successfully bridged the gap into this "new guard" mentality. This brutal exercise in compartmentalization ultimately transformed Microsoft from a regulatory pariah into a trusted, default software and services provider for governments across the free world.

- **Regulatory Transformation:** Microsoft went from being the chief target of the DOJ to the only major tech giant not under active federal antitrust investigation today. - **The Peace Strategy:** Disregarding past grievances to prioritize corporate survival and institutional compliance over being "right."

[01:32:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5531s) The Enterprise Juggernaut and the Client-Server Flywheel

While handling the legal fires, Ballmer simultaneously executed the third item on his agenda: ensuring long-term commercial success. Bill Gates remained deeply involved as a full-time employee in his role as Chief Software Architect, meaning Ballmer still had Gates as the technical leader guiding future products. However, this era marked a fundamental, permanent shift for Microsoft. It began a transformational journey from being a broad consumer and enterprise technology company to becoming *the* dominant Enterprise technology powerhouse. The hosts point out that the massive enterprise juggernaut Microsoft is today was almost entirely built post-Department of Justice (DOJ) lawsuit.

The speakers break down the brilliance of Microsoft's corporate strategy during the client-server era, characterizing it as "the most incredible three-sided technology flywheel ever built." The enterprise ecosystem relied heavily on a tight integration between Windows client devices, Office applications, and a new suite of multi-billion dollar server products. By ensuring that these three components operated seamlessly together, Microsoft created a highly sticky environment. A key corporate benefit of this tight integration was that nothing Microsoft made actually had to be a "best of breed" point solution. Because everything was bundled and worked smoothly together out of the box, corporate IT departments saw no practical reason to look anywhere else.

- **Core Enterprise Products Built in the Ballmer Era:** SQL Server, Active Directory, Exchange, Dynamics, and SharePoint (the latter technically residing within the Office division). - **The IT Realization:** Microsoft discovered that the enterprise market is not fundamentally about individual users; it is about IT departments and centralized systems.

[01:36:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5769s) The Anatomy of Microsoft's $4 Billion Pillars

To illustrate how this empire was constructed, the hosts look back at a 1996 interview Bill Gates did with Kevin Kelly for *Wired* magazine. In that interview, Gates categorized Microsoft into four distinct buckets: Windows, NT/BackOffice, Office (which Gates noted was a $4 billion-a-year business at the time), and a final 10% bucket of "everything else" (including MSN, e-commerce, games, and joint ventures like DreamWorks and MSNBC). The hosts chuckle at how Gates referred to the server side as "NT/BackOffice," a naming convention that exposed his view of it as just "stuff that businesses buy," whereas Ballmer possessed the true operational passion required to weaponize it into a commercial monster.

The product momentum truly shifted in 1998 when Microsoft upgraded SQL Server into a legitimate, enterprise-ready database capable of rivaling stalwarts like IBM mainframes and Oracle. Crucially, unlike IBM, SQL Server ran on cheap, standard x86 Intel architecture. Microsoft’s sales pitch to enterprise IT was simple: lower your total cost of ownership by ditching expensive mainframes, buying cheap Dell windows boxes, and running the Microsoft stack. This database stack perfectly set the stage for Exchange, which arrived exactly as email was becoming the definitive "killer app" for corporate America. Microsoft aggressively bundled Exchange with Outlook, easily displacing legacy competitors like Lotus Notes and giving corporate buyers an undeniable, all-in-one communications package.

- **Active Directory as the Ultimate Anchor:** Once a company adopted Active Directory to manage its canonical list of users and devices, it inevitably "tipped" and bought the rest of the Microsoft software suite because every other application needed to reference that single source of truth. - **The Enterprise Agreement:** Ballmer pioneered an aggressive bundling strategy through Enterprise Agreements. If a customer was only using 30% of the bundle, they would automatically adopt other Microsoft tools for new business needs because those tools were essentially perceived as "free."

---

[01:33:52](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5632s) Sponsor Interlude: ServiceNow & Microsoft Co-pilot

The hosts take a brief detour to thank their season sponsor, ServiceNow, noting that it has evolved to join the ranks of companies like Microsoft as one of the most critical enterprise technology vendors in the world, with over 85% of the Fortune 500 running on its platform. They highlight a massive, recently expanded partnership between the two giants that mirrors the historic 1997 Macworld announcement where Bill Gates appeared via video to save Apple by committing to Microsoft Office for the Mac.

In this modern iteration, Satya Nadella appeared at ServiceNow's annual "Knowledge" keynote to announce the deep integration of their respective enterprise AI assistants. Beginning in August, users will be able to interact with ServiceNow's "Now Assist" directly within Microsoft Co-pilot, starting inside Microsoft Teams. The systems are designed to seamlessly pass context and retrieve internal company data with proper user permissions without actually exposing or sharing underlying data securely, allowing enterprise users to effortlessly trigger complex corporate workflows using standard office applications.

[01:45:55](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6355s) The Unseen Genius of the Enterprise Agreement

The sheer scale of business model innovation required to transform Microsoft from a "big boring Enterprise company" into an absolute juggernaut often goes overlooked. By understanding that software carries zero marginal costs, Microsoft realized it could bundle additional products together for free to engineer massive customer stickiness, driving long, durable, compounding revenue streams. This "land and expand" strategy thrived because their enterprise customers were expanding alongside them. Microsoft was pioneering how to build and sell technology systems in an entirely new era where enterprise systems actually had end-users—a dynamic that didn't exist during the preceding IBM era.

This led to the introduction of the Enterprise Agreement (EA). Rather than relying on a salesman to pitch Windows piece-by-piece, Microsoft amortized its go-to-market costs across its entire software suite. Under the EA, companies paid a flat annual fee per device—which practically meant per employee at the time—to grant their workforce blanket access to everything, including Windows and Office. This masterstroke converted a unpredictable, one-time software sale into a steadily growing annual annuity that scaled automatically with a client's headcount.

- **The Three-Year Alignment:** Enterprise Agreements were structured as three-year contracts. Conveniently, Microsoft aligned its development cycle to release a major new version of Windows or Office roughly once every three years, ensuring every client was guaranteed at least one major upgrade during their contract lifecycle.

---

[01:48:39](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6519s) Rogue Agents and the IT vs. User Tug-of-War

The shift to the Enterprise Agreement exposed a deep, inherent friction between what an enterprise IT buyer wants and what an actual employee wants. For an individual worker sitting at an office PC, the ideal machine is flexible, unrestrictive, and easy to use. Users want to procrastinate, play games, browse the internet, and run custom VBA macros—often willing to trade off security or use pirated software if it makes them more efficient. They act with their own agency rather than the corporation's best interests.

Conversely, corporate IT administrators view these employees as "Rogue agents" operating inside their systems. IT administrators demand total control to restrict user behavior, block unauthorized software installations, disable macros, and filter websites. Their priorities lean heavily toward security, privacy, legal compliance, and data protection. Microsoft solved this tension by offering deep customization through the EA, selling IT buyers the exact administrative control they craved.

- **The Innovation Trade-off:** While users constantly demand the latest and greatest software features, IT administrators have a strong incentive to keep things exactly the same to avoid the logistical nightmare of retraining thousands of employees.

> "When I was in Office, I always thought we could stop bundling new features for 10 years and it would be fine no one would notice I think people would probably pay more for it." — *Anonymous Former Microsoft Office Executive*

---

[01:51:27](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6687s) The Tyranny of Backward Compatibility

The pressure from the corporate sales force severely constrained Microsoft's product managers. Even when user research indicated that people wanted innovative features, the sales team would beg developers not to implement major UI updates, like the Office Ribbon, because radical changes introduced sales friction and customer objections. This environment turned corporate tech into a heavily locked-down ecosystem. For instance, an investment banking analyst entering the Wall Street workforce at UBS in mid-summer 2007 would find themselves using an incredibly restrictive Windows XP and Office 2003 setup. Everything was heavily firewalled to the point where even access to simple casual gaming sites like miniclip.com was rapidly shut down by management.

Enterprise clients did not want groundbreaking, disruptive upgrades; they wanted cost-efficient total cost of ownership (TCO) and absolute backward compatibility. Microsoft secured its enterprise dominance by promising that no matter what custom software or modifications a company relied on, future updates would never break them. While this built an unshakeable enterprise brand, it would eventually act as a massive anchor, holding the company back creatively when it needed to pivot.

- **Enterprise Revenue Dominance (2007):** By 2007, market analysts estimated that a staggering 40% of Microsoft’s $51 billion total revenue came directly from multi-year (three-year) Enterprise Agreements. - **Single-Year EAs:** An additional 15% of total revenue flowed from single-year Enterprise Agreements, meaning over half the company’s revenue (55%) was generated by this enterprise motion just seven years into Steve Ballmer's tenure as CEO. - **The Revenue Split:** Of the remaining revenue, 30% came from the traditional OEM Windows business, leaving a mere 15% for everything else in the Microsoft portfolio (such as Xbox).

---

[01:55:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6936s) Ditching Retail Boxes for the OEM Network Effect

Microsoft completely revolutionized its financial health by shifting away from the traditional retail boxed software model in favor of two highly optimized channels: Enterprise Agreements and OEMs. In the 1990s, selling physical boxes of Windows at retail carried an abysmal gross margin of just 29% due to the material costs of printing discs (especially multi-floppy sets), packaging, shipping, retail profit splits, and heavy marketing expenses.

By pivoting to an OEM model—selling licenses directly to PC builders like Dell and Lenovo—the gross margin skyrocketed to 75%. Microsoft simply delivered the software bits once, letting the hardware manufacturers handle the entire distribution and scaling process. This functioned similarly to a "networks of networks" model, allowing the independent scale of every PC manufacturer to accrue financial value directly to Microsoft without any added operational workload. They successfully hunted whales with their enterprise sales force while letting OEMs distribute their consumer operating systems, completely eliminating the inefficient retail box model. However, prioritizing enterprise buyers meant that user-facing software quality slipped from being priority number one—a blind spot that went unpunished until the iPhone arrived in 2007.

---

[01:58:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7111s) Windows XP and the Dual-Kernel Triumph

To understand the core state of Microsoft during the early 2000s, one has to look at the storytelling surrounding its operating system releases, starting with Windows XP. It was a staggering technological and cultural achievement, especially considering Microsoft developed and shipped it right in the middle of intense Department of Justice antitrust pressure, including the hostile rulings of 2000 and the eventual settlement proposals of 2001.

Technically, Windows XP achieved what many thought was impossible: it unified Microsoft's fractured codebase. For nearly a decade, Microsoft ran two completely parallel development tracks that shared an API but differed entirely in implementation: Windows NT for high-end enterprise systems and Windows 9x (Windows 95 and 98) for consumers. NT was powerful but required beefy, expensive hardware and lacked an intuitive feel, while the consumer 9x lineage looked friendly but was unstable and fundamentally weak. Windows XP killed off the aging 9x codebase entirely, successfully wrapping the approachable, friendly ease-of-use of Windows 95/98 directly over the robust, enterprise-grade NT kernel.

- **XP Editions:** The operating system launched with two primary editions: Home and Professional. Tech enthusiasts and PC builders routinely opted for "Pro" simply to get the most comprehensive features, even if they weren't working in a corporate office. - **The Digital Era:** The release placed a heavy emphasis on digital media—including digital photography, music, home networking, and video. While Apple branded this era as their "digital hub" strategy, XP machines became the actual workhorses for these tasks, heavily populated by millions of active Napster clients.

---

[02:01:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7314s) A Surreal Moment in Tech History: The XP Keynote

The official launch keynote for Windows XP stands out as one of the most bizarre and historically grounded product events ever recorded, completely removed from the polished aesthetic of a modern Apple WWDC or Google I/O. The presentation opened surreal caught-in-time imagery: a gospel choir singing "America the Beautiful," followed immediately by Bill Gates and New York City Mayor Rudy Giuliani walking onto the stage together to deliver remarks about the resilience of the city and the evils of terrorism.

The launch took place in New York City in October 2001, exactly one month after the September 11 terrorist attacks. Tech executives faced the agonizing dilemma of whether to cancel the launch entirely or reframe it to honor first responders. This heavy, historical backdrop anchored the commercial product launch to a profound moment in global history, creating an atmosphere unlike any other tech event. Amidst this backdrop, another subtle organizational shift was on display: though Steve Ballmer had already assumed the mantle of CEO, it was still Bill Gates who commanded the stage to introduce Microsoft's vision to the world.

[02:03:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7404s) The "Windows is Gospel" Strategy

The Windows XP launch kicked off with Bill Gates appearing alongside Rudy Giuliani. While it seemed natural for Gates, as the public face and founder, to lead the keynote, it highlights a strange organizational dynamic: why wasn't the CEO doing the keynote? Alongside XP, Microsoft executed a classic, highly effective move by releasing a brand-new version of Office on day one. By ensuring their powerhouse applications immediately adopted the latest Windows platform, Microsoft boosted the value of both the OS and the software suite.

This symbiotic relationship—where Windows success directly begot Office success—became an unshakeable gospel inside the company. During the earlier DOS era, rival applications like Lotus 1-2-3 and WordPerfect had completely "smoked" Microsoft in its own backyard because Microsoft's own productivity apps failed to gain real adoption on DOS. Swearing "never again," the company ensured tight integration moving forward. For years, this strategy was brilliant, but Microsoft may have held onto it a decade too long. Up until 2014, the prevailing corporate mindset was that as Windows goes, so goes the entire company.

- **Enterprise Ecosystem Integration:** As Microsoft transitioned into an enterprise-focused giant, Windows became the anchor of the "Enterprise Agreement." The core value proposition of Microsoft's server technologies was that they integrated seamlessly with Windows devices across a corporate network. - **The Disney Analogy:** The speakers compare Microsoft's absolute reliance on Windows to the old Walt Disney adage: "with animation goes the company."

[02:06:06](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7566s) Riding the Secular PC Wave

The Windows XP era represented the absolute showcase moment for Microsoft's platform strategy, backed by one of the greatest secular growth trends in business history: the explosion of the personal computer. Between Microsoft's founding in 1975 and its IPO in 1986, PC shipments grew at a staggering rate of 98% per year. Astoundingly, this massive tailwind was still pushing Microsoft forward in 2001.

The speakers express genuine amazement at how low computer penetration actually was in American households during the early decades. By the 1986 IPO, only about 10% of US homes had a PC (up from 8% pre-IPO). That number climbed to 37% by 1997, and reached 62% in 2003 after XP had been in the market for a couple of years. For a company holding what the US government legally defined as a monopoly, this meant Microsoft effectively acted as a tracker for the entire market—as the world bought computers, Microsoft grew automatically.

- **2003 Penetration:** The hosts note how crazy it feels that even in a relatively "modern" historical moment like 2003, nearly 40% of US households still did not own a personal computer.

[02:07:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7674s) The True Cost of Xbox

Recognizing that Microsoft fans might be "gripping their phones with all their strength" wondering when gaming would be mentioned, the hosts dive into the history of Xbox. Microsoft's gaming roots didn't start with a console; they began with Windows 95 and the rollout of DirectX. This incredibly clever set of APIs bypassed standard Windows layers to connect directly with hardware acceleration, putting Microsoft on the map for PC gaming and enabling legendary late-90s titles like *Quake*, *Counter-Strike*, and *Half-Life*. This software competency directly paved the way for Microsoft's high-stakes entry into the Home Console market with the launch of the original Xbox in November 2001—just weeks after Windows XP launched, and right in the middle of the company's frantic anti-trust battle with the Department of Justice.

However, the hosts argue that while the Xbox became a massive cultural phenomenon, it was a historical "folly" and a financial "sinkhole" for Microsoft's business. Reviewing historical financial data from the old "Entertainment and Devices" division, the business operated at near-break-even or outright losses for years. Five years after launch, in 2006, the division generated $4 billion in revenue but suffered a massive $1.4 billion operating loss. Even by 2010, at the steady-state height of the Xbox 360 era, it generated $8 billion in revenue but only pulled in $700 million in operating income. For a company generating roughly $20 billion in overall profit at the time, a few hundred million was practically a rounding error.

The real cost of Xbox wasn't capital—it was the drain on Microsoft’s elite talent. Reflecting on a conversation with Steve Ballmer, the hosts note that because Microsoft was never capital-constrained, money wasn't the scarce resource; time, talent, and focus were. The absolute best product designers, hard-driving cultural leaders, and engineers were poured into the Xbox division to build complex systems like Xbox Live (which reached 40 million subscribers by 2012). Had that world-class design sensibility and relentless work ethic been deployed across the rest of Microsoft's core software products, the company's trajectory might have looked entirely different.

- **The Ballmer Perspective:** When discussing failed or low-margin acquisitions, former CEO Steve Ballmer humorously remarked to the hosts, "We only lost money." - **The Gates Philosophy:** In past interviews, Bill Gates explicitly stated that financial investments were never limited by dollars, but strictly by whether the company had enough smart people to pull the project off. - **The Counterargument (Azure DNA):** The hosts concede that Xbox did provide two massive benefits: it kept Microsoft relevant to a younger consumer base, and building the infrastructure for Xbox Live created the pioneering online subscription DNA that directly birthed AWS-competitor Azure.

[02:14:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8097s) The Longhorn Disarray and the Road to Vista

The transition from the triumph of Windows XP to its successor, Windows Vista, is framed as a legendary corporate disaster. The journey began with internal code names tied to favorite Pacific Northwest ski destinations. Windows XP was codenamed "Whistler" (the famous mountain near Vancouver), and its theoretical, major next-generation successor was codenamed "Blackcomb" (the mountain right next to Whistler). However, Blackcomb quickly became a technically complicated, deeply hairy project, forcing Microsoft to push its timeline out.

To bridge the multi-year gap before Blackcomb, developers proposed a modest intermediate release. Naturally, they named this stopgap "Longhorn," after the famous Longhorn Saloon ski lodge nestled right between the two mountains. The plan was a fast, simple follow-up, but Microsoft made the fatal mistake of aggressively marketing Longhorn's conceptual features via public developer and fan blogs years before it had an actual ship date. Tech-enthusiasts—including a teenage David Rosenthal—were downloading custom desktop skins just to mimic the promised Longhorn UI sidebar and clock, while behind the scenes, the operating system was spiraling into five years of development hell.

- **The Security Sidetrack:** The timeline was severely disrupted because Windows XP, despite its popularity, was deeply insecure. Microsoft initially planned a quick three-month detour to build a patch, but ended up spending nearly two full years iterating on what became the legendary Windows XP Service Pack 2 (SP2) so enterprises could finally trust it against viruses. This massive effort completely tied up the engineering talent needed for the next OS. - **The Avalon/Aero Pivot:** Longhorn's initial core pillars included "Avalon," a new graphics engine designed to bake DirectX directly into the OS, utilizing raw GPU hardware acceleration. While visually impressive, it required heavy gaming-PC-spec hardware. This completely missed the market trend of the mid-2000s, leaving hardware OEMs furious because they desperately needed a lightweight OS to drive sales of low-spec netbooks. - **The Aqua Ripple:** When Vista finally did ship with its heavily accelerated, translucent "Aero" interface, the hosts joke that it was essentially a blatant attempt at "ripping off Mac OS's Aqua" design.

[02:20:37](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8437s) The Longhorn Fever Dream and the WinFS Pitfall

The development of the Windows operating system faced a fundamental architectural shift that ultimately did not pan out. The speakers compare this era to the "Chicago" (Windows 95) and "Cairo" days, recalling how Cairo was a theoretical, next-generation operating system that never actually shipped. With Longhorn (which became Windows Vista), Microsoft essentially tried the same highly ambitious approach again. Bill Gates was a massive fan of this technically complex vision, which centered around an object-oriented file system called WinFS. Under this architecture, the file system itself specified data types—such as a standardized calendar invite with native fields for date, time, and notes—allowing applications to plug directly into it. This meant the operating system could look inside files stored by applications to make everything easily searchable without constantly traversing directory trees.

However, WinFS, the third pillar of Longhorn alongside Blackcomb, became a monument to engineering wheel-spinning. The team held endless offsites and architecture reviews celebrating its potential, but in practice, application developers never actually showed any demand for it. Even worse, the OS development team wanted to bake.NET directly into the core bits of the operating system shipped in the box, putting it everywhere. This created a massive bottleneck; other groups felt they could not implement their features until the WinFS specification and implementation were finalized.

The organizational structure fractured as different groups compiled their own isolated sub-projects. They could run them independently, but when it came time to actually integrate the projects into a unified build of the operating system, it simply would not compile. After years of restarts and moving targets, they never successfully built the original vision of Longhorn. Instead, the engineers had to completely re-fork from an older Windows Server version and add features back piece-meal. By year four of development, the goal shifted from a revolutionary architecture to finding the minimum acceptable set of features to make the OS look and feel new enough to ship to consumers.

[02:24:02](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8642s) The Vista Disaster and the Culture of "Alchemy"

The launch of Windows Vista in 2006 was a verifiable disaster for Microsoft, though executive keynotes tried desperately to spin it otherwise. Even though Steve Ballmer had been CEO for six years, Bill Gates was still the one delivering the keynote. The speakers note that the presenters clearly had no idea what flagship features to highlight. They leaned heavily on the 3D "Alt-Tab" window switcher, desktop widgets, the sidebar, and the translucent Aero aesthetics. Meanwhile, the one feature users vividly remember was User Account Control (UAC)—a security measure that made sense in theory but completely overwhelmed users with constant dialogue boxes, prompting a massive user revolt as everyone tried to figure out how to turn it off.

The marketing message for the launch was "the wow starts now," but consumers and businesses found the release completely incoherent. The market flatly rejected it; by 2009—three and a half years after launch—roughly three-quarters of corporate PCs were still running Windows XP and refusing to upgrade to Vista. The situation was so brutal that Microsoft's OEM partners, facing consumers who refused to buy Vista machines, forced Microsoft to extend the rights to sell Windows XP on new machines for an additional two years.

This failure highlighted the worst aspects of the internal Windows culture at the time. One speaker, drawing on his background working in the Office division, notes that Office was strictly ship-date driven; they would set a release date three years in advance and hit it exactly using robust triage, escalation, and "zero bug bounce" processes run by PM, Dev, and Test triads. The Windows team, by contrast, resisted these software development principles under the guise that they were doing "systems-level programming" and "alchemy." This marked the failure mode of a culture that had historically enabled technical genius but had devolved into "the company smoking their own Supply." They believed they were smarter than the market, choosing to spend years fighting over hallucinated technical concepts and trying to force them onto consumers who had absolutely no interest in them.

[02:27:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8823s) The Exodus to Amazon and the Absence of the "Governor"

To salvage something shippable, Microsoft brought Brian Valentine over from the Exchange and Windows Server teams to manage the crisis. Valentine aggressively cut features and abandoned the core pillars of the original Longhorn vision, yet it still took an additional two years just to get Vista out the door. Immediately after shipping it, Valentine left Microsoft alongside a wave of other talented systems engineers. They drove "across the lake" to Amazon, where Valentine became a top senior executive leading the entire engineering platform team for Amazon.com.

Reflecting on this era, the speakers speculate on how the shadow of the government's antitrust case and Bill Gates's role transition disrupted the company. When Gates stepped down as CEO to become Chief Software Architect, it altered the feedback loop. While a CEO is forced to stay grounded by engaging with OEM partners, enterprises, and customers, Gates was left to focus purely on technical architecture. Without a business "governor" balancing his technical ambitions, the team fell into re-architecting Windows simply for the sake of re-architecting it.

The speakers emphasize that nobody should sell Gates short by labeling him merely a technical genius; he was also one of the greatest business minds and deal negotiators of all time. However, Microsoft suffered from having only one Bill Gates. While he excelled at law, engineering, and business development, he lacked Ballmer's genuine passion for empowering and building relationships with the enterprise.

- Brian Valentine left Microsoft immediately after delivering Vista to run Amazon's engineering platform. - The antitrust landscape heavily fractured the internal execution focus of Microsoft's leadership. - A massive influx of core systems talent migrated from Redmond to Amazon, laying the early groundwork for what would become AWS.

[02:29:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8986s) The "Mac vs. PC" Era and Losing Developer Hearts and Minds

The timing of the Vista release in late 2006 could not have been worse, coinciding precisely with the launch of Apple's iconic "Mac vs. PC" advertising campaign. Even though Mac's market share was tiny at the time—around 2% to 3% compared to roughly 8% in 2024—the ads hit a deeply sensitive nerve within the halls of Microsoft because so many of the jokes "were straight up true," such as the PC character constantly crashing. Apple was firing on all cylinders under Steve Jobs; the iPod was a smash hit, and the "digital hub" strategy was actively driving Mac sales, particularly among college students.

This era served as a critical training ground for consumers. While Vista came to symbolize friction and dialogue boxes, Apple spent years training the public on the brand promise that its products "just worked." This predictability built immense pent-up demand and set the stage for the iPhone. By 2005 and 2006, consumers had a visible glint in their eyes regarding rumors of an Apple phone, purely because existing mobile devices were so poor and Apple's brand had become synonymous with user-friendly design.

Ultimately, the Vista debacle marked the death of Microsoft as a relevant consumer technology leader and a compelling platform for third-party developers. Innovators completely stopped writing applications for the Win32 API. Unless a company specifically required a heavy desktop footprint due to sheer user volume, developers shifted entirely to writing web applications. The only people left building dedicated Windows applications were enterprise developers writing custom internal software and legacy giants like Adobe. Disruptive new web players, such as Facebook, completely bypassed building Vista apps.

- Apple's "Mac vs. PC" campaign successfully weaponized Vista's instability to define the Apple brand promise. - Microsoft completely lost the hearts and minds of independent desktop software developers during this cycle.

[02:33:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9216s) The Talent Black Hole and Steve Ballmer's Regret

The true cost of Vista was not just lost market share or developer interest; it was a devastating black hole for human capital. As the project faltered, it continuously sucked in executive talent and distinguished engineers from other thriving divisions to try to fix it. This total systemic distraction left Microsoft completely flat-footed just as the technology landscape was shifting toward social media, mobile, search, and a renewed browser war where Internet Explorer was about to fall off a cliff. While it is fair to note that Steve Ballmer was not a consumer-oriented technologist, the deeper issue was that Vista had taken Microsoft's best engineering talent completely off the field.

In an exclusive conversation with the hosts, Steve Ballmer later reflected on his tenure, explicitly noting that the Vista era was his single worst moment as CEO. He remarked, "my Acquisitions, my mistakes—we just lost money on the bad ones." Because money was an infinite resource for Microsoft, bad acquisitions did not ultimately matter; however, paralyzing a massive percentage of the company's elite engineering minds was near-fatal.

The disruption was so severe that it even pulled Brian Valentine away from the Exchange division, which had been absolutely dominating the enterprise market. The speakers marvel at the ultimate irony: Microsoft's top systems experts left the company to work for what was then considered an online book seller, only to build that book seller into the absolute market leader in enterprise cloud compute. The hosts tease that this narrative thread connects search, mobile, Windows 8, and the Zune directly into a cohesive story that culminates in the creation of Microsoft Azure.

- Steve Ballmer cited the Vista timeline as the absolute worst moment of his tenure as CEO because it sidelined his best assets. - The resource drain crippled Microsoft's agility in entering the mobile, search, and social markets. - The episode transitions into a sponsor shout-out for pilot.com, noting that its CEO, Wasim Daher, formerly interviewed Bill Gates while studying at MIT.

Sponsor Interlude: Focus on What Makes Your Beer Taste Better

The hosts share a brief historical throwback, noting a PDF linked in the episode sources where Bill Gates discussed Microsoft’s forthcoming 40 GB portable media center. Seamlessly transitioning into a sponsor slot for Pilot, the conversation pivots back to another Seattle neighbor often discussed on *Acquired*: Jeff Bezos. They recall the AWS-inspired axiom that startups should strictly focus on "what makes your beer taste better." In other words, companies should invest their finite time and resources solely into moving the needle for their product and customers while outsourcing everything else. Accounting stands out as the ultimate example of this rule; it must be handled by a professional to mitigate risk, yet it has zero impact on improving the core product.

Pilot solves this problem by setting up and operating a company's entire financial stack. Their services span everything from accounting, tax, and general ledgers to higher-level CFO functions like budgeting, investor reporting, and financial sections for board decks. Having served thousands of startups across Silicon Valley and beyond—including prominent names like OpenAI, Airtable, and Scale AI—Pilot possesses the unique capability to support companies as they scale into their growth phases.

---

The $47 Billion Acquisition That Wasn’t

The discussion moves into a core chapter on search history, playfully alternative-titled "an acquisition that wasn't." The hosts reflect on a "lost to history moment" from 2008: Microsoft's attempted $47 billion acquisition of Yahoo. Setting the context, they trace the landscape of programmatic advertising technology at the time. Two major entities led the charge in serving and targeting online search ads: market leader DoubleClick and its competitor, Aquantive. After losing the race to acquire DoubleClick to Google, a desperate Microsoft bought Aquantive for $7 billion—a move that went so poorly they eventually declared the entire venture a write-off.

Desperate for search market share, Microsoft’s internal initiatives like MSN Search and Windows Live Search were failing to make headway. Concurrently, Internet Explorer had utterly languished. Having completely "taken their eye off the ball" since the browser wars a decade prior, Microsoft left users and web developers saddled with a widely disliked browser that required tedious custom workarounds. Meanwhile, Firefox burst onto the scene around 2007, introducing game-changing features like tabs, which Internet Explorer and Safari lacked. Crucially, Firefox featured Google as its default search engine, highlighting an emerging reality: "search was the business for the browser."

* **Market Share Context:** Microsoft was effectively locked out of search momentum while Google weaponized browser partnerships. * **The Power of the Front Door:** Even if a company like Apple or Mozilla didn't own a search engine, managing the browser meant controlling user attention. It is noted that today, in 2024, rumors suggest Apple secures roughly $20 billion annually from Google just to remain the default search engine on iPhones—a line item captured as Traffic Acquisition Cost (TAC) in Google's financials.

---

Search Scale and The Great Airport Showdown

The hosts explore why Microsoft failed to grasp the magnitude of search early on. When Google debuted in 1998, widespread skepticism surrounded whether auction-based advertising could truly scale. Even after Google's highly profitable 2004 IPO, observers failed to recognize the staggering "returns to scale" inherent to search. A massive search volume yields superior user data, which optimizes search results. This data advantage creates marketplace liquidity, drawing in the highest-bidding advertisers and allowing the market leader to out-earn competitors "by a country mile." This flywheel funds heavier R&D and data center infrastructure, cementing dominance. The hosts marvel that "search as monetized by an ad-based auction is one of the world's true Marvels," prompting a joke that they must discuss a future Google episode privately "after the kids go to bed."

Back in 2002 and 2003, Microsoft treated search as a mere feature of MSN or Windows Live rather than the next paradigm shift. By 2008, realizing they needed a Hail Mary to stay relevant before launching Bing in 2009, Microsoft put a $47 billion offer on the table for Yahoo. The tense climax of the negotiations culminated in Yahoo founders David Filo and Jerry Yang flying to Seattle. Steve Ballmer met them right at the airport at Boeing Field. Ultimately, the deal collapsed amid conflicting reports. Bill Gates and Steve Ballmer looked the Yahoo founders in the eye and concluded they were being "jerked around" by leaders who had no desire to be part of Microsoft, fearing an expensive "organ rejection."

* **The Bull Case for the Yahoo Deal:** Though listeners might think $47 billion for Yahoo was absurd, Yahoo held a 15% search market share and two massive hidden assets: Yahoo Japan and a 40% stake in Alibaba. * **The Valuation Breakdown:** Because the Alibaba and Yahoo Japan stakes were worth over $30 billion combined, Microsoft was effectively paying only $15 billion for a 15% slice of the hyper-profitable search market. Given Alphabet’s 2023 revenue exceeded $300 billion, paying "a billion dollar per percent of market share" would have been an incredibly lucrative steal.

---

The Back-End Deal, Satya Nadella, and Cloud Foundations

Following the collapse of the acquisition, Yahoo's valuation cratered, eventually selling for a mere $5 billion to Verizon and Apollo. However, Microsoft still managed to pivot by signing a commercial deal where Bing powered Yahoo's back-end search. While Yahoo retained 88% of the ad revenue for the first five years and kept user-facing relationships, Microsoft captured the invaluable search data and advertiser liquidity. This massive influx of traffic forced Bing to evolve into a highly sophisticated, distributed computing system operating 24/7 with low latency—the exact infrastructure DNA required to build a modern cloud platform. Alongside Xbox Live's 40 million users and Microsoft's long-running consumer web service Hotmail (acquired in 1997), the technical ingredients for Azure were quietly falling into place.

The conversation then shifts to a striking bit of corporate history: a brief period where a young Satya Nadella led the Bing division. The hosts swap obscure Nadella trivia, noting he originally joined Microsoft in 1992 from Sun Microsystems as an evangelist for Windows NT. His very first product management role was on "Tiger Server," a project Born from the mid-90s "interactive television information superhighway fever dream" that never actually shipped. They laugh over how Nadella's path defied the standard corporate narrative, winding through marketing-adjacent product roles, Microsoft's CRM competitor Dynamics, BizTalk Server, and Bing, before he was ultimately plucked to run Server and Tools and, eventually, Azure.

* **The Alibaba Miss:** Had Microsoft completed the Yahoo purchase, they would have inherited the 40% Alibaba stake. At its 2014 IPO, that single stake was worth a staggering $92 billion, making it "one of the greatest Venture Investments of all time." * **The Shift in Business Models:** The segment closes on Microsoft's historical obsession with search. As a traditional software company relying on upfront license sales, Microsoft watched in awe as a new business model emerged that appeared to scale infinitely.

The Existential Shift from Selling Software to Monetizing "Everything"

The economics of the software business underwent a massive, existential shift with the advent of search and modern ad-based monetization. If a technology company has the choice between selling a retail copy of Windows to 7 billion people or letting those same 7 billion people use a search engine for free, the latter is vastly more lucrative. Monetizing via advertisers yields roughly 3x to 5x more money per user than traditional software licensing. This realization fundamentally changed how tech giants viewed their core products. Windows revenue was not magically disappearing, but the next generation of software economics was explicitly decoupled from selling licenses.

This structural evolution transitioned companies from selling software as a technology product to effectively "selling everything." The offline economy—encompassing everything from tents to airline tickets—is exponentially larger than the software economy alone. Because every single business requires customer acquisition, a massive pool of global capital is constantly funneled into advertising. When individuals audit their monthly credit card statements, the amount spent on actual software is a tiny fraction compared to general consumption. Consequently, even capturing a small sliver of that broader "everything else" budget via advertising inventory results in a pool of wealth far larger than any traditional software budget.

This reality highlights the divergent paths taken by Microsoft and Google regarding web browsers. Microsoft built Internet Explorer and used it defensively to protect their core operating system monopoly. They did not monetize the browser traffic directly; they used it to safeguard the OS licenses they sold. Years later, Google created Chrome, also offering it for free, but they knew exactly how to play offense. Google leveraged the browser to monetize every single bit of traffic flowing through it, executing a monetization strategy that was vastly superior to Microsoft's licensing model.

* **Monetization Multiplier:** Monetizing free users via advertising generates 3x to 5x more revenue per user than selling software licenses directly. * **Browser Strategies:** Microsoft used its dominant browser market share for defense (protecting Windows), while Google built a browser explicitly for offense (capturing and monetizing search traffic).

Microsoft’s $15 Billion Intertwined History with Facebook

In October 2007, Microsoft found itself in a vulnerable position. Having missed the initial search engine wave, the executive team realized that social networking was the next major paradigm shift arriving five or six years later. Regretting that they were not a bigger player in the online advertising landscape, Microsoft was determined not to let history repeat itself. The company knew it could not organically build a social competitor—humorously noting they "know better than to do Google+"—meaning they had to look outward. While Microsoft was not capital-constrained, they were severely constrained by talent, product DNA, and the specific brand authority required to win in social.

Microsoft's solution was an aggressive attempt to acquire Facebook outright. They put a highly complex deal structure on the table designed to let Facebook shareholders cash out over an extended period as the company's valuation scaled. At the time, news outlets reported the offer valued Facebook at a staggering $24 billion, which was remarkably high for a company only three years old and not long after Yahoo's famous $1 billion acquisition offer. Mark Zuckerberg famously did not even respond to the acquisition proposal, despite some of his lieutenants indicating to Microsoft that an offer in that range would at least be discussed.

Because an outright acquisition was dead on arrival, the two companies instead brokered a strategic investment and commercial partnership. In October 2007, Microsoft invested $240 million into Facebook in exchange for a 1.6% equity stake, implying a total valuation of $15 billion. Crucially, the deal granted Microsoft the exclusive right to sell international banner ads on Facebook until 2011. This gave Microsoft a massive injection of ad inventory to bootstrap the advertiser side of its marketplace. Financially, if Microsoft had sold that stake at Facebook's 2012 IPO, it would have represented a 7x return; holding until 2014 would have yielded a 14x return. However, cash returns were secondary to Microsoft's true strategic goals of acquiring execution focus and developer alignment.

* **Investment Details:** In October 2007, Microsoft invested $240 million for a 1.6% stake in Facebook, cementing a $15 billion valuation. * **Commercial Terms:** Microsoft secured the exclusive rights to sell Facebook's international banner advertising inventory through 2011. * **Talent Ecosystem:** The partnership fostered a friendly relationship, leading many notable Microsoft employees to transition to Facebook early on, such as VJ Raji (later COO of StatSig).

Bill Gates' Retirement and the Loss of Developer Excitement

In June 2008, Bill Gates stepped down from his full-time role as Chief Software Architect to focus entirely on the Bill & Melinda Gates Foundation, remaining only as the chairman of the board. This true retirement coincided almost exactly with a massive technological transformation: the release of Apple's iOS 2.0, the launch of the iPhone SDK, and the opening of the App Store. At the time of his departure, Time Magazine published a blistering, highly reductionist quote suggesting Gates was exiting technology at the right time because it was "not really a business for nerds anymore." The article argued that while Gates mastered the PC and internet revolutions, he lacked the intuition for aesthetic revolutions (the iPod), social revolutions (MySpace), or entertainment revolutions (YouTube).

While the podcasters dismiss the Time Magazine quote as overly harsh and personal, it accurately captured the prevailing public sentiment that Microsoft was becoming irrelevant. During the Windows 95 era, such a critique would have been unthinkable. However, Microsoft's fierce obsession with the Enterprise market caused them to completely fumble consumer markets and ignore the open-source web ecosystems where actual developer excitement lived. While Microsoft corporate clients were locked into proprietary infrastructure, independent developers were flocking to the LAMP stack (Linux, Apache, MySQL, PHP) to build platforms like Facebook.

This divergence exposed a critical flaw in Microsoft's armor: consumer excitement and developer excitement go hand in hand. If a platform does not inspire external developers, the parent company is forced to build every interesting application themselves, which is an impossible task. Furthermore, the launch of the iPhone sparked a massive user revolt against rigid corporate IT departments through "Shadow IT" and "Bring Your Own Device" (BYOD). Employees demanded the right to use beautiful, highly functional consumer hardware and software at work, completely shattering the enterprise lock-in that Microsoft spent a decade cultivating. Internal innovative projects at Microsoft, like the Courier tablet or the Kin mobile phone, were routinely stifled or killed by an unyielding corporate culture that insisted Windows must remain the center of the universe.

* **The LAMP Stack:** Linux, Apache, MySQL, and PHP defined the web development era of the mid-2000s, operating entirely outside of Microsoft's enterprise ecosystem. * **Corporate Gatekeeping:** Promising internal consumer innovations at Microsoft were routinely killed if they did not directly serve or make the core Windows operating system look great.

Windows Mobile and Steve Ballmer's Infamous iPhone Critique

Long before smartphones achieved cultural dominance, Microsoft and Bill Gates were deeply obsessed with mobile computing. Gates' vision of the future heavily featured natural user interfaces, multi-touch, pen computing, and tablet PCs—to the point where a primary bullet point on the 2001 Windows XP launch presentation was explicitly dedicated to mobile computing. To realize this vision, Microsoft developed Windows Mobile. However, Windows Mobile was designed in the image of their traditional desktop business model: it was an operating system built for handset OEMs to license, tweak, and install on devices that largely resembled physical blackberries with small screens and tactile keyboards.

Windows Mobile was built entirely for business users and corporate networks, prioritizing seamless integration with Exchange and desktop PCs. Microsoft viewed the OS merely as an ingredient for original equipment manufacturers (OEMs). Unlike the PC market, where Windows remained uniform regardless of whether it ran on a Dell or an HP, mobile carriers and handset manufacturers heavily modified Microsoft's code. When a customer bought a Windows Mobile phone, they were at the mercy of the hardware manufacturer's software optimizations. In an attempt to buy their way into consumer mobile culture, Microsoft even acquired Danger—the company behind the popular T-Mobile Sidekick phone—notably co-founded by Andy Rubin before he left to build Android, the very platform that would later upend Microsoft's mobile ambitions.

When Steve Jobs announced the iPhone in January 2007, it completely reset the industry paradigm. While consumers were left in awe by the most spectacular technology demonstration in decades, entrenched mobile incumbents like BlackBerry and Palm dismissed Apple's ability to compete. Famously, Microsoft CEO Steve Ballmer scoffed at the device, stating that it would "never work at $500" because it lacked a physical keyboard and was far too expensive for business users accustomed to subsidized $100 devices. While Ballmer was playing the role of defensive corporate salesperson, his critique missed a massive business model innovation: the upcoming carrier subsidies that would make the expensive device accessible to the masses. Microsoft viewed mobile through a legacy paradigm of cheap embedded systems, failing to realize the iPhone was a fully-fledged Mac scaled down to fit into a pocket.

> "It's never going to work at $500." — Steve Ballmer's definitive reaction to the iPhone announcement, showcasing Microsoft's reliance on the legacy enterprise handset paradigm.

[03:11:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11471s) The Shift to Windows Phone and the Metro UI

In the early days of embedded systems, hardware choices came down to a couple of cents either way, yet they ultimately determined whether a phone would sell. By 2008, iPhones were selling incredibly well; by 2009, apps were coming out, and sales were really picking up. This momentum finally prompted Microsoft to react. They decided to repurpose an old asset, Windows Mobile, to create a brand-new operating system called Windows Phone.

However, everything Microsoft had to optimize for was completely different. The new ecosystem expectation demanded a super high-quality user experience. Traditionally, Microsoft worked with hardware partners by making software work on the "crappiest Hardware you can think of"—a strategy similar to how Roku operates with embedded TV manufacturers today. To compete with the iPhone, Microsoft had to completely flip this playbook. They began dictating intense hardware requirements, backing the product with the Microsoft brand first, and defining a breakthrough user interface called Metro, which humorously traced its lineage back to the Zune.

[03:12:45](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11565s) Google's Disruptive Android Business Model

Microsoft initially tried to maintain its legacy business model with Windows Phone: selling the operating system to original equipment manufacturers (OEMs) and charging a royalty per device. Because the iPhone had set a high bar for quality, Microsoft assumed manufacturers could generate nice margins on good phones and would willingly pay a licensing fee. The problem was that Microsoft's competitive set had changed entirely. Google had acquired Android and transformed it from a BlackBerry clone into an iPhone clone. Crucially, the software was open source.

Google went to the exact same manufacturers Microsoft used to rely on—like HTC and Motorola—and offered an unmatchable three-part value proposition. First, the operating system was completely free. Second, manufacturers were given the source code. Third, Google used anti-Microsoft sentiment to pitch the deal, warning PC makers that they made zero profit dollars because all the value in the chain accrued to the software vendor. Google told them, "It's literally the same people who did that to the PC makers, why would you let them do that to you?" In a low-margin mobile business where every cent mattered, the difference between $0 and a licensing fee of even $5 was a mountain.

- **Google's Three Deal Points:** * $0 license fee (Free). * Full access to the source code. * Appeal to anti-Microsoft sentiment regarding historical PC profit margins. - **Monetization Strategy:** Google only cared about its services, which were monetized via advertising. - **The Services Bundle:** Google required manufacturers to use its services. While open source Android could theoretically be taken alone, access to the crucial Google Play Store required bundling all other Google services.

[03:15:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11710s) The Interconnected Path of Google Search and Mobile Domination

If one traces the logic step-by-step from around 2002—when Google's search business model hardened into a ludicrously profitable, high-value-per-user machine—it becomes evident that Microsoft was structurally positioned to lose in mobile. The circuitous but direct path required Google to protect its core business. Because Google made its money on search, step two was to capture all possible searches, step three was to own the "front door" to search, and step four was to secure the next major computing platform.

To guarantee that mobile search volume would always flow back to them, Google bought or invented a mobile operating system and gave it away for free. Unless Microsoft was willing to completely abandon its legacy software-licensing roots and adopt Google's advertising-driven business model, they were immediately screwed. Microsoft's true competitor in mobile was not Apple and the iPhone; it was Android.

- **The Droid Factor:** The door firmly shut on Microsoft when Verizon freaked out over Apple's exclusive carrier deal with AT&T. Needing an answer, Verizon put a "gajillion dollars" behind the Moto Droid advertising campaign, cementing a two-horse race between iOS and Android. - **The Culture Clash:** Microsoft did have Bing at the time and theoretically could have offered Windows Phone for free to monetize via search. However, it would have required a massive cultural shift to declare themselves an advertising company. At that point in time, Microsoft was fundamentally a software company that made its money via enterprise agreements and software licenses.

[03:18:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11890s) Google's War on Microsoft's Core Products

Following the same strategic playbook, Google began targeting other high-margin software products that Microsoft sold. They looked at Microsoft's core value propositions, calculated which ones were the cheapest per user to run, and gave them away for free to consumers. They targeted Outlook and Exchange with Gmail, and took aim at Word, Excel, and PowerPoint with G Suite (now Google Workspace).

Even if these free alternatives did not directly make a massive amount of money for Google, they served a dual purpose. They brought more data and a closer relationship to the user, driving more queries and platform interactions to feed Google's core ad engine. Furthermore, because Google made so much money from search, hammering Microsoft's core business lines was a win in its own right. Despite this success in the consumer world, Microsoft's lock-in remains as strong as ever in the enterprise space. Google has still not truly figured out how to be an enterprise company, and global enterprise spreadsheet work is still done in Excel, full stop.

[03:19:53](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11993s) Steve Ballmer's Hardware Pivot and the Dodge Camry of Operating Systems

While Steve Ballmer often misses out on historical credit, he realized early on that the profit pools in mobile were shaking out differently than they did in the PC era. Under Bill Gates, Microsoft enjoyed tremendous profits while PC manufacturers saw their margins squeezed to nothing because all the value accrued to the software platform. Ballmer recognized that future mobile hardware platforms were not guaranteed to yield those same software profits due to the disruptive pressure of Android. He aggressively pushed Microsoft into the hardware business, a notoriously difficult transition since it is fundamentally hard for a high-margin software company to enter a lower-margin hardware or cloud business. To his credit, Ballmer acted by releasing the Surface and attempting to buy Nokia.

Before Windows 8 and the Surface pushed Microsoft into hardware, however, the company achieved a massive operational victory with Windows 7, led by Steven Sinowski. Brought over from the disciplined, ship-on-time culture of the Office division after the Vista disaster, Sinowski completely turned the Windows organization around. He used a brilliant analogy, noting that Windows did not need a radical technical vision; it didn't need to be a wild "Dodge Viper" like the failed Longhorn project had attempted to be. Instead, it needed to be a "Toyota Camry."

- **Windows 7 Success:** Released as a predictable, stable, and highly functional operating system. It gave consumers and enterprise users exactly what they wanted: a modern version of XP or a fixed version of Vista that didn't crash devices or networks, featured a normal start button, a fast file system, and quick search. - **Organizational Overhaul:** Sinowski slimmed down the team into a disciplined Dev/Test/PM structure that proved Microsoft could successfully plan, execute, and hit a strict ship date.

[03:24:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12254s) The Arrival of the iPad and the Vision for Windows 8

With the infrastructure tightly in place after shipping Windows 7, the team began planning for the future. The design team kicked off the vision for Windows 8, placing a big bet on the idea that touch and tablets would become the dominant computing paradigm of the future. While Bill Gates had championed tablets for years, the concept hadn't yet manifested successfully in the market. However, the launch of the iPhone changed everything by proving multi-touch was viable. When Apple launched the iPad in 2010, the Windows 8 team felt completely validated, realizing their core product vision was tracking perfectly with the market.

At the same time, the iPad absolutely terrified Steve Ballmer and Microsoft's leadership. They had just watched Apple disrupt the phone market, and the iPad looked poised to do the same to Microsoft's core PC market. In his original keynote, Steve Jobs laid out a threatening vision where the PC would become the "pickup truck" and the iPad would become the everyday "car." While history eventually showed that the iPad would not replace the PC, in 2010, it looked like a catastrophic threat to Microsoft's enterprise stronghold.

To fight back, Windows 8 attempted to kill two birds with one stone. First, they wanted to out-Apple Apple by creating a touch-centric tablet OS to protect their core PC market. Second, they desperately needed to win back developers, who were increasingly building for the web rather than Windows. To appeal to these web developers, Microsoft chose to build the new Metro UI development environment for Windows 8 entirely around HTML5, leaning directly into the tech that developers were already using.

[03:28:07](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12487s) The "Two-Headed Dragon" and the Shock of Windows 8

The strategy behind Windows 8 emerged as a "two-headed dragon": establishing a new developer platform focused on HTML5 and touch-first interactions, while making it run smoothly on ARM processors—exemplified by Microsoft's own hardware, the Surface RT. The ultimate manifestation of this touch-first vision in the operating system was radical. The traditional desktop was relegated to being just an app. Upon booting up Windows 8, users were no longer met with a desktop, but a full-screen, tablet-tiled start screen.

While the speakers note that finding the traditional desktop wasn't technically difficult to figure out—requiring only about five seconds to learn that clicking the bottom left corner would collapse the start screen—the "shock value" for consumers was severe. One speaker analogizes the experience to buying a Toyota Camry expecting a standard car, only to find out it is a scooter. This fundamental shift mixed user interface metaphors, creating deep confusion when trying to switch between "Metro" modern UI apps and legacy desktop applications.

The origin of this confusing hybrid layout came down to internal corporate politics. The original vision for the live-tile touch interface was meant to ship strictly for tablets, while a separate, Windows 7-like desktop version was being prepared for standard PCs. However, word came down "from on high" that "Windows is Windows," declaring there could only be one operating system shipped across all devices. Because massive career capital, momentum, and political betting had already been poured into the HTML5 Metro UI developer community, Microsoft chose to smash both concepts together into a single product.

[03:30:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12636s) The Bungled Rollout and Public Backlash

While the resulting operating system wasn't quite as bad as Windows Vista, its rollout was thoroughly bungled. Interestingly, tech pundits who spent time with the software figured it out relatively quickly, but the public narrative was poisoned before reviews even dropped. The loudest opposition came from what Microsoft employees internally referred to as "the basement"—the 0.001% of power users who dominated online discourse and heavily tainted the product's reputation.

Original Equipment Manufacturers (OEMs) were equally frustrated. They hadn't signed up to build touch-screen laptops, yet Microsoft was putting all its marketing muscle behind a touch-optimized OS. Consumers were left asking whether good standard laptops even existed anymore, or if they were seriously expected to reach out and touch their desktop screens.

Furthermore, the hardware in 2012 simply wasn't ready to meet the dual demand of running on low-power mobile ARM processors while functioning as full-fledged laptops. The speakers point out that Apple succeeded with the iPad because it was built as a "scaled-up version of the iPhone, not a scaled-down version of the Mac." However, looking at the technology available in 2026, the speakers reflect on how different the landscape is today.

- **The Modern Tablet Dream:** The speakers lament the current separation of the iPad and MacBook, wishing Apple would use Apple Silicon to create a single "awesome pane of glass" that can do everything. - **The Dream Machine:** A speaker describes their ultimate setup as a 13-inch M3 MacBook Air that could flip all the way around like a Lenovo Yoga, executing a universal binary to effortlessly run iOS counterparts of Mac apps with perfectly synced, cached data.

[03:33:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12780s) The Commercial Failure and the Dark Era of Consumer Microsoft

Ultimately, Windows 8 was a complete commercial failure. The developer ecosystem for Windows 8 apps never galvanized, and by the 2013–2014 period, the developer community was right back where it started, with no interest in Microsoft's platform. Instead, all industry energy shifted toward web apps and mobile apps—neither of which belonged to the Microsoft ecosystem.

This period marked the definitive "death of Microsoft as a consumer company." The brand faced an undeniable string of high-profile consumer failures: Zune was dead, Bing held a minuscule market share, Windows Phone lost completely to Android, and Windows 8 crashed. Consequently, the company's stock price languished, stuck at around $30 for an entire decade. It was an incredibly dark period for the company's public image.

Yet, paradoxically, the company's financial statements were shining bright. Outside of the Great Recession in 2008, Microsoft's massive Enterprise engine "never had a down year," chugging along with tremendous revenue and profit growth right through the worst of the Windows public relations disasters. Behind the scenes, the real story of Microsoft's salvation was already humming: Azure and the Cloud.

[03:35:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12910s) The Secret Genesis of Project Red Dog (Azure)

The general public assumes that Steve Ballmer was simply obsessed with Windows, and that Azure only launched and succeeded once Satya Nadella took over. The speakers clarify that "that's not really what happened." The true origin of Azure dates back to 2004 and 2005, when Bill Gates was planning his retirement from the Chief Software Architect role. Ballmer, fully recognizing that he was not a deep technologist, knew he needed a "Bill" to handle the technical vision. Realizing that Gates could not be replaced by a single person, they looked for "two Bills."

Internally, Craig Mundy filled one slot, but for the other, they acquired Groove Networks to recruit Ray Ozzie—the legendary creator of Lotus Notes. Ozzie had an incredibly unique history; he originally built Lotus Notes at his own independent startup using Lotus merely as a publisher, which allowed him to maintain close, confidential relationships with Microsoft that even Lotus wasn't privy to.

When Gates officially announced his retirement timeline in June 2006, Ozzie was named his successor as Chief Software Architect. Gates and Ballmer essentially told him to look at their massive pool of assets and talent and figure out the next generation of technology. In response, Ozzie penned a foundational, prophetic memo in October 2005 titled *Internet Services Disruption*, which laid out the exact economic and technological realities of modern cloud computing.

> "The environment has changed yet again... businesses are increasingly considering what services-based economics of scale might do to help them reduce infrastructure costs or deploy Solutions as needed and on a subscription basis."

[03:38:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13128s) Bypassing Corporate Gatekeepers and Building the Cloud

By January 2006, with Ballmer's full blessing, Ozzie began recruiting for a highly secretive incubation project code-named "Red Dog." Crucially, this project was kept entirely outside of Microsoft's massive, existing Server and Tools division. Had it been birthed inside the Windows Server team, it would have been strangled by corporate antibodies. The Windows Server business model relied on selling on-premise licensing to software operating in a client's own data center, relying heavily on lucrative implementation partnerships with firms like Accenture and OEMs like Dell.

A cloud platform threatened this entire ecosystem. Furthermore, it completely broke internal Key Performance Indicators (KPIs). Everyone in the Enterprise division was compensated based on how many copies of Windows software they licensed. Ozzie’s new cloud strategy proposed spending billions of dollars in capital expenditure building out massive server farms and generating zero upfront licensing dollars, simply hoping people would rent the hardware later.

To build this architecture, Ozzie brought in the ultimate engineering heavyweights: Dave Cutler, the legendary architect of Windows NT, and Amitabh Srivastava, a powerhouse engineer who had helped pull Windows Vista across the finish line. Remarkably, Cutler built a brand-new hypervisor—the critical software layer that virtualizes hardware to run multiple tenants—completely from scratch by himself, without using any open-source code. Ballmer stood by the team, protecting the project and deploying billions of dollars to construct data centers, despite intense political pressure from internal Windows teams.

[03:43:58](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13438s) The Initial Limits of Windows Azure

While the creation of Azure was a massive undertaking, one speaker challenges the notion that it was a pure "bet the company" move in its earliest days. First, Microsoft was only risking cash, which was a resource they had in abundance. Second, the project's initial iteration carefully avoided threatening the existing Windows-centric paradigm. When it first debuted, it was called "Windows Azure," it ran Windows Server, acted strictly as a Platform as a Service (PaaS), and fiercely rejected open-source software.

At this point in history, Microsoft still viewed open source as a "complete cancer." Because Microsoft made its fortune charging for proprietary assets—from operating systems and development environments to programming languages—the rise of free open-source software represented an existential threat. While they successfully protected their business for years without embracing open source, it meant that day-one Azure was fundamentally limited. It would take until the 2014–2015 era before Microsoft would finally pivot to Infrastructure as a Service (IaaS) and allow developers to run Linux and the LAMP stack on their cloud infrastructure.

[03:45:39](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13539s) The Strategy Behind Azure and the Rise of Satya Nadella

Azure entered the cloud market with an aggressive, highly distinct point of view by positioning itself as a platform as a service (PaaS), contrasting sharply with AWS's infrastructure as a service (IaaS) framework. While there was initial pushback that Microsoft’s enterprise base would "organ reject this so hard," the company pushed forward, fast-following with software as a service (SaaS) applications. An early pilot program with the battery company Energizer successfully proved the model by delivering SharePoint and other productivity tools as a service.

By 2010, massive organizational shifts were underway. Chief Software Architect Ray Ozzie departed the company, but not before he and Steve Ballmer renamed the "Red Dog" project back to Azure and folded it into the server and tools business. In a historic moment designed to align the company internally, Ballmer gave a speech at the University of Washington declaring that Microsoft was "all in and we are betting the company on cloud and on Azure." Shortly after, Ballmer replaced division head Bob Muglia—who was running a highly successful $12 billion business growing 30% to 40% annually—because he wanted a leader unburdened by legacy success to build a "cloud first organization." Ballmer famously tapped a rising, highly talented executive from the Bing division to lead this transformation: Satya Nadella.

- **Energizer Pilot:** Early SaaS testing program utilizing SharePoint as a service. - **Bob Muglia's Success:** Muglia was crushing it as head of server and tools before being replaced; he later went on to become the CEO of Snowflake. - **Satya Nadella's Promotion:** Handpicked by Ballmer from Bing to gain the right executive exposure for a future CEO path.

[03:47:56](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13676s) The Financial Powerhouse of Intelligent Cloud

To understand how well this cloud transformation went, one only has to look at Microsoft’s modern financial reporting structure. The company splits its operations into three reporting segments: Productivity and Business Process (including Office 365 and LinkedIn), More Personal Computing (Windows, Surface, and hardware), and Intelligent Cloud (Azure). Today, Intelligent Cloud has become the undisputed crown jewel of Microsoft—standing as the largest segment by both revenue and profit, and tracking as the company's fastest-growing business while traditional segments like Windows experience declines. In fiscal year 2023 alone, the Intelligent Cloud segment pulled in a staggering $88 billion in revenue.

While critics occasionally note that the Intelligent Cloud segment artificially inflates its numbers by grouping legacy businesses like SQL Server and Windows Server alongside Azure, the hosts point out that this pairing was actually Microsoft’s ultimate weapon against AWS. In the early days of the transition, having a robust on-premise server business allowed Microsoft to walk into enterprise accounts and pitch a "hybrid cloud" alternative. It was a massive competitive advantage, assuring clients that their world-class public cloud infrastructure would interface perfectly with existing on-prem systems. The cloud market ultimately proved so vast that it acted as an absolute tracker on a secular tailwind, absorbing all other corporate missteps.

- **Fiscal Year 2023 Revenue:** Intelligent Cloud generated $88 billion. - **Three Segments:** Productivity and Business Process, More Personal Computing, and Intelligent Cloud. - **Hybrid Cloud Advantage:** Grouping Windows Server and SQL Server with Azure allowed Microsoft to dominate enterprise accounts via hybrid infrastructure.

[03:52:20](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13940s) Debunking the 'Loser Era' Myth

There is a prevalent public narrative that the Steve Ballmer era was a period of unmitigated loss for Microsoft, but the reality is that the foundation for their greatest victory was built right under his watch. Azure's development spent seven full years incubating under Ballmer before Satya Nadella ever took the reigns as CEO. The technical ingredients were quietly harvested from all across the company's existing portfolio: Bing provided expertise in distributed systems; Xbox Live contributed an always-on, real-time, low-latency cloud service serving 40 million subscribers; MSN brought the scale of a high-traffic web property with 750 million registered users; and Hotmail offered deep experience running massive web applications.

Embracing this new frontier required overcoming intense internal skepticism. An internal Azure doubter sitting on top of the highly profitable Windows monopoly would naturally ask why they should jeopardize a business with self-reinforcing network effects and massive switching costs. Moving to the cloud meant building massive, energy-expensive data centers that lacked the zero-marginal-cost perfection of traditional software. Clients could suddenly run open-source software or competitor products on Microsoft's servers without buying Windows licenses. Despite these terrifying structural realities, Ballmer gave Ray Ozzie’s incubation group the necessary "air cover" to build an orthogonal business model that ultimately saved the company.

- **Xbox Live Scale:** Offered real-time cloud infrastructure with 40 million subscribers. - **MSN Scale:** Managed 750 million registered users, proving web property capability. - **The Core Conflict:** Azure represented a non-zero marginal cost business that threatened to cannibalize the high-margin Windows ecosystem.

[03:56:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14174s) Inside Microsoft's Windows-Centric Inertia and the 'Shelf Party'

Despite the quiet successes of Azure, Microsoft clung to its Windows-centric identity for far too long, leading to intense internal counter-strategies. Ben shares a personal anecdote from his time working at Microsoft from 2011 to 2014. After interning on the word web app, he returned to work on a deeply secret, highly anticipated "Office for iPad" project. In 2013, a team of 200 engineers spent years getting the application ready for its ship date. Instead of a launch, however, the team was told the product was being shelved to protect the newly released Windows 8 and Surface tablet marketing push, which promised that real Office was "first and best on Windows." Instead of a ship party, the devastated team threw a "shelf party."

> "Oh this company has its head in the sand, this is ridiculous. What users want is we have a good version of Word, Excel, PowerPoint that people can run on their iPads and we've decided not to ship it to try and Advantage surface..."

With the benefit of hindsight, the hosts reflect on why an incumbent thinks this way. Holding the product back for a year didn't cost Microsoft a dollar in enterprise agreements, and rushing to support an iPad ecosystem that felt like an existential threat carried high risks with low immediate upside. Ultimately, holding the product gave Satya Nadella an incredibly easy cultural win during his first year as CEO. Nadella immediately shipped Office for iPad, using it as a high-profile proof point that Microsoft had officially shifted from a devices and services company into a "cloud first, mobile first" powerhouse.

- **Ben's Microsoft Tenure:** Worked at the company from 2011 to 2014. - **The Shelf Party (2013):** Office for iPad was fully completed but shelved to artificially protect the Surface tablet launch. - **The Pivot:** Nadella used the delayed release to instantly signal a massive corporate culture shift.

[04:00:04](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14404s) Android's 'Less Than Free' Model and the Nokia Acquisition

The conversation transitions to the dramatic conclusion of the mobile wars and the end of Steve Ballmer's tenure. By 2011, Microsoft's Windows Phone was fundamentally doomed to fail against Android because it simply couldn't compete with a "less than free" business model. As venture capitalist Bill Gurley famously pointed out at the time, Google wasn't just giving Android away for free—they were actively paying carriers, funding marketing campaigns like the "Droid" launch, and incentivizing retail salespeople to push Android over the iPhone.

Nokia was the only major phone OEM left willing to play ball with Microsoft. Nokia's legacy Symbian operating system had hit a dead end, and under the leadership of CEO Stephen Elop—a former Microsoft executive—the company signed a deal in February 2011 to adopt Windows Phone as its primary smartphone OS. By 2013, however, the strategy was failing. Nokia approached Microsoft with a stark ultimatum: they were going to defect to Android unless Microsoft bought them out. Ballmer viewed the subsequent $7 billion acquisition through a highly pragmatic lens, noting that it was "only money." At the time, Microsoft had a market cap of roughly $300 billion, meaning they risked just 2.3% of the company on a final, massive smartphone gamble.

- **Android's Subsidy Model:** Google subsidized Android adoption via carrier incentives and massive ad campaigns. - **Nokia Deal (Feb 2011):** Nokia abandoned its internal OS efforts to adopt Windows Phone under former Microsoft exec Stephen Elop. - **The $7 Billion Ultimatum (2013):** Nokia threatened to switch to Android, prompting Microsoft to acquire their hardware business for 2.3% of Microsoft's $300 billion market cap.

[04:03:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14611s) The Pragmatics of a Venture Capital Bet

The dynamics of Microsoft's financial positioning in 2013 reveal a stark disconnect between theoretical valuation and pragmatic business execution. While standard financial theory dictates that cash reserves contribute directly to a corporation's market capitalization, Microsoft's massive "cash flow geyser" was fundamentally unappreciated by Wall Street at the time, leaving the stock price deeply depressed. With an operating income of $27 billion in 2013 alone, the proposed $7 billion to $7.5 billion deployment for Nokia's mobile unit represented surplus cash that the company simply had no alternative use for.

From the perspective of leadership, spending uncredited cash on a high-upside gamble is a highly rational choice. If a strategic bet requires only cash—rather than diverting scarce organizational resources like top talent or executive focus—it operates essentially as a venture capital play. For Steve Ballmer, the calculation was simple: the market was giving him zero credit for generating billions in cash, so there was little to lose by risking a small percentage of the company's worth on a move that could fundamentally alter their mobile trajectory, especially while Nokia was "basically holding a gun" to Microsoft's head.

[04:04:58](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14698s) The Nokia Coup and the Changing of the Guard

The acquisition of Nokia's mobile unit triggered a profound internal crisis, exposing deep rifts between executive leadership, the division heads, and the board of directors. Satya Nadella's own account reveals that a straw poll conducted among the company's top leaders showed a clear majority opposed to the acquisition, with Nadella himself among the dissenters. When the board confronted Ballmer with this absolute lack of internal backing, a tense sequence of summer negotiations unfolded, culminating in a historic, rapid-fire operational shakeup.

* **August 23, 2013:** Steve Ballmer shocks the industry by announcing his retirement as CEO, effective within the next 12 months, triggering an immediate successor search. * **September 3, 2013:** Merely ten days later, Microsoft officially agrees to acquire Nokia's mobile unit for $7 billion, leaving an ambiguous historical legacy as to who truly drove the final deal across the finish line. * **February 4, 2014:** Satya Nadella is officially introduced as Microsoft's next CEO. Concurrently, Steve Ballmer steps down, Bill Gates resigns as chairman of the board, and John Thompson assumes the chairmanship.

This total transformation represented a wholesale "Changing of the Guard" that was strictly necessary to purge 40 years of accumulated organizational baggage, including old tech wars, antitrust scars, and disastrous Windows releases. By clearing the path of its original founders in one fell swoop, the company could finally embrace a culture reset and allow its massive, latent growth engine—Azure—to be fully appreciated by the market, safeguarding everyone's personal net worth.

[04:08:17](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14897s) A Decadal Look at the Numbers: The Ballmer Era Scorecard

An evaluation of Steve Ballmer's 14-year tenure requires separating raw operational performance from the brutal reality of market multiple compression. When Ballmer took the helm in 2000, he inherited an nearly impossible "garbage sandwich": the company's price-to-earnings (P/E) ratio stood at a staggering 75x, right at the peak of the dot-com bubble and immediately prior to the devastating Department of Justice antitrust verdict. Over his tenure, Ballmer dramatically scaled the core business, yet the stock remained stubbornly flat because massive profit growth was constantly offset by the market rationalizing that historical, unsustainable valuation multiple down to 14x.

| Metric | Year 2000 (Start) | Year 2014 (End) | Growth / Change | |---|---|---|---| | **Annual Revenue** | $23 Billion | $84 Billion | 3.5x Increase | | **Operating Income** | $12 Billion | $30 Billion | ~3x Increase | | **P/E Multiple** | 75x | 14x | Compressed by ~80% | | **Market Capitalization** | $600 Billion | $330 Billion | Decreased by ~45% |

The investor skepticism during this era was so severe that Microsoft stock was practically given away on the cheap; buyers could have acquired the business for a mere 2.1x annual revenue in 2009, and just 3x revenue as late as 2013. Wall Street fundamentally doubted the stickiness of the core Windows and Office franchises, incorrectly assuming they would easily fall victim to new technological eras, while entirely missing their incredible, long-term durability.

[04:14:22](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15262s) The Power of Storytelling and the Peer Survival Rate

The core failure of the Ballmer era was not operational, but narrative: "telling stories is the most important thing," and Microsoft simply failed to communicate a cohesive vision to consumers, developers, or Wall Street. While corporate buyers understood the product value, the broader market was distracted by high-profile consumer stumbles like the Zune losing to the iPod, searches being captured by Google, and systemic losses in mobile. This highlights a critical lesson for any corporate leader: you must repeat your core strategic message "about 10 times more than you think it is" necessary. When Satya Nadella took over, his immediate brilliance was resetting the entire narrative around the clear, rhythmic mantra of "mobile first, cloud first."

Despite these narrative failures, a macro analysis against the 2000 peer set proves that Ballmer's tenure was an extraordinary success of capital preservation. While iconic tech and media giants of that era like Yahoo, AOL, HP, and Nortel completely collapsed or fractured, Microsoft survived alongside Dell and Oracle, maintaining its core talent and business continuity over decades. Ultimately, while consumer-facing products languished and lacked user excitement, the unglamorous enterprise engines met customer needs perfectly and preserved the foundation for Microsoft to eventually reclaim its title as "The Once and Future King"—the most valuable company in the world, scaling to a $3.5 trillion market cap 10 years later.

[04:19:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15549s) Structural Analysis: Applying the Seven Powers

Transitioning into a formal business analysis using Hamilton Helmer's *Seven Powers* framework reveals how Microsoft maintained persistent, differential returns over its competitors. The framework outlines seven distinct categories that enable sustainable, long-term profitability: counter-positioning, scale economies, switching costs, network economies, process power, branding, and cornered resources. While the Microsoft of the 1990s commanded nearly all of these dynamics simultaneously, its position as a massive incumbent by the 2000s fundamentally altered its strategic options.

* **Counter-Positioning Limitations:** As an entrenched market incumbent, Microsoft was structurally blocked from utilizing counter-positioning broadly, and was instead actively counter-positioned by nimbler rivals—such as Google offering mobile operating systems entirely for free to undermine Windows Mobile. * **The Azure Exception:** Microsoft successfully leveraged a form of counter-positioning against AWS during the early development of Azure, pitching itself to Fortune 500 companies as a trusted, hybrid-cloud partner that understood legacy enterprise infrastructure in a way Amazon could not replicate. * **The Stickiness of Legacy:** Despite intense user and developer dissatisfaction with consumer products during this period, the underlying enterprise business lines retained bulletproof switching costs and network economies that insulated the core financials from macro disruption.

[04:20:42](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15642s) The Powers of Scale Economies and Enterprise Moats

Microsoft's single greatest asset during this era was its massive scale economies. Any investment the company made could be amortized over an enormous user and customer base, making heavy R&D spending and strategic acquisitions highly lucrative. This dynamic translated directly, and perhaps even more intensely, into the cloud era where returns on scale economics are incredibly high. However, the company noticeably lost its process power and consumer branding during this period. The "Blackcomb, Longhorn, Vista thing" serves as a primary example of this decline. Microsoft went from shipping the deeply beloved Windows 95 and repeating that success under intense antitrust pressure with Windows XP—with virtually no blunders in between, aside from the superficial "fresh coat of paint" that was Windows ME—to completely forgetting why those franchises held economic value when they engineered Longhorn and Windows 8.

Conversely, Microsoft drastically strengthened its trust and power within the enterprise market. If offered the exact same service from Microsoft and AWS, enterprise clients were overwhelmingly more willing to pay Microsoft due to unbelievable switching costs rooted in Enterprise Agreements (EAs). This reality exposes a fascinating irony regarding the Department of Justice's historical antitrust focus. While the government and the DOJ fiercely targeted Microsoft's product tying on the consumer side, the true monopoly power was being cemented via product tying in the enterprise sector. Under these corporate frameworks, companies paid a flat dollar amount per network device to receive a sweeping bundle of all Microsoft software. Even though superior point-solution software existed for hundreds of individual features, switching away from the bundled ecosystem was virtually impossible for IT departments.

- **Dial-Up Switching Costs:** In the early internet era, literal switching costs made bundling a browser helpful for consumers, as downloading an alternative like Netscape could take anywhere from 5 to 24 hours over dial-up. - **Eroding Network Economies:** Traditional operating system network economies began to thin during this era as interoperability and file format standardization took hold, allowing the same documents to open seamlessly on both Macs and PCs. - **The Seven Powers Check:** Evaluating the company against classic business strategy frameworks reveals that Microsoft lacked meaningful counterpositioning or cornered resources, leaving scale economies and high switching costs as their sole structural moats.

[04:24:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15871s) The Zero-Sum Culture and the "Guns at Each Other" Org Chart

The flatlining of Microsoft’s stock price over a prolonged period triggered a devastating psychological and cultural shift inside the company. When the stock remained stagnant year after year, employees became deeply convinced that it would simply stay flat forever. This belief warped the corporate environment into a toxic, zero-sum game. Because workers felt that nothing they did could increase the total value of the company, they concluded that the only way to grow their personal value and advance their careers was "to win at the expense of someone else at this company." Teams and products began aggressively devouring one another, and promotions were secured by making immediate peers look foolish.

This internal warfare perfectly manifested in the famous cartoon organization chart of Microsoft, which depicted the various product divisions pointing pistols at one another. This hyper-competitive toxicity was severely amplified by the company's aggressive stack ranking system. Under this performance evaluation framework, every employee was rigidly graded on a scale from one to five. Because managers were strictly capped on how many "ones" and "twos" they could hand out, colleagues were forced into head-to-head bake-offs against their close friends every six to twelve months. Because the economic pie wasn't growing, internal destruction became the default mechanism for individual survival.

[04:26:26](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15986s) The Strategic Failure of Fighting the User Tide

Throughout the 2000s, Microsoft exerted massive energy trying to fight the natural tide of technology, aggressively resisting the rise of open-source software and the open web. The company attempted to erect high barriers to steer users forcibly back into its proprietary ecosystem. Ironically, the modern playbook that Satya Nadella deployed to revitalize the company is a direct return to a classic Microsoft strategy: "Embrace and extend." Instead of fighting user behavior, a successful strategy requires figuring out what people want—whether it's using iPhones, hosting web apps, or running open-source software—embracing it fully, and then inventing a clever business model that extends that behavior to capture revenue.

This antagonistic relationship with users heavily impacted Microsoft's enterprise business, which historically focused on controlling corporate networks and preventing employees from altering them. When the iPhone launched, that restrictive dam broke entirely; corporate IT departments could no longer hold back employees from using the devices they wanted. Shifting to the cloud became a massive strategic triumph because it forced Microsoft to stop paying mere lip service to customer satisfaction and actually become a partner to its users. The ultimate challenge of this transition was navigating the underlying economics of user desire: consumers fundamentally want free value, but a business cannot survive giving a dollar away for ninety cents. Winning required aligning the business model so that users were genuinely happy with the value creation and willing to pay premium rates.

[04:28:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16134s) Strategic Accuracy vs. Tactical Taste and Execution

Looking back at Microsoft's product roadmap from Windows 98 onward, a harsh reality emerges: the company possessed brilliant long-term foresight but suffered from terrible timing, implementation, and a profound lack of taste. Strategically, Bill Gates was incredibly accurate about the trajectory of technology. He repeatedly championed the "natural user interface" long before touch computing took off; he accurately predicted interactive television; and he knew mobile would dominate the computing landscape. The fatal flaw was that Microsoft began building these systems 5 to 20 years before the underlying hardware and infrastructure were mature enough to support them, causing them to bet on the wrong paradigms.

- **Touch Computing Missteps:** Microsoft built touch interfaces around resistive screens requiring a stylus, whereas the winning market standard turned out to be capacitive multi-touch. - **Tablet Architecture:** Microsoft treated tablets as scaled-down desktop PCs, while history proved they were actually scaled-up cousins of smartphones. - **Interactive TV and Bandwidth:** Gates envisioned interactive TV arriving before the widespread internet as a layer on top of cable companies like Comcast. In reality, it required massive internet bandwidth to support ad-hoc streaming platforms like YouTube and Netflix. - **Mobile Paradox:** Microsoft’s early mobile efforts arrived five years too early and treated phones like embedded devices rather than pocket-sized computers.

Historically, Microsoft mitigated these risks through an internal practice known as "bracketing." This strategy involved developing two separate products concurrently—one aimed below current technical capabilities and one aimed ambitiously above. As shipping deadlines or market realities approached, executives would choose whether to elevate the low-end product or trim features from the high-end version. In the IBM era, this was Windows versus OS/2; in the internet era, it was the web browser versus interactive TV; and during the Longhorn development cycle, it pitted a minor, iterative update against the hyper-ambitious, ultimately cancelled Blackcomb project. During this lost era, however, this crucial operational optionality collapsed, leaving the company unable to execute its great insights at the right time.

[04:32:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16323s) Positive-Sum Leadership and the Curse of Ecosystem Externalities

The partnership between Bill Gates and Steve Ballmer represents a classic case of positive-sum leadership, where the duo was profoundly more effective together than apart. When humming in unison, their collaboration was a clear case of "one plus one equals five." Left entirely to his own devices, Gates was prone to getting trapped in "theoretical technologies," with the over-engineered WinFS file system serving as the perfect example of his tendency to drift into academic engineering. Ballmer, on the other hand, desperately needed a premier technology partner who commanded the absolute, extreme loyalty of Microsoft's thousands of brilliant engineers to align them behind a single corporate vision.

The two leaders possessed beautifully complementary styles regarding data and momentum. Gates was a relentless machine for processing new information; if presented with compelling data, he was completely willing to pivot the entire company overnight regardless of how much momentum an existing project had, famously illustrated by his rapid "Internet Tidal Wave" memo. Conversely, Ballmer viewed the company as an massive aircraft carrier that required immense operational alignment to steer, alongside an entire ecosystem of external partners. Together, they balanced adaptability with execution. However, this massive external ecosystem eventually became a golden cage. While Microsoft generated historic profits for its independent software vendors, resellers, and retail partners, the sheer weight of these dependencies made it impossible to hit the reset button. When Steve Jobs returned to Apple, he executed a total developer and product purge because Apple had lost the desktop war and had nothing left to lose. Microsoft could not adapt cleanly to mobile because it was crushed by the massive switching costs of preserving its legacy ecosystem.

[04:35:21](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16521s) Landing the Plane: The Narrative Failure of the "Lost Years"

Reflecting on the exhausting mountain of historical research and interviews, a powerful, contrasting conclusion emerges about this specific era of Microsoft's history. The prevailing public narrative that defines these years as a period of absolute, uniform failure for the company is fundamentally incomplete and inaccurate. While there were undeniable and high-profile losses, characterizing this stretch as entirely "the losing years" misses the monumental foundational victories that occurred away from public view.

The ultimate tragedy of Microsoft during this timeframe was not a failure of business engineering, but a failure of storytelling. "This was the biggest failure of the company during this period: they did not tell their story right." Much of the modern perception of their historic decline—and the entirety of the psychological weight baked into the frozen, flatlined stock price—stems from this communication breakdown. Steve Ballmer stepped into the chief executive role at the absolute zenith of the historic tech bubble, inheriting a hyper-inflated valuation multiple that was destined to correct regardless of performance, masking the massive operational engine he was actively building underneath.

[04:37:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16634s) The Power of Corporate Narratives: Microsoft vs. Amazon

The hosts dive into why Microsoft's stock price languished in the $30s throughout Steve Balmer's entire tenure. They argue it wasn't a lack of achievement, but a failure to "tell the story right." During this period, the public narrative devolved into a self-reinforcing "Microsoft sucks" loop, branding the tech giant as an irrelevant, failing entity incapable of conquering the consumer market. They contrast this with Amazon, which successfully advanced a consistent identity: "we are a company who invents and wanders."

This corporate positioning allowed Amazon to survive massive, public failures—such as their high-profile smartphone collapse, various grocery iterations, Amazon Go, local initiatives, and restaurant delivery services—while maintaining a reputation for bold, commendable risk-taking. Even the long-term status of Alexa is debated, with the hosts noting it might just serve as an eventual distribution option for large language models (LLMs). Ultimately, while Amazon's narrative was celebrated as a beautiful philosophy imbued by Jeff Bezos, Microsoft was trapped in a narrative where they simply "can't get anything right."

[04:38:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16734s) Tech Disruption and Balmer’s True Legacy

Shifting to alternative takeaways, the hosts reflect on Bill Gates's core philosophy: technology companies are perpetually at extreme risk of disruption. Even the most dominant players can easily lose their footing and become irrelevant tomorrow. While sticky enterprise models allow legacy giants like IBM to generate massive revenues for decades, Microsoft faced a fast-moving, dynamic landscape. The hosts argue that even without the Department of Justice antitrust distractions, Microsoft was fundamentally structured to miss the mobile shift. They could not have anticipated Google's disruptive strategy of giving away mobile software "for less than free."

Because the technological landscape moves so rapidly, the hosts emphasize that Steve Balmer and his era deserve far more credit than they typically receive. Unlike IBM, Microsoft did not fade into large-scale irrelevance; instead, they successfully positioned themselves to become the most valuable company in the world through Azure, the cloud, and AI. Balmer's era effectively provided a "free option" to dominate the modern tech landscape while simultaneously tripling both revenue and profits.

[04:40:41](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16841s) Carve-outs: From Sleep Buds to the Model Y

The conversation transitions into the show's traditional "carve-outs" segment, highlighting favorite recent products. The first recommendation is a re-carve-out of the Ray-Ban Meta smart glasses. The ambient audio feature is praised as a brilliant alternative to traditional earbuds, specifically serving as a perfect tool for monitoring a baby crib while still remaining present to talk with family and friends. A related recommendation is the Oslo sleep buds, a product engineered by a former Bose team that left to launch a startup after Bose discontinued a similar line. Built specifically for side-sleepers, these tiny buds do not protrude or jam into the ear, prompting an immediate request for a purchase link.

The recommendations stick to highly functional, everyday tech as the next choice is the M3 MacBook Air. Described as an incredibly fast, lightweight, and beautiful machine, it has become the ultimate on-the-go travel companion, a major relief compared to lugging a heavy 16-inch M1 MacBook Pro on flights. Finally, the Tesla Model Y receives high praise following a weekend trip to Orcas Island. The car handled multiple hours of driving, a ferry ride, and island navigation without a single charge, returning home with 18% battery remaining. The hosts celebrate the vehicle's speed and its seamless, "iPhone with wheels" design.

- **Ray-Ban Meta Glasses:** Excellent ambient audio utility, specifically useful as a non-isolated baby monitor. - **Oslo Sleep Buds:** Designed by former Bose engineers specifically for side-sleepers to comfortably stream audio at night. - **M3 MacBook Air:** Praised for its mobility and speed over older, bulkier 16-inch MacBook Pro models. - **Tesla Model Y:** Noted for its extensive battery range on an island road trip and its robust supporting Supercharger network.

[04:43:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=17034s) Acknowledgments and Expert Contributions

The episode wraps up with a massive round of gratitude for the show's core sponsors—JP Morgan Payments, ServiceNow, and Pilot—alongside an extensive list of over 20 industry insiders who provided research and firsthand accounts for the episode. The hosts shout out legendary Windows leaders like Brad Silverberg (Windows 95) and Steven Sinowski (Windows 7 and 8, Office), whose 1,000-page textbook-like blog *Hardcore Software* served as a foundational text. Other critical contributors included early Internet Explorer developer Thomas Reardon, Office for iPad collaborator Anen Rajaswaran, and former Microsoft corp dev strategist Fritz Lanman, who now serves as the CEO of the combined ClassPass and Mindbody.

The thank-you list extends to brilliant tech minds like investor Charlie Songhurst, former Apple and Palm engineering executive John Rubinstein, and Ray Ozzie, who charmed the hosts over Zoom by showing off museum-worthy 1970s and 80s hardware from his office. Deep insights were also provided by RealNetworks founder Rob Glazer, long-time Windows Phone frontman Joe Belfiore, and antitrust comms expert Vi Vara. Crucially, Steve Balmer himself was thanked for his generosity with his time, balancing the interview alongside his current "day job" running the LA Clippers. Historical data spreadsheets were supplied by the *Science of Hitting* substack, while archival, unreleased reporter recordings of Bill Gates and Steve Balmer from the early 2000s were provided by Todd Bishop of GeekWire.

- **Brad Silverberg:** Led the development of Windows 95. - **Thomas Reardon:** Original IE team member; later founded CTRL-labs (acquired by Meta). - **Steven Sinowski:** Author of the massive insider chronicle *Hardcore Software*. - **Julie Larson-Green:** Worked intimately on user experiences for Office and Windows. - **Ray Ozzie:** Legendary tech creator now running the startup Blues Wireless. - **Joe Belfiore:** Famous for iconic Microsoft live software demonstrations alongside figures like Jay Leno and Regis Philbin. - **Mary Jo Foley:** Veteran tech journalist and Editor-in-Chief at *Directions on Microsoft*. - **Dave Marquardt:** The sole outside venture investor (from TVI) before Microsoft's IPO, serving on the board for 33 years.

[04:50:01](https://www.youtube.com/watch?v=6oER1bdQa_s&t=17401s) Upcoming Live Event and Farewells

For listeners hungry for more content, the hosts recommend checking out *Microsoft Volume 1*, their AWS coverage, and their *Nvidia* series, which intersects cleanly with the PC gaming evolution of this era. They issue a final energetic plug for their upcoming massive live event, Acquired Cleanroom, happening on September 10, 2024, at the Chase Center in San Francisco. Featuring a live appearance by "Mark freaking Zuckerberg," the event is framed as the ultimate gathering for business and technology nerds—a West Coast equivalent to the iconic Berkshire Hathaway annual meeting in Omaha. With that final invitation, the hosts sign off to their traditional outro music.

## Notable quotes > [17:41](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1061s) When I heard Steven talk about what was happening at Cornell, I began to take the internet quite seriously. — Bill Gates

> [23:51](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1431s) You go to school, you do your research, you leave and then they try and sue your business. Had I known this would happen, I would have gone to Stanford. — Marc Andreessen

> [35:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2115s) Perhaps you have already seen memos from me or others here about the importance of the internet. I have gone through several stages of increasing my views of its importance. Now I assign the internet the highest level of importance. In this memo I want to make clear that our focus on the... — Bill Gates

> [35:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2148s) reduce Windows to a poorly debugged set of device drivers — Marc Andreessen

> [01:28:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5311s) developers, developers, developers — Steve Ballmer

> [01:30:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5403s) It's time to make peace. — Brad Smith

> [02:03:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7404s) with animation goes the company. — Walt Disney

> [02:07:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7674s) We only lost money. — Steve Ballmer

> [02:33:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9216s) my Acquisitions, my mistakes—we just lost money on the bad ones. — Steve Ballmer

> [02:33:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9216s) what makes your beer taste better — Jeff Bezos

> [02:33:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9216s) It's never going to work at $500. — Steve Ballmer

> [03:38:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13128s) The environment has changed yet again... businesses are increasingly considering what services-based economics of scale might do to help them reduce infrastructure costs or deploy Solutions as needed and on a subscription basis. — Ray Ozzie

> [03:56:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14174s) Oh this company has its head in the sand, this is ridiculous. What users want is we have a good version of Word, Excel, PowerPoint that people can run on their iPads and we've decided not to ship it to try and Advantage surface... — Ben

> [04:35:21](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16521s) This was the biggest failure of the company during this period: they did not tell their story right.

> [04:37:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16634s) we are a company who invents and wanders

## Mentioned Ben Gilbert, David Rosenthal, Microsoft, Ballmer, Satya Nadella, Windows XP, Xbox, Zune, Department of Justice, Chase Center, JP Morgan Payments, Mark Zuckerberg, Meta, Hit Refresh, Manu Cornet, Windows 95, Jay Leno, Rolling Stones, CompuServe, Prodigy, AOL, H&R Block, Steve Case, Paul Allen, Project Marvel, MSN, Information Superhighway, Comcast, Time Warner Cable, DirecTV, Cablesoft, Tiger, Silicon Graphics, SGI, Jurassic Park, Jim Clark, John Malone, TCI, Gerald Levin, Time Warner, Michael Ovitz, MSNBC, The Road Ahead, Netscape, James Allard, J. Allard, Mosaic, National Center for Supercomputing Applications, NCSA, University of Illinois, Marc Andreessen, Steven Sinowski, Cornell University, Electric Media, Nintendo, N64, Kleiner Perkins, John Doerr, Mosaic Communications Corporation, Mosaic Navigator, Spyglass, Booklink, Internet Explorer, Thomas Reardon, Windows 97, Plus Pack, Apple, Macworld, Steve Jobs, Microsoft Office for Mac, Safari, Sun Microsystems, Oracle, Google, Facebook, Federal Trade Commission, FTC, Office, OEMs, Windows Explorer, Thomas Penfield Jackson, Sherman Antitrust Act, YouTube, David Boies, World War 3.0, Steve Ballmer, Standard Oil, iOS, Clinton Administration, Bush Administration, Windows 98, NT, RealNetworks, European Union, Brad Smith, SQL Server, Active Directory, Exchange, Dynamics, SharePoint, Kevin Kelly, Wired, BackOffice, DreamWorks, IBM, Intel, Dell, Outlook, Lotus Notes, Enterprise Agreement, ServiceNow, Fortune 500, Mac, Knowledge, Now Assist, Microsoft Co-pilot, Microsoft Teams, Office 2003, UBS, miniclip.com, Lenovo, iPhone, Windows NT, Windows 9x, Napster, Bill Gates, Rudy Giuliani, New York City, Lotus 1-2-3, WordPerfect, DOS, DirectX, Quake, Counter-Strike, Half-Life, Xbox Live, Azure, AWS, Windows Vista, Whistler, Blackcomb, Longhorn, Longhorn Saloon, Service Pack 2, Avalon, Aero, Aqua, Mac OS, Chicago, Cairo, WinFS, .NET, User Account Control, Brian Valentine, Windows Server, Amazon, iPod, Win32 API, Adobe, Pilot, Wasim Daher, MIT, Jeff Bezos, OpenAI, Airtable, Scale AI, Yahoo, DoubleClick, Aquantive, MSN Search, Windows Live Search, Firefox, Windows Live, Bing, David Filo, Jerry Yang, Boeing Field, Yahoo Japan, Alibaba, Alphabet, Verizon, Apollo, Hotmail, Tiger Server, BizTalk Server, Server and Tools, Windows, Chrome, Google+, VJ Raji, StatSig, Bill & Melinda Gates Foundation, iOS 2.0, iPhone SDK, App Store, Time Magazine, MySpace, LAMP stack, Linux, Apache, MySQL, PHP, Courier, Kin, Windows Mobile, HP, Danger, T-Mobile Sidekick, Andy Rubin, Android, BlackBerry, Palm, Windows Phone, Metro UI, Roku, HTC, Motorola, Google Play Store, Moto Droid, AT&T, Gmail, Word, Excel, PowerPoint, G Suite, Google Workspace, Surface, Nokia, Windows 7, Office division, Vista, iPad, Windows 8, HTML5, ARM processors, Surface RT, Lenovo Yoga, M3 MacBook Air, Apple Silicon, Great Recession, Project Red Dog, Craig Mundy, Groove Networks, Lotus, Accenture, Dave Cutler, Withamitabh Srivastava, Windows Azure, Energizer, Bob Muglia, Snowflake, University of Washington, Office 365, LinkedIn, Intelligent Cloud, Ben, Symbian, Stephen Elop, Bill Gurley, Wall Street, John Thompson, Nortel, Hamilton Helmer, Seven Powers, Windows ME, PCs, Netflix, OS/2, Ray-Ban Meta, Oslo Sleep buds, Bose, 16-inch M1 MacBook Pro, Tesla Model Y, Orcas Island, Brad Silverberg, Hardcore Software, CTRL-labs, Anen Rajaswaran, Fritz Lanman, ClassPass, Mindbody, Charlie Songhurst, John Rubinstein, Ray Ozzie, Blues Wireless, Rob Glazer, Joe Belfiore, Vi Vara, LA Clippers, Science of Hitting, Todd Bishop, GeekWire, Julie Larson-Green, Mary Jo Foley, Directions on Microsoft, Dave Marquardt, TVI, Nvidia, Acquired Cleanroom

[Source](https://www.youtube.com/watch?v=6oER1bdQa_s)

中文简报

鲍尔默掌舵时期微软营收翻三倍，企业业务筑牢生存根基尽管在消费移动领域遭遇备受瞩目的失败且饱受反垄断冲击，鲍尔默激进的企业捆绑策略和多年期企业协议仍构建起强大的财务引擎，不仅捍卫了公司的垄断地位，也为其未来的云端主导地位提供了资金支持。本篇详尽的分析回顾打破了史蒂夫·鲍尔默（Steve Ballmer）时期微软所谓“由盛转衰”的传统叙事，阐述了尽管在消费领域遭遇多起引人注目的挫折，该公司仍如何转型为具有主导地位的企业级科技巨头。内容详细梳理了关键运营脉络，包括美国司法部的反垄断审讯、Windows Vista 和 Windows 8 系统级架构的延期，以及未能抗衡 Google“比免费更优惠”的 Android 移动模式的战略失误。该分析最终展示了鲍尔默如何保护并资助了以 Project Red Dog 为代表的早期云基础设施，从而打造出企业业务引擎和数据中心布局，为萨提亚·纳德拉（Satya Nadella）领导下 Azure 最终斩获数万亿美元规模的胜利奠定了坚实基石。关于史蒂夫·鲍尔默执掌时期微软“由盛转衰”的传统说法并不准确；通过深耕企业软件销售，公司年营收从1995年的60亿美元成功跃升至2014年的800亿美元。 (0:32) 比尔·盖茨在意识到互联网的爆发式增长后，系统性地将微软整体战略重心转向互联网，放弃了此前关于“信息高速公路”和互动电视的业务构想。 (11:03) 微软采取了激进的捆绑策略，在 Windows 95 中免费提供 Internet Explorer，以此切断 Netscape 的收入来源，并在2000年确立了对浏览器市场的完全主导地位。 (40:14) 具有里程碑意义的1998年“美国诉微软公司案”反垄断诉讼揭露了其激进的商业行为，并最终在2000年导致了拆分令，重创了微软的估值和内部文化。 (52:57) 布拉德·史密斯（Brad Smith）主导了化解遗留监管纠纷的战略转向，使微软从监管机构眼中的弃儿转变为全球各国政府值得信赖的软件供应商。 (90:03)

Reading summary

Ben Gilbert and David Rosenthal argue that while popular memory depicts Steve Ballmer’s tenure at Microsoft as an era of consumer blunders and stagnation, it actually entrenched an immensely lucrative enterprise software monopoly and quietly laid the technical groundwork for Azure. Antitrust Fallout, Bureaucratic Drag, and Cultural Decay (84:52) The protracted antitrust battle with the Department of Justice imposed severe operational and psychological costs on Microsoft. The hosts contend that the legal onslaught neutralized Bill Gates—the company's greatest strategic advantage—and saddled software development with protocol documentation requirements that permanently slowed iteration. Following Gates's resignation as CEO in 2000, Steve Ballmer promoted Brad Smith to General Counsel with a mandate to make peace. Smith systematically settled outstanding regulatory disputes, transitioning Microsoft from a perceived corporate predator into an institutional enterprise partner. Simultaneously, a flatlining stock price fostered an internal zero-sum culture. Because employees believed corporate valuation would not grow, advancement required outmaneuvering internal rivals. This dynamic was exacerbated by rigid stack-ranking performance reviews that forced managers into head-to-head employee bake-offs, visually symbolized by the popular cartoon of Microsoft product divisions pointing handguns at one another. The Longhorn Architectural Quagmire and Talent Drain (140:37) Microsoft’s ambitious vision for "Longhorn" (eventually released as Windows Vista) turned into an engineering sinkhole centered around WinFS, an object-oriented file system that independent developers never adopted. With Gates serving strictly as Chief Software Architect, engineering lacked a pragmatic commercial "governor" to rein in technical overreach and halt attempts to weave .NET directly into the operating system core. Salvaging the operating system required executive Brian Valentine to aggressively strip features, yet the crisis drained elite talent across the company. Valentine and key systems engineers departed for Amazon immediately after Vista shipped. In discussions with the hosts, Ballmer cited Vista as his single worst mistake as CEO, observing that while bad acquisitions only cost money, Vista took the company's best technical minds off the field during critical shifts in search and mobile. Enterprise Optimization and Repeated Consumer Blind Spots (115:36) Microsoft transformed its financial profile by moving away from retail boxed software—which carried gross margins of just 29%—toward Enterprise Agreements and OEM pre-installs boasting 75% gross margins. Providing centralized IT administrators with administrative control over worker behavior delivered massive enterprise lock-in, but it detached Microsoft from end-user product taste and consumer developer ecosystems. Consequently, the company faltered across consumer shifts despite early strategic foresight. Windows 8 alienated traditional PC users and OEMs by prematurely forcing a touch-centric tablet interface onto desktops before hardware was ready. In mobile, Windows Phone was undercut by Google's "less than free" Android business model, which subsidized carriers and manufacturers. Furthermore, Microsoft initially treated search as a minor portal feature rather than recognizing that ad-monetized search generated three to five times more revenue per user than software licensing. The Incubation of Azure Under Ballmer (215:10) Contrary to the narrative that cloud computing was conceived under Satya Nadella, the technical and strategic foundation of Azure spent seven years incubating during the Ballmer era. The initiative began under Chief Software Architect Ray Ozzie in 2005 as "Project Red Dog," relying on infrastructure, real-time networking, and distributed systems harvested from MSN, Hotmail, Xbox Live, and Bing. Ballmer provided vital executive air cover against internal critics reluctant to jeopardize high-margin Windows server licensing with capital-intensive data centers. By 2010, Ballmer declared Microsoft "all in" on cloud computing and replaced server division head Bob Muglia with Satya Nadella to build a cloud-first organization. The hosts emphasize that Microsoft's greatest failure in this era was narrative rather than strategic, as inheriting an inflated dot-com valuation multiple masked substantial operational execution. The hosts conclude that dismissing the Ballmer era as a lost decade overlooks how enterprise profits and cloud incubation enabled Microsoft's subsequent multi-trillion-dollar valuation, though whether the company could have preserved its enterprise base while simultaneously winning consumer platforms remains an enduring historical question.

精炼阅读摘要

Acquired 节目主播 Ben Gilbert 与 David Rosenthal 指出，外界将史蒂夫·鲍尔默（Steve Ballmer）执掌微软的时期简单归咎于全面衰退是片面的；尽管微软因反垄断诉讼、消费级产品失策与内耗错失了移动和搜索先机，但其通过企业级业务与长达七年的 Azure 孵化，为后来的复兴奠定了核心基础。反垄断诉讼重创研发效率与领导层精力 (84:52) 微软在 1990 年代凭借“拥抱与扩展”策略迅速统治互联网，但捆绑 Internet Explorer 的行为直接引发了 1998 年美国司法部与微软的反垄断大案。诉讼源于 1990 年联邦贸易委员会（FTC）针对按处理器许可的调查，虽然 FTC 在 1993 年以 2-2 投票僵局未作处罚，但司法部破例接手并在 1994 年达成和解协议。1997 年司法部再次指控其违规捆绑，主审法官 Thomas Penfield Jackson 曾下令要求提供不含 IE 的 Windows，微软照办后系统却彻底崩溃；其后 Jackson 法官因私下向记者泄露未公开裁决而被上诉法院撤换，最终因 2001 年“9·11”事件及政府换届推动了全面和解。节目主播强调，反垄断案给微软造成的最沉重打击在于核心资产比尔·盖茨的极度分心以及企业战略焦点的瓦解。诉讼期间繁重庞杂的底层协议文档化合规要求，给工程团队施加了长期拖累，严重减缓了产品迭代。面对动荡，盖茨于 2000 年辞去首席执行官职务并转任首席软件架构师，由鲍尔默接任；新任总法律顾问 Brad Smith 随后以一张仅写着“该讲和了”的幻灯片为准则，转变强硬态度全面达成监管和解，促使微软从监管众矢之的转型为各国政府值得信赖的服务商。 B端企业合同锁死基本盘却滋生零和内耗 (115:36) 在硬件渠道端，微软果断放弃毛利率仅 29% 的零售盒装软件模式，全面转向利润率高达 75% 的 OEM 预装授权与面向大客户的企业协议（Enterprise Agreement, EA）。这种企业销售模式精准迎合了企业 IT 管理员的需求，IT 部门将随意下载软件、运行自定义宏的普通员工视作潜在的安全威胁，极度渴求系统层面的行政限制与合规控制，而微软的 EA 框架恰好提供了这种集中管控能力，带来了极其稳固且高利润的企业现金流。然而，企业级市场的绝对主导导致微软轻视了普通用户的软件体验。更为严重的是，在股价常年横盘的漫长停滞期，员工对公司总价值增长丧失信心，将内部竞争视为晋升的唯一通道。严苛的末位淘汰制（Stack Ranking）强制要求经理对员工进行 1 到 5 分的定额打分，导致团队间彼此拆台甚至互相倾轧，形成了经典的“各部门拿枪互指”的组织内耗，研发人才在内部零和博弈中被严重空耗。 Vista陷阱抽干人才与搜索商业模式错判 (140:37) Windows Vista（原代号 Longhorn）的研发成为微软史上惨痛的工程黑洞。由于盖茨转任首席软件架构师后缺乏接地气的商业制衡，开发团队执着于构建过于理想化的 WinFS 面向对象文件系统，并试图将 .NET 深度嵌入系统底层，最终导致项目失控延期。微软不得不从当时称霸企业市场的 Exchange 团队抽调 Brian Valentine 等核心系统骨干进场止损并砍掉多项特性；在系统艰难交付后，这批精英技术力量随即大量出走亚马逊，成为支撑亚马逊电商平台的核心高管。鲍尔默日后坦承，Vista 周期是其 CEO 任内最糟糕的时刻，其代价并非资金损失，而是全公司最顶尖的工程技术心智被彻底冻结与抽干。在搜索领域，微软将搜索仅视为 MSN 或 Windows Live 的附加功能，未能洞察搜索广告竞价体系所蕴含的惊人规模效益与飞轮效应。相较于传统售卖软件许可证，广告变现直接面向规模极其庞大的线下商业消费预算，单用户变现价值是传统软件的 3 至 5 倍。2008 年微软曾开出 470 亿美元天价洽购雅虎以求翻盘，甚至在波音机场进行紧急谈判，但最终破裂，彻底落后于依靠搜索广告盈利反向补贴生产力办公套件的谷歌。战略高瞻远瞩却因落地品味与时机屡屡碰壁从硬件与消费交互来看，微软在前瞻战略上往往具有惊人的敏锐度，盖茨极早便预言了自然用户界面、平板电脑与移动设备的普及。然而，微软在具体战术品味与落地上却屡屡产生严重偏差：往往比底层基础设施与芯片成熟期提早了 5 到 20 年立项，并押注在错误的设计范式上。例如在触控技术上长期固守依赖手写笔的电阻屏，忽视了电容多点触控；将平板电脑简单粗暴地视作精简版的桌面 PC，而苹果 iPad 则证明平板电脑本质是放大的智能手机。 2010 年 iPad 带来的颠覆性威胁促使微软仓促推出主打触控的 Windows 8，试图以源自 Zune 的 Metro UI 一箭双雕，但该系统在低功耗 ARM 硬件环境与桌面 PC 之间未能协调平衡，激怒了未准备好触屏硬件的 OEM 厂商并遭到核心极客群体的猛烈抵制。移动生态上，面对谷歌以极具侵略性的“低于免费”模式（直接补贴运营商与销售渠道）大肆推广 Android，Windows Phone 阵营仅剩诺基亚支持。2013 年诺基亚以倒向 Android 相要挟，鲍尔默在公司内部高管普遍反对的情况下力主推进 70 亿美元的收购案，这一激进化豪赌最终成为其黯然退休与领导层彻底交棒的导火索。长达七年的云端静默孵化打破失落神话 (215:10) 主流舆论普遍认为鲍尔默时代只知死守 Windows，云端成功完全归功于其继任者，但主播明确反驳了这一观点。Azure（原代号 Red Dog）的实质性孵化早在 2004 至 2005 年盖茨筹划退休时便已启动。鲍尔默深知自身技术视野有限，收购 Groove Networks 请来 Lotus Notes 创造者 Ray Ozzie 出任首席软件架构师。Ozzie 在 2005 年 10 月发布预见性的《互联网服务颠覆》备忘录，正式拉开云转型序幕。在企业级客户对运行开源代码可能损害 Windows 暴利心存顾虑之际，鲍尔默顶住内部压力为云计算孵化提供了关键保护伞，整合了 Bing 的分布式系统、Xbox Live 的 4000 万实时联网架构及 Hotmail 的大规模 Web 经验。2010 年鲍尔默公开宣告微软“全面押注云计算与 Azure”，并大胆撤换当时掌管年增长 30% 到 40%、体量 120 亿美元服务器业务的高管 Bob Muglia，提拔 Bing 业务出身的萨提亚·纳德拉接管云端。纳德拉在接任 CEO 之前，Azure 已经在鲍尔默麾下完整孕育了七年，微软的“失落岁月”本质上是一场未能向外界有效传达其底层战略推进的故事讲述失败。微软在鲍尔默时代的真正悲剧并非缺乏技术预见或战略商业工程能力，而是高昂的监管与研发试错代价分散了组织心智，加之叙事能力的欠缺掩盖了其深层的工程蓄力。随着 2014 年初鲍尔默与盖茨双双彻底退出管理层与董事会，积压四十年的历史包袱得以出清，为纳德拉激活云计算与企业服务底盘铺平了道路。

Detailed notes — zh

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=0s) 微软第二卷：颠覆认知的二十年与企业软件帝国节目一开场，Ben Gilbert 调侃自己今天嗓子有点哑（"I'm a little hoarse today"），打趣希望接下来不用说太多话，David Rosenthal 则笑着祝他好运。这是《Acquired》第 14 季的收官集（第 6 集）——备受期待的“微软第二卷”。在第一卷梳理了微软早期的创业史之后，第二卷终于来到了听众们亲身经历并伴随成长的黄金与动荡岁月：互联网浪潮、Windows XP、Xbox、浏览器之战、搜索引擎以及移动时代的全面溃败。在这一时期，微软常常拥有极具前瞻性的绝妙构想，却在执行时机和落地手法上屡屡摔跟头，从 Zune 播放器到 Bing 搜索莫不如是。然而，表面的产品挫败并不能掩盖商业上的惊人扩张。从 1995 年到 2014 年这二十年间，微软的年营业收入从 60 亿美元一路狂飙至 800 亿美元。它不仅没有倒下，反而彻底参透了企业级软件销售的商业密码。Ben 坦言，刚开始做研究时，他以为第一卷讲述的是“崛起”，第二卷必然是关于“衰落”、文化毒瘤、消费电子产品连环惨败以及反垄断诉讼；但历时数月的深入挖掘彻底颠覆了这种二元论构想。谈及 1998 年美国司法部针对微软发起的世纪反垄断诉讼，人人都知道这桩案子，但极少有人清楚其真实的审判走向与结局——微软真的输掉了这场官司吗？答案并非简单的胜负，而是充满了戏剧性的制度博弈与精妙细节。 - 叙事时间跨度：从 1995 年（Windows 95 发布）到 2014 年（萨提亚·纳德拉接任 CEO）。 - 财务增长奇迹：微软年收入在 1995 至 2014 年间从 60 亿美元飙升至 800 亿美元。 - 核心叙事反思：推翻“早期辉煌、中期滑铁卢、晚期由萨提亚力挽狂澜”的粗糙滤镜，重新审视企业级业务带来的深厚护城河。 --- ### [02:04](https://www.youtube.com/watch?v=6oER1bdQa_s&t=124s) 重磅预告：Acquired 旧金山 Chase Center 现场秀与特别嘉宾马克·扎克伯格在正题展开之前，两位主播公布了《Acquired》开播以来的重磅里程碑活动：他们将于 9 月 10 日在旧金山金州勇士队的主场 Chase Center 举办大型现场秀。这场活动由摩根大通支付部门（J.P. Morgan Payments）联合呈现，现场不仅为全球奔赴而来的听众准备了多重社交与互动环节，更邀请到了当之无愧的重量级嘉宾——Meta 创始人兼 CEO 马克·扎克伯格（Mark Zuckerberg）。扎克伯格不仅主导了节目此前反复拆解过的科技史上最成功的几笔并购案（如 Instagram、WhatsApp），如今 Meta 也正凭借开源大模型与先进 AI 战略深刻定义着未来十年的计算形态。主播提醒大家可以前往 acquired.fm/sf 登记购票提醒，并惯例给出了免责声明：本节目内容不构成任何投资建议，主播与听众可能持有节目中所讨论公司的股票或指数基金。 --- ### [03:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=204s) 盛名之下的迷思：萨提亚的自白与“相互指枪”的组织架构关于微软在 1995 年至 2014 年间的这二十年，科技界早已形成了一套根深蒂固的标准叙事：比尔·盖茨退居二线，史蒂夫·鲍尔默执掌下的微软错失了每一个浪潮，陷入大公司病，直到 2014 年萨提亚·纳德拉（Satya Nadella）挂帅才重获新生。就连萨提亚自己在 2017 年出版的自传《刷新》（*Hit Refresh*）开篇第一段都亲自坐实了这种观点。他在书中深情写道，自己于 1992 年加入微软，是因为渴望与一群坚信自己在改变世界的人共事；但在连续多年将所有竞争对手甩在身后之后，公司内部悄然变质，“创新被官僚主义取代，团队协作被内部政治取代，我们开始掉队了”。 ```mermaid flowchart TD accTitle: 微软大众叙事与实际情况的对比 accDescr: 对比科技界流传的简单二元论叙事与播客调研展现的复杂真实商业状况 subgraph PopularNarrative ["流行叙事模式"] P1["1995年前：狂飙胜利"] --> P2["1995-2014：内斗与衰落 （鲍尔默时期）"] P2 --> P3["2014年至今：萨提亚刷新 重返巅峰"] end subgraph ActualReality ["实际调研真相"] R1["巨额财务增长 （收入增长逾13倍）"] R2["企业级市场大获全胜 （奠定云服务与B端底座）"] R3["极其复杂的制度抗衡 （反垄断博弈与战略惯性）"] end ``` 这种内耗甚至催生了漫画家兼软件工程师 Manu Cornet 那幅传遍互联网的名作——微软组织架构图被画成了各大业务部门之间互相拔枪对峙的黑色幽默。市面上的主流看法可以粗暴概括为：微软先是赢麻了，接着烂了很久，然后全靠萨提亚挽狂澜于既倒。但两位主播在筹划本集时反问：这真的符合客观事实吗？为了拆解这段历史，他们采访了二三十位微软亲历者和高管，访谈人数是过往任何一期单集节目的四到五倍，最终挖出的复杂现实远比这种脸谱化故事精彩得多。 - 萨提亚·纳德拉在《刷新》中的名言揭示了当年微软从技术使命感滑向官僚内耗的阵痛。 - 著名的“枪口互指组织架构图”（Gun pointing org chart）形象刻画了 Windows、Office 与其他新兴业务部门之间争夺资源与话语权的血腥内斗。 - 本集调研广度空前，采访名单长达二十余人，力求还原鲍尔默时代真实的商业逻辑与执行得失。 --- ### [05:08](https://www.youtube.com/watch?v=6oER1bdQa_s&t=308s) Windows 95 的狂欢背后：那个唯一缺席的关键角色故事的起点必须回到第一卷的终点——1995 年 8 月 Windows 95 的世纪发布会。那是一场对个人电脑软件的纯粹狂欢，微软请来了名嘴 Jay Leno 担任主持，斥巨资买下滚石乐队的《Start Me Up》作为广告主题曲，隆重推出了革命性的“开始”按钮。整个科技产业都沉浸在这一狂热之中，但站在消费者的角度审视，Windows 95 初始版本其实戏剧性地遗漏了一件未来最重要的核心工具：一个网页浏览器（Internet Browser）。在回溯历史时，主播们坦言之前刻意没在 Windows 95 发布节点深聊互联网，因为一旦涉足网络，整部微软历史就必须彻底翻入下一篇章。多年后再看，Windows 95 真正奠定历史地位的原因，不仅是其图形界面的成熟，更是因为它是全人类第一次大规模接入互联网的物理跳板。当时的业界流行一个词叫“互联网时间”（Internet Time），意味着所有商业变革都是以“周”而非“年”来计算。然而，在 1992 年至 1995 年初的消费者世界中，“上网”的含义与如今完全不同。普通用户拨号连接的是 CompuServe、Prodigy，以及巨无霸 AOL（美国在线）。这些服务根本不是开放的万维网，而是封闭的“围墙花园”（Walled Gardens），它们通过电话拨号调制解调器将专有内容和网络通道打包兜售。想要在上面发布内容，普通人根本不可能自己插上一台服务器就搭建网站，必须拥有极强的商业议价能力，并动用私人关系与 AOL 这样的平台巨头签订准入协议。 - CompuServe 的离奇背景：当时全美最大的拨号在线服务之一 CompuServe，其母公司竟然是总部位于哥伦布的报税连锁机构 H&R Block。 - 互联网基础设施对比：早期在线体验完全受制于拨号运营商的专有网络生态，与任何人皆可搭建主页、自由链接的开放 Web 截然不同。 --- ### [07:50](https://www.youtube.com/watch?v=6oER1bdQa_s&t=470s) Project Marvel、AOL 与“信息高速公路”的电视迷思作为当时科技世界的统治者，微软自然不能坐视在线服务的风口旁落。1993 年，微软管理层开始试探性地接触 AOL，探寻收购后者的可能性。然而 AOL 创始人 Steve Case 断然拒绝了卖身要约。更具戏剧性的是，微软联合创始人 Paul Allen 随后擅自通过个人渠道大举买入 AOL 的大宗股份，这一举动引发了微软高层的内部尴尬与战略混乱。既然无法完成收购，微软决定正面宣战，在内部启动了一个代号为“Project Marvel”的绝密项目，旨在打造一套抗衡 AOL 的专有在线服务——这便是 MSN（The Microsoft Network）的最初雏形。 ```mermaid flowchart LR accTitle: 90年代初两大联网网络构想路径 accDescr: 展示当时被主流追捧的有线电视信息高速公路方案与个人电脑开放互联网方案的对比 subgraph Vision1 ["当时被主流信奉的坦途：有线电视网络"] direction TB C1["基础设施：已入户的有线/卫星电缆"] --> C2["终端呈现：家家户户的电视屏幕"] C2 --> C3["控制与交互：双向机顶盒与遥控器"] C3 --> C4["内容供给：成熟的好莱坞与有线电视传媒"] end subgraph Vision2 ["当时被视为不可思议的狂想：PC 互联网"] direction TB P1["基础设施：需全新铺设/改造的数据通信网络"] --> P2["终端呈现：昂贵的个人电脑（PC）"] P2 --> P3["交互硬件：键盘与鼠标操作门槛"] P3 --> P4["内容供给：需从零培育的数字原生网站"] end ``` 在这里，Ben 敏锐地指出了历史记载中的一个偷梁换柱之处：“Project Marvel”在立项之初完全是一个类似 AOL 的封闭专有网络；直到该模式彻底破产后，微软才把 MSN 这个品牌重新包装套用在基于开放 Web 的门户媒体属性上。而在 90 年代初期，不论是科技界、微软管理层还是华尔街投资者，相当多的人都深信这种封闭式的拨号在线服务只是过渡方案，真正的未来属于他们口中的“信息高速公路”（Information Super Highway）。令人啼笑皆非的是，当时全行业认定的信息高速公路终极形态并非电脑网络，而是由付费电视提供商（Comcast、Charter、Time Warner Cable、DirecTV 等有线和卫星电视巨头）主导的“交互式电视”（Interactive Television）。站在 1993 年的视角，这套逻辑无可挑剔： 1. **终端普及度**：电视机早已进驻千家万户，而昂贵的 PC 仍是少数人的工具； 2. **交互基础设施**：有线电缆早已铺设完毕，只需升级双向机顶盒即可实现信号回传与遥控交互； 3. **内容生态**：成熟的媒体制作公司和娱乐内容全部围绕电视运转。当时如果有人向投资人推销“未来每个人都会专门花几千美元购买一台名为 PC 的新型设备，全社会将铺设全新的专用网络线路，并由一批全新的网络原生公司来创造内容”，这在任何人听来都是异想天开的疯狂赌博。因此，微软在嗅到联网风向的第一反应，正是调转船头去追逐有线电视与机顶盒的交互式网络，历史的发展在此刻埋下了巨大的误判伏笔。 - 关键战略代号：Project Marvel，微软对抗 AOL 的封闭网络项目，失败后被重组为门户时代的 MSN。 - 时代主流预判：认为“信息高速公路”的载体是有线电视网络而非 PC，科技巨头纷纷押注交互式电视与机顶盒操作系统。 - 基础设施权衡：电视在屏幕、布线、遥控反馈和影视内容上拥有压倒性的现成存量优势，导致当时极少有人坚信开放 PC 互联网能凭空逆袭。 ### [10:45](https://www.youtube.com/watch?v=6oER1bdQa_s&t=645s) “CableSoft”与信息高速公路的三角同盟 1993年夏天，科技界到处流传着微软正酝酿与各大有线电视公司成立大型合资企业的传言。这个被称为“CableSoft”的构想野心极大：有线电视巨头掌控物理网络管道、庞大的客户关系以及绝大多数内容资源；微软则负责开发系统软件，既包括装在千家万户客厅里的机顶盒系统，也涵盖后端服务器软件，该软件项目代号为“Tiger”。为了构建这条通往“信息高速公路”（Information Super Highway）的宏伟蓝图，他们拉拢了第三个关键盟友——硬件巨头 Silicon Graphics（SGI）。由于机顶盒与服务器都需要极强的图形与运算处理能力，刚凭借震撼业界的图形工作站技术支持了《侏罗纪公园》（Jurassic Park）CGI特效制作的 SGI 成了不二之选，其创始人兼董事长正是硅谷传奇人物吉姆·克拉克（Jim Clark）。华尔街对这场联姻陷入了狂热，分析师与媒体纷纷惊呼这是一个万亿美元级的超级市场。比尔·盖茨开始频繁与 TCI 的约翰·马龙（John Malone）、时代华纳（Time Warner）的杰拉德·莱文（Gerald Levin）会面，甚至被狗仔队拍到秘密同行的照片。与此同时，盖茨还与好莱坞超级经纪人迈克尔·奥维茨（Michael Ovitz）探讨微软如何分食内容产业的蛋糕，无论是通过筹备中的 MSN 项目，还是其他多媒体探索——这也为日后微软与 NBC 联合组建 MSNBC 有线新闻网埋下了伏笔。 - **三方合作分工**：有线电视运营商掌控网络管道、入户关系与内容分发；微软通过代号“Tiger”的项目提供机顶盒及后端服务器软件；SGI 提供高性能计算与图形硬件。 - **狂热的市场预期**：整个行业将交互式电视与信息高速公路视作下一次计算革命，引发资本市场与好莱坞媒体的全面追捧。 ```mermaid flowchart LR accTitle: CableSoft Alliance accDescr: Tripartite alliance between Cable Companies, Microsoft, and SGI for the Information Super Highway. subgraph Alliance["CableSoft 联盟"] direction TB C["有线电视巨头 (TCI / Time Warner) 管道、客户与内容"] M["微软 (Tiger) 机顶盒与服务器软件"] S["SGI (Jim Clark) 高性能计算与图形硬件"] end C --- M M --- S S --- C ``` --- ### [12:30](https://www.youtube.com/watch?v=6oER1bdQa_s&t=750s) 微软的多方押注与《未来之路》的尴尬改版当时微软面对正在萌芽的交互式网络计算，手中同时押注了两个庞大计划：一个是代号为 Marvel 的 MSN 在线服务，另一个则是信息高速公路机顶盒项目。耐人寻味的是，这两项探索起初都与真正的开放互联网或网页浏览器无关。这精准折射出了微软长久以来的核心商业策略：在未来的终极形态尚未完全清晰前，先在所有看起来最具潜力的方向上广泛下注、四处占位，确保自己始终留在牌桌中央。这种策略在微软早期的 PC 时代屡试不爽，而在此刻也被全面套用。为了向公众布道“信息高速公路”和“信息触手可及”（Information at your fingertips）的愿景，比尔·盖茨决定撰写一本名为《未来之路》（The Road Ahead）的专著。然而传统图书出版周期的漫长，让这本书陷入了极其尴尬的时间差：该书精装版迟至1995年11月才正式面世，而彼时网景（Netscape）早已完成轰动全球的 IPO，Windows 95 也已正式发售。最富戏剧性的一幕发生在实体书的索引对比中。在1995年出版的精装版中，盖茨对互联网的讨论仅仅散落在全书的三处地方；而到了1996年紧急修订推出的平装版里，微软几乎把全文里的“信息高速公路”全选替换成了“互联网”与“网页浏览器”。在平装版的书末索引中，“Internet”一词竟然单独占据了整整一整页，甚至需要为其设立密密麻麻的二级细分子索引。这一页纸的悬殊反差，生动记录了科技巨头在短短一年间从迷茫到骤然惊醒的历史转折。 --- ### [14:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=871s) J Allard 的备忘录：“拥抱并扩展”战略的诞生精装版《未来之路》所反映的认知局限，正是1994年1月微软高层的真实写照。就在那个月，微软年轻的 Windows 网络工程师詹姆斯·阿拉德（James Allard，人称 J Allard）向比尔·盖茨及公司核心高管递交了一份石破天惊的备忘录，标题直截了当地命名为《Windows：互联网的下一个杀手级应用》（Windows: The Next Killer Application for the Internet）。阿拉德在备忘录中警示管理层，伊利诺伊大学国家超级计算应用中心（NCSA）诞生了一款名为 Mosaic 的软件，正以野火燎原般的速度在全网蔓延，而它据传只是一个名叫马克·安德森（Marc Andreessen）的年轻程序员利用业余时间写出来的。阿拉德敏锐地指出，互联网与 Mosaic 网页浏览器正在呈现指数级爆发的态势，这为微软带来了不可思议的战略机遇。他在备忘录中首次提出了后来深深烙印在微软基因里的口号——对互联网采取“拥抱、扩展、创新”（Embrace, extend, innovate）策略，将网络协议与浏览器能力彻底融入 Windows 原生系统。这一策略后来被外界及竞争对手戏谑并攻击为“拥抱、扩展再消灭”（Embrace, extend, extinguish），但在商业逻辑上却极其高明。当时 Mosaic 已经完成了早期的产品市场匹配（PMF）探索，证明了大众对通过超文本访问网络内容的强烈渴求；微软无需承担试错风险，只需直接“拥抱”这一崭新行为，再将其“扩展”进自身垄断全球的 Windows OEM 商业模式之中，让每一个购买新电脑的用户自然获得这项能力。在网景公司甚至还没注册成立的1994年初，年轻的 J Allard 就已经为微软推演出了日后整场浏览器大战的终局打法。 --- ### [16:53](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1013s) 康奈尔大学的暴风雪与“指数级现象不可忽视” 紧接着在1994年2月，又一场偶然事件进一步推倒了微软转向的多米诺骨牌。比尔·盖茨的时任技术助理兼“影子”（Technical Assistant / Shadow，微软内部一项极具声望的精英岗位）史蒂文·辛诺夫斯基（Steven Sinofsky）前往母校康奈尔大学（Cornell University）执行校园招聘任务。不巧的是，一场突如其来的暴风雪将整个校园彻底封锁，辛诺夫斯基被迫在伊萨卡多滞留了数日。在这段与世隔绝的时光里，他惊奇地发现，被大雪困在宿舍里的大学生们全都在狂热地使用互联网。在辛诺夫斯基过去的学术认知里，互联网不过是科学家和学者用来交流论文、传输科研数据的古板网络，绝无商业或大众娱乐价值；然而眼前康奈尔的年轻人们，却在用互联网谈情说爱、在线选课、互相发送无关学业的即时信息与闲聊邮件。深受震撼的辛诺夫斯基随即写下了另一份名载史册的内部备忘录：《康奈尔触网了！》（Cornell is wired!）。微软的历史常常由一系列影响深远的内部备忘录串联而成。读完这份报告后，盖茨公开坦言：“当我听完史蒂文讲述康奈尔校园里发生的一切时，我才开始真正严肃地看待互联网。” 1994年4月5日，盖茨与辛诺夫斯基召集了 J Allard 以及所有负责网络前沿探索的高管团队，举行了一场划时代的“互联网战略研讨闭门会”（Internet Offsite）。在这场会议上，盖茨彻底“皈依”了互联网信仰，并确立了一条影响深远的准则：互联网正处于不可逆的指数级暴发期，而“指数级现象绝不可忽视”正是微软自创立以来最核心的铁律。整座微软帝国最初就是建立在摩尔定律这根指数级杠杆之上的——坚信芯片性能会呈指数级翻倍，从而敢于提前数年开发常人难以想象的软件产品。 --- ### [19:58](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1198s) Jim Clark 的辞职邮件与 Netscape 的横空出世最具戏剧性的是历史的时间线交错：微软召开那场至关重要的互联网闭门会议的前一天——1994年4月4日，网景公司刚刚完成商业注册。视线转回之前埋下伏笔的 SGI 创始人吉姆·克拉克。早在1994年2月，克拉克就因不满董事会与管理层在信息高速公路机遇上的保守与迟缓，愤而辞去了自己一手创办的 SGI 董事长职务。在离开 SGI 的最后一天，克拉克敲出了一封改变互联网历史的陌生拜访邮件（cold email），径直发给了远在伊利诺伊州的年轻小子马克·安德森。克拉克在信中坦言自己注意到了 Mosaic 惊人的普及势头，并直言如果有机会携手合作，自己将“非常感兴趣”。初出茅庐的安德森自然无法拒绝这位硅谷殿堂级传奇人物抛来的橄榄枝。两人迅速达成共识，并于1994年4月4日正式创立了公司，最初命名为 Electric Media（后更名为 Mosaic Communications，最终定名为 Netscape）。克拉克与安德森最初的设想，依然是打算利用安德森卓越的编程才能去承接有线电视网络与信息高速公路的浪潮，做 SGI 没能全力以赴去做的事。他们未曾预料到，正是这家初创公司，即将把整座科技行业乃至比尔·盖茨的微软帝国，全面拖入一场空前残酷的互联网浏览器大战。 ### [21:22](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1282s) 信息高速公路的执念与任天堂 N64 机顶盒计划当吉姆·克拉克（Jim Clark）与马克·安德森（Marc Andreessen）初次联手创办公司时，他们的最初设想甚至不是深耕互联网，而是要做“信息高速公路（Information Superhighway）”的机顶盒。回过头来看，这种偏离让人觉得不可思议——安德森本是全世界最理解互联网呈指数级爆发潜力的人，彼时他甚至已经把 `<img>` 标签引入了 HTML，实现了在浏览器中直接渲染图片。然而，当克拉克发邮件给他时，两人却一头扎进了所谓的“信息高速公路”泡沫中。当时克拉克刚离开 SGI（Silicon Graphics），SGI 曾为好莱坞制作《侏罗纪公园》等大片的图形工作站，并且还是任天堂 N64 游戏机图形引擎的技术合作伙伴。克拉克凭借这层人脉，设想把即将发售、即将进入千家万户客厅电视的 N64 改造为信息高速公路的机顶盒设备。但讽刺的是，历史上被热炒的“信息高速公路”根本从未真正落地：有线电视公司只在极少数社区做过连接二三百户人家的封闭实验，最后沦为充斥着过多参与方的“PPT 概念（Slide wear）”。甚至连安德森自己当时都没坚定敲桌子认定互联网才是未来，人类大脑天然缺乏对复利增长的理解直觉，安德森当时的心态更像是：“我已经做了一个玩具般的学术项目成了名，现在是时候听从专家们的建议去做点正经的‘大人业务’了。” ```mermaid flowchart LR accTitle: SGI与任天堂合作转向机顶盒构想 accDescr: 展示克拉克与安德森最初计划利用任天堂N64打造机顶盒，随后放弃转向浏览器的过程。 subgraph SGI["克拉克的资源"] A["SGI 图形技术"] --> B["任天堂 N64 图形引擎"] end subgraph Plan["早期创业构想"] B --> C["N64 改装为信息高速公路机顶盒"] end subgraph Pivot["战略转向"] C -->|"信息高速公路未落地"| D["彻底放弃机顶盒"] D --> E["成立 Mosaic 商业化公司"] end ``` ### [23:52](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1432s) 商业化 Mosaic 的诞生与改名 Netscape 到了 1994 年深春，伊利诺伊大学的 NCSA Mosaic 活跃用户数已达百万，比尔·盖茨（Bill Gates）也开始高度关注这一趋势。克拉克与安德森终于醒悟，果断放弃了 N64 机顶盒计划，转而专注 Mosaic 的商业化。他们拉来了传奇风投机构凯鹏华盈（Kleiner Perkins）的顶级投资人约翰·杜尔（John Doerr）领投并加入董事会。1994 年 10 月，新成立的“马赛克通信公司（Mosaic Communications Corporation）”在网上推出了第一款免费下载的浏览器——Mosaic Navigator。其商业模式非常清晰：客户端浏览器免费向大众赠送以抢占市场，盈利则靠向需要搭建网站的企业客户销售高价的 Web 服务器软件。值得注意的是，安德森虽然沿用了“Mosaic”这个名字借势宣传，但所有代码都是团队从零全新编写的商业版本，并没有直接抄袭大学时期的代码。与此同时，伊利诺伊大学意识到自己手中原本的 Mosaic 代码极具商业价值，便把原始 Mosaic 的商业授权独家卖给了一家名为 Spyglass 的本地公司。这就造成了巨大的冲突：每当安德森的新 Mosaic 团队去推销服务器软件时，Spyglass 就会跑去向客户发出法律威胁声称要起诉侵权，直接搅黄交易。面对生死存亡的法律与市场阻击，安德森的团队不得不反诉 Spyglass 叫停骚扰，最终双方达成和解。为了彻底摆脱法律纠纷，这家初创公司在 1994 年秋天正式更名为网景（Netscape）。安德森当时在媒体面前留下了一句极为经典的凡尔赛名言：“你去上大学，做了研究，毕业离开了，然后他们反过来想毁掉你的生意。早知道会这样，我当初就去斯坦福了。”（尽管这只是安德森的事后调侃，因为来自中西部的他当时并没有去斯坦福的经济条件和名气）。 - **商业模式设定**：个人端浏览器 Mosaic Navigator 完全免费下载，通过向搭建网站的企业端售卖 Web 服务器软件赚钱。 - **命名纠纷核心**：安德森团队拥有的是“从零编写”的新商业代码，而伊利诺伊大学通过 Spyglass 持有“原始 Mosaic”的商业授权。 - **改名结局**：诉讼和解后，Mosaic Communications 正式更名为 Netscape。 ### [27:28](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1648s) 浏览器争夺战与微软收购落空在互联网呈爆发态势的同时，各大在线服务巨头（如 CompuServe、Prodigy、AOL 等）敏锐地察觉到了时代变局。1993 年末至 1994 年初是一个划时代的节点：AOL 等封闭网络服务开始与开放互联网打通，用户不仅可以在自家的专属网络内浏览，还可以冲浪访问开放的万维网。这些平台急需将浏览器整合进自己的客户端。 Netscape 志在自立生态、出售服务器软件，无意将浏览器技术授权给别人。于是，Spyglass 开始对外大肆授权原始 Mosaic 的代码；与此同时，一家名为 Booklink 的初创小公司也独立写出了另一款浏览器并寻求授权合作。1994 年 5 月，刚刚开完内部“互联网务虚会（Internet Retreat）”的比尔·盖茨和史蒂芬·辛诺夫斯基（Steven Sinofsky）下定决心要将互联网作为微软和 Windows 的核心战略，便前去与 Booklink 洽谈，探讨以约 200 万美元的价格拿下其浏览器的授权。然而中途杀出程咬金，AOL 直接豪掷 3000 万美元把 Booklink 整家公司买断。当时市面上成熟的独立浏览器仅有三家： 1. **Booklink**：被 AOL 闪电全资收购； 2. **Netscape Navigator**：坚持独立发展，拒绝技术对外授权； 3. **Mosaic（Spyglass）**：唯一剩下的授权渠道。错失 Booklink 的微软别无选择，只得转向 Spyglass，最终以 200 万美元拿下了原始 Mosaic 源代码的授权许可，这也构成了日后 Internet Explorer（IE）的基础框架代码。早期的 IE 浏览器在“关于”界面中甚至直接标注着“基于 NCSA Mosaic，由 Spyglass 授权分发”。 ### [29:40](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1780s) Windows 97 的宏大构想与 IE 真正的技术溯源大众通常熟知的历史叙事是“微软从 Spyglass 买来 Mosaic 代码包装成了 IE”，但这只是流于表面的版本。在微软内部，战略从来不是盖茨和辛诺夫斯基几个人坐在一起想明白后下一道圣旨、其他人排排坐执行那么简单；真实的历史充满了各方势力的多线并行与混乱竞争。早在 1994 年底 Windows 95 发布之前，由托马斯·里尔登（Thomas Reardon）领导的一支 Windows 95 核心团队成员就被抽调出来，着手规划下一代代号为“Windows 97”的战略级功能。里尔登团队的核心思想非常激进——“全天候、全方位拥抱互联网（All Internet, all the time）”。该小组不仅构想了类似于现代 Zoom 的虚拟会议软件、专为公共互联网而非企业局域网设计的专属邮件客户端，还提出了一个颠覆性的深层愿景：**将整个 Windows 外壳（Shell）彻底变成一个浏览器**。在他们的设想中，操作系统层面的每个交互视觉组件，都应该是一个能解析 HTML、能与线上服务器通信的在线实体。从底层工程逻辑来看，HTTP 协议不过是跑在 TCP/IP 之上的一个普通协议，因此操作系统理应直接内置 HTTP 通信支持，并向所有第三方应用提供一个标准、可复用的底层 HTML 渲染器组件。这种将浏览器技术直接下沉为操作系统基础构件的设计思路，不仅深刻塑造了 IE 的底层架构，更为数年后震动硅谷的微软反垄断世纪诉讼埋下了关键伏笔。 ### [32:22](https://www.youtube.com/watch?v=6oER1bdQa_s&t=1942s) 浏览器的组件化战略与来自 Spyglass 的代码波折微软最初在浏览器上的策略并不是简单地写一个孤立的应用程序，而是构建底层构建模块（Building Blocks），包括深度的 HTTP 协议栈、HTML 渲染引擎以及围绕它们的外层包装，使任何第三方也能调用这些能力来渲染网页。微软甚至曾主动找上门对网景（Netscape）说：“看，我们有这么优秀的 HTTP 栈和 HTML 引擎组件，你们何必从头重写？直接用我们打算内置进 Windows 的现成代码吧！”网景断然拒绝了这一提议。最终，Internet Explorer（IE）反而成了唯一深度使用这套底层 Windows 组件的应用程序。与此同时，IE 团队在推进过程中说服了 Windows 领导层：我们动作可以极快，不必苦等下一次 Windows 大版本迭代，完全可以赶上 Windows 95 的发售周期。当时围绕 IE 的技术来源流传着一个版本——微软拿到了 NCSA Mosaic 的代码。但实际情况极富戏剧性：微软无法直接授权最初由马克·安德森（Marc Andreessen）等人编写的原版 NCSA Mosaic，只能转而向拥有商业授权的 Spyglass 公司寻求授权。然而 Spyglass 当时为了打造自己的独立浏览器，已经对 Mosaic 代码进行了大刀阔斧的魔改，改得甚至相当糟糕。微软接手后，不得不花费大量精力把 Spyglass 改动的部分一点点“倒带”剥离回去，随后再在上面搭建自己的 Explorer 体系。 - 整个过程与微软当年从西雅图计算机产品公司（Seattle Computer Products）收购 QDOS（Quick and Dirty Operating System）演化为 MS-DOS 的故事如出一辙：买来的只是一张门票，后续海量的工程重构全由微软自己完成。 - 尽管如此，如果在早期的 IE 浏览器中点击“关于”（About）窗口，版权信息依然标明致谢 Mosaic 与 Spyglass，这种历史的讽刺意味堪称“美味”（delicious）。 - 这段插曲凸显了微软的核心图谋：他们不是为了做一款浏览器小应用，而是要让整个 Windows 操作系统在底层深度“网络化”（Web-enabled）。IE 运行所需的许多关键协议实现，本就是直接写进 Windows 系统内核的代码。 ```mermaid flowchart TD accTitle: IE 的技术整合与 Windows 深度绑定架构 accDescr: 展示微软如何从外部代码中剥离重构并深度嵌入 Windows 系统底层的技术链路 subgraph External["外部技术代码"] A["NCSA Mosaic 原始代码 （无法直接授权）"] -.-> B["Spyglass 深度修改版代码 （微软最终获取）"] end subgraph Microsoft["微软重构与系统级集成"] C["代码剥离与清理 （撤销大量 Spyglass 改动）"] D["Windows 底层协议栈 （HTTP 栈 / HTML 渲染引擎）"] E["Internet Explorer 应用程序"] end B --> C C --> D D --> E D -. 曾提议提供.-> F["Netscape Navigator （网景拒绝采用）"] ``` --- ### [35:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2115s) “互联网海啸”备忘录与网景的狂飙突进在 Windows 95 紧锣密鼓准备发售的 1995 年春季，比尔·盖茨写下了一篇名垂青史、日后更被印证为硅谷转折点的公开备忘录——《互联网海啸》（The Internet Tidal Wave Memo）。盖茨在文中毫不掩饰自己的危机感，明确写道： > “或许大家已经看过我或其他人之前关于互联网重要性的备忘录。我对它的重要性认知经历了数个不断提升的阶段。现在，我将互联网赋予最高等级的重要性。我希望借此备忘录明确一点：我们对互联网的专注，对我们业务的每一个部分都至关重要。互联网是自 1981 年推出 IBM PC 以来最重要的单项进展，它甚至比图形用户界面（GUI）的到来更为重要。” 1995 年 8 月 24 日是 Windows 95 的既定发布日。而在发布两周前的 8 月 9 日，网景公司抢先登陆纳斯达克。这笔轰动全球的 IPO 瞬间让这家成立不久的公司市值飙升至 30 亿美元。当时网景在短短一年内用户规模从 100 万疯狂膨胀到 1500 万，展示出无可挑剔、立竿见影的“产品市场匹配”（PMF）。科技界过去常为某项新技术“到底会不会成气候”争论不休，但自 1994 年起，从未有人怀疑过互联网的未来。在伴随 IPO 的媒体狂欢中，马克·安德森抛出了一句令微软背脊发凉的挑衅名言：“网景很快就会将 Windows 贬低为一套充斥着未调试 Bug 的设备驱动程序集合（a poorly debugged set of device drivers）。” --- ### [37:06](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2226s) 终极特洛伊木马：操作系统之争升级为平台之战安德森那句充满狂气的断言背后，揭示了计算机产业底层权力的致命转移。Windows 的商业帝国之所以坚不可摧，是因为它是所有独立软件开发商（ISV）赖以编写企业与消费级软件的平台；用户为了运行这些软件，不得不购买 Windows。但安德森指出的未来图景是：那些在浏览器中展示的简陋静态网页只是第一步，随着 JavaScript、CSS 以及日后的 Java 和 Flash 落地，浏览器自身将演化为富 Web 应用的开发平台。如果未来所有的软件都是在浏览器内部运行，那么底层运行的究竟是 Windows、Mac 还是其他操作系统，对用户和开发者而言将毫无区别。操作系统将被降级为仅仅用来开机、驱动屏幕、鼠标和键盘的外设管理器。这触动了微软最深层的生存恐惧——一旦 Web 成为未来平台，微软耗费十数年建立的生态网络效应与软件护城河将在瞬间土崩瓦解。更为残酷的黑色幽默在于：网景构想的这个摧毁微软帝国的“未来平台”，本身却必须寄生在微软的 Windows 系统上发行。当时 Windows 坐拥 80% 至 90% 以上的市场份额，网景正是靠着庞大的 Windows 装机量才得以触达千家万户。网景借助微软的渠道反噬微软，这无异于一场特洛伊木马式的突袭。 - 盖茨在《互联网海啸》备忘录中对网景的威胁洞若观火，日后该段文字甚至成为美国司法部（DOJ）反垄断起诉中的核心呈堂证供：“一个在互联网上诞生的新竞争对手是网景。他们的浏览器占据了 70% 的主导使用份额，使他们有权决定哪些网络扩展标准能够流行……他们正推行多平台策略，将关键的 API 移入客户端，以商品化（commoditize）底层操作系统。” - 这场对决在微软看来是绝对生死存亡的“零和博弈”。为了斩断网景的命脉，微软必须把浏览器以 IE 的形式彻底熔接进 Windows，死守应用平台霸主的地位。 --- ### [40:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2414s) 零美元免费与系统捆绑：网景的空气供给被彻底切断 1995 年 8 月 Windows 95 零售盒装版正式发售时，IE 并没有默认内置在系统里。用户需要额外花费 50 美元购买名为“Plus! Pack”的扩展包才能安装 IE。虽然微软能从中赚取零售软件收入，但面对网景浏览器在网络上的免费下载和试用狂潮，收费的 IE 根本无法形成阻击。当时全社会对“信息高速公路”的所有憧憬与投机狂热全盘倒向网景，网景股价在 9 月、10 月、11 月间连续翻了三倍，市值逼近 100 亿美元——尽管当时其企业端服务器软件几乎还没产生像样的利润。微软迅速认清了现实：首要目标绝不是靠浏览器赚零售软件的钱，而是以最快速度铺开规模，“切断网景的氧气供应”（Cut off their air supply），彻底掌控互联网入口。 - **1995 年 12 月 7 日的宣战**：比尔·盖茨正式宣布，Internet Explorer 将对所有用户永久免费，并且在未来将直接捆绑进每一份出厂和销售的 Windows 95 系统中。 - **市场判决**：就在宣布免费捆绑当天，网景的股票应声暴跌约三分之一，且此后再也未能重回巅峰。这一天成为了网景命运不可逆转的“最高水位线”。这一战役揭示了商业世界一条残酷至极的铁律：**如果你的核心分发渠道（Distribution）决定下场亲自与你竞争，并将其列为最高优先级战略，你的商业模式会在顷刻间化为乌有。** 为了扩大战果，微软在 1996 年雷厉风行地与当时把持拨号入口的几大主流在线服务商——AOL（美国在线）、CompuServe 和 Prodigy 达成排他性合作协议，迫使它们弃用原有浏览器并全线捆绑 IE。IE 的市场扩张势如破竹： - **1996 年底**：IE 市场份额突破 20%； - **1997 年**：IE 市场份额突破 40%； - **1998 年**：IE 市场份额突破 60%； - **到 2000 年前后**：IE 实际上斩获了全球近乎 100% 的浏览器市场份额，在彻底绞杀网景的同时，也将微软一步步推向了反垄断司法的风暴中心。 ### [43:08](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2588s) IE 市场份额的兴衰与网景商业模式的崩塌如果回头拉出 Internet Explorer 历年的市场份额走势图，那几乎是你见过形态最完美的一座“山丘”：从 1995 年的 0% 起步，到 2000 年飙升至接近 100%，随后在 2010 年一路跌回原点。微软究竟是如何丢掉曾经固若金汤的浏览器垄断地位的，这构成了下一阶段的核心叙事。但在进入那段历史之前，值得先停下来反思：网景（Netscape）的业务当年为什么会彻底枯竭？网景的商业模式本质上是靠销售服务器软件（Server Software）赚钱。而在当时，要卖出最好的服务器软件，前提是你必须牢牢掌控客户端（Client）。网站开发者和企业客户极为关注自己的网站能否在每个人都在用的软件环境中完美呈现；一旦网景不再能底气十足地声称“全网绝大多数网民都在用网景浏览器”，客户的算盘就变了：我为什么非得买你的服务器软件？我完全可以去市场上找价格更低、性价比更高、功能更丰富的替代品。网景在收入端的竞争优势瞬间瓦解。与此同时，面对股价 80% 甚至更大幅度的暴跌，任何初创企业都极难逆境翻盘——当初支撑狂热估值的是未来的想象空间，而非实打实的营收；一旦泡沫散去，员工手中的期权大幅缩水，人才流失便成为致命打击。反观微软，他们当时甚至根本没有改变自己的商业模式。微软没有去开辟全新的变现路径，仅仅是通过将 IE 与 Windows 捆绑，冷酷地扑灭了一个可能颠覆自己的潜在未来。他们依然做着老本行：继续通过 OEM 厂商向 PC 预装 Windows 许可证，或者在零售渠道向个人消费者卖软件盒子。 ```mermaid flowchart TD accTitle: 网景业务崩溃的恶性循环 accDescr: 解释网景失去客户端份额如何导致服务器收入崩盘及人才流失的因果关系链 subgraph Client["客户端失守"] IE["微软 Windows 免费捆绑 IE"] --> Drop["网景浏览器份额急剧暴跌"] end subgraph Revenue["营收端坍塌"] Drop --> ServerLoss["失去客户端支配力 企业客户不再采购网景服务器软件"] ServerLoss --> PriceWar["转向更具性价比竞品 商业模式造血能力枯竭"] end subgraph Capital["资本与人才崩盘"] PriceWar --> StockCrash["股价暴跌逾 80% 资本狂热退潮"] StockCrash --> TalentDrain["员工期权归零/人才流失 公司彻底丧失自救能力"] end ``` --- ### [44:47](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2687s) 1997 年 Macworld 大会：乔布斯、盖茨与拯救苹果的四项协议在这个关键节点上，出现了一段戏剧性极强的插曲。1997 年 8 月，著名的 Macworld 大会召开，刚刚重返苹果的史蒂夫·乔布斯（Steve Jobs）登台演讲，而比尔·盖茨巨大的面孔通过卫星连线出现在大屏幕上。这一幕早已载入科技史册，但如果透过当时的战略格局重新审视，背后隐藏着此前常被忽视的一层深意。盖茨与乔布斯在舞台上宣布的合作实际上包含四个核心要点，而非外界津津乐道的那一两个： - **第一点**：微软向苹果注资 1.5 亿美元。按照当时苹果仅约 20 亿美元的市值计算，微软直接收购了苹果约 8% 的无投票权优先股。 - **第二点**：微软做出为期 5 年的承诺，继续为 Mac 平台开发并发布 Office 办公套件（Office for Mac）。 - **第三点**：双方同意全面化解所有专利纠纷，为此前多年旷日持久的版权与专利官司彻底画上句号。 - **第四点**：Internet Explorer 将取代网景，成为苹果 Mac 电脑上的默认浏览器。这项默认设置从 1997 年一直延续到 2003 年苹果推出自研的 Safari 为止。前两项条款在当时几乎挽救了苹果的性命。当时的苹果正深陷破产边缘，而 Mac 用户对微软 Office 的依赖是生存级别的——如果微软决定停止为 Mac 开发 Office，消费者就会彻底放弃购买 Mac，苹果也将随之死亡。而在第四项条款中，盖茨真正的意图展露无遗。乔布斯对外界的说辞是：微软看好 Mac 作为 Office 的优秀运行平台，认可苹果并施以援手帮助伙伴渡过难关。但实际的谈判博弈更为老练精明：乔布斯敏锐地抓住了时机，暗示苹果手里还有未了结的技术侵权筹码，同时心知肚明美国司法部正因反垄断问题盯紧微软。苹果向微软释放信号：与其在反垄断监管下四面楚歌，不如展示出与生态巨头友好合作的姿态。而微软的回应则是，他们势必要让 IE 占领每一个角落。于是双方一拍即合——跨许可专利、免除诉讼，苹果拿到了活命的资金与 Office 支持，而微软则借此彻底将 IE 钉死在苹果的平台上，一举断了网景在 Mac 阵营的后路。 --- ### [48:17](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2897s) 网景谢幕与微软在消费科技领域的绝对顶峰 1998 年 11 月，网景的命运迎来了终局：美国在线（AOL）以略超 40 亿美元的全股票交易收购了网景。具有讽刺意味的是，此时距离时代华纳（Time Warner）与 AOL 那场世纪大并购仅剩一年多时间。这一刻标志着微软作为消费科技公司的绝对巅峰，甚至可能是整部人类消费科技史上前所未有的权力顶峰： - **对苹果的绝对生杀大权**：当时的苹果在生存上对微软产生着生死存亡般的依赖，微软一句话就能决定其去留。 - **消灭一切威胁**：网景被彻底击溃，微软成功“切断了它的氧气供应”，桌面互联网的入口被全面接管。 - **竞争真空期**：此时的科技版图上一片荒芜——距离谷歌成立还有 3 年，距离 Facebook 创立还有 9 年，市场上除了雅虎以及企业级市场的 Sun Microsystems 和 Oracle 之外，在整个消费级计算生态中，微软拥有毫无争议的终极权力。然而，这种依靠自身垄断优势直接“掐断对手氧气”的打法，也随之引来了一个无法回避的问题：一家公司真能凭借自己的体量随意宣判最具威胁的创新者死刑吗？这种行为在法律意义上难道不应该被判定为违法吗？ --- ### [49:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=2986s) 赞助商插播：JP Morgan Payments 的开发者平台讨论完 20 世纪 90 年代以 Windows 95 和 Web 为代表的“开发者平台”概念，节目顺势引入了本期赞助商摩根大通支付（JP Morgan Payments）的开发者平台。打造一个出色的开发者平台，核心在于赋能他人的企业文化以及超长期的战略投资眼光。 - **开发者话语权的崛起**：摩根大通每年在技术与研发上的投入高达 70 亿美元。现代数字化业务（如网约车、电商和在线旅游平台）高度依赖产品内原生的无缝支付体验，技术选型的决策权已经不仅属于传统财务人员，开发者、工程师和产品经理同样在支付平台的挑选上拥有了核心决策席位。 - **一站式开发者门户与沙箱环境**：摩根大通推出了全新的支付开发者门户（Payments Developer Portal），提供覆盖资金管理、商业运作、嵌入式金融乃至账户验证的端到端 API 路线图。开发者可以自行注册、创建项目、生成访问令牌，并在文档完备的沙箱（Sandbox）环境中测试资金收付接口。 - **巨大的业务体量与信誉**：背靠 200 年的银行经验，摩根大通每天处理的高达 10 万亿美元的资金流转，全美超过 50% 的电子商务交易都会流经其支付系统。 --- ### [52:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3177s) 巅峰之下的暴风雨：1998 年美国司法部反垄断诉讼拉开帷幕随着网景战事的落幕，历史终于走到了那个最具决定性的分水岭：1998 年美国司法部对微软发起的反垄断世纪审判（United States v. Microsoft Corp.）。回顾浏览器大战的收尾，主持人们不由得感慨，此前调侃“微软的做法难道不该算违法吗”其实正是当年整个司法审判的核心争论点。微软当时聚集在手中的市场支配力，在整个商业史上都前所未见。而这种前所未有的权力，也注定将这家庞大帝国推向反垄断风暴的最中央。 ### [53:26](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3206s) 微软垄断之辩：工程师的初心、公众指责与“消费者福利”标准当权力和市场份额集中在一家公司手中时，往往会引出核心法律问题：这是否违法？微软真的做错了吗？对于当年在微软一线写代码的绝大多数工程师而言，他们根本不考虑商业战略或竞争壁垒，他们的目标极其纯粹：“我只想发布优秀的软件，做出大家爱用的产品，并和我喜欢的人一起共事。”若去问他们，他们只会觉得大家只是在竭尽全力做出市场上最好的软件。然而，这种工程师视角与当时外界愈演愈烈的公众舆论产生了剧烈撕裂——在竞争对手的推波助澜下，微软被广泛描绘成一个仗势欺人的行业霸凌者（bully）。但是，商业感知上的“反竞争”与反垄断法层面上的“违法”完全是两码事。更深维度的讨论在于消费者体验：作为一名普通用户，在买到的操作系统中免费内置了一款世界级的浏览器，这难道不是一件好事吗？这直接触及了美国反垄断法体系的核心根基——“消费者福利标准”（Consumer Welfare Standard）。如果消费者的利益没有受损甚至获得了便利，垄断指控究竟是否成立？ --- ### [55:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3300s) FTC 的调查与司法部（DOJ）前所未有的接手微软的反垄断风暴并非始于 1998 年的司法部诉讼，其源头可以一直追溯到 1990 年的美国联邦贸易委员会（FTC）。当时 FTC 就微软涉嫌违反反垄断法展开了调查，核心针对的是所谓的“按处理器许可”（per-processor licensing）政策。历经数年调查，1993 年 7 月，FTC 委员会就“微软是否构成垄断且需采取进一步处罚措施”进行投票，结果以 2 比 2 形成死锁僵局。这意味着联邦政府不会对微软采取任何执法行动，对微软来说是一场重大胜利，针对微软的反垄断调查理论上应在此画上句号。出乎所有人意料的是，就在 FTC 放弃起诉的次月——1993 年 8 月，美国司法部（DOJ）突然打破常规，强行介入接手案件。一个联邦政府部门刚决定不起诉，另一个部门立刻跳出来说“我们觉得你们办得不对，我们要重新来查”，这让微软陷入了因同一罪名在联邦层面“被审判两次”的荒谬境地（宛如违反了刑事中的禁止双重追诉原则 Double Jeopardy）。数名 FTC 委员对此流程极为反感，甚至拒绝向司法部移交案件笔记与材料。尽管遭遇阻力，司法部仍紧咬不放。为了尽快息事宁人、结束漫长的政府纠缠，微软在 1994 年 7 月选择妥协，与司法部达成和解，双方签署了著名的同意令（Consent Decree）： - **禁止捆绑销售应用程序**：微软同意不将独立应用软件（如 Office 办公套件）的销售强行与 Windows 操作系统的销售捆绑。 - **保留功能集成自由**：同意令中明确允许微软继续在 Windows 操作系统中自由集成额外的新“功能”（features）。 ```mermaid flowchart TD accTitle: 微软反垄断调查与同意令的演变 accDescr: 展示从1990年FTC调查、僵局，到司法部介入并最终达成1994年同意令的历程 subgraph Phase1["FTC 调查阶段 (1990-1993)"] A["1990年 FTC 立案调查 (按处理器许可)"] --> B["1993年7月委员投票 2:2 僵局 (FTC 决定不采取行动)"] end subgraph Phase2["DOJ 介入与和解 (1993-1994)"] B -.->|"1993年8月 DOJ 破例接盘"| C["DOJ 展开独立调查"] C --> D["1994年7月双方达成和解 签署《同意令》 (Consent Decree)"] end subgraph Terms["同意令的核心条款界限"] D --- E["禁止捆绑销售应用程序 (Applications) 例: 不得强迫购买 Office"] D --- F["允许自由集成新功能 (Features) 埋下 Internet Explorer 争议隐患"] end ``` --- ### [58:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3490s) 灰色地带：Internet Explorer 到底是一个独立应用，还是一项系统功能？这一纸同意令直接埋下了未来数年最核心的法理争议：Internet Explorer（IE 浏览器）到底是一个应用（application product），还是一个功能（feature）？从软件平台演进的规律来看，操作系统的发展史本就是平台为了用户体验不断吸收过去由独立应用承担的功能。几年前还是独立软件的东西，几年后就会自然成为底层平台的一部分。在今天，任何人都无法想象买一台设备，其内置操作系统里竟然没有网络浏览器。然而在当时，这处于极度模棱两可的灰色地带。从比尔·盖茨（Bill Gates）与 J Allard 当年具有前瞻性的互联网备忘录来看，微软打从一开始就将浏览器视为 Windows 操作系统的核心组成部分；但不可否认的是，微软全力研发并捆绑自身浏览器的战略动机，恰恰是为了“切断潜在竞争对手的氧气供应”，保护 Windows 的垄断地位，防止被下一代网络平台架空。 --- ### [01:00:12](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3612s) 司法部发难、杰克逊法官的强制令与上诉法院的反转 1997 年 10 月，美国司法部在联邦地方法院提起动议，指控微软在 Windows 中捆绑 IE 浏览器的行为直接违反了 1994 年同意令中关于禁止捆绑销售产品的规定。司法部的诉求非常紧迫：不要等什么正在开发的 Windows 98 了，微软必须立即停止在当前出货的 Windows 系统中捆绑 IE。对此，微软辩称 IE 与 Windows 代码深度集成，根本无法直接剥离。当时用户在 Windows 资源管理器里直接输入网址就能解析网页，IE 已经深深嵌入了 Windows Shell；强行拆解会导致系统彻底崩溃。然而，微软在法律事实层面上也存在硬伤：最初出货的 Windows 95 版本里本来就没有 IE。审理该案的联邦地区法官托马斯·彭菲尔德·杰克逊（Thomas Penfield Jackson）并不买账，强令微软必须向 PC 制造厂商（OEM）提供一份完全移除 IE 的 Windows 版本，供厂商自行选择是否装机。微软无奈执行，直接把依赖的相关底层代码禁用，结果不出所料——系统出现了大面积崩溃和功能失效。这一举动让杰克逊法官极为愤怒，认为微软是在“钻法律字眼遵守条文，却公然蔑视法律精神”，态度傲慢无礼；而 PC 厂商自然也绝不可能出货这种残缺崩溃的系统。司法部随即要求法院判处微软藐视法庭。微软迅速提出上诉。1998 年 5 月初，联邦上诉法院作出有利于微软的裁决：推翻杰克逊法官的命令，裁定微软完全可以在 Windows 中继续捆绑 IE，只要这种集成能让消费者受益，微软就有权整合任何其希望提供的功能。上诉法院的判决再次锚定了反垄断法中的“消费者福利标准”——只要企业没有损害消费者利益、没有借此抬高价格，这种技术整合就是合法的。 --- ### [01:03:26](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3806s) 1998年5月18日：DOJ 祭出反垄断核武器全面宣战上诉法院的裁决宣布仅仅一周后，1998 年 5 月 18 日，美国司法部联合多州总检察长，正式对微软发起了全新的、规模极其庞大的全面反垄断诉讼，指控其违反《谢尔曼反垄断法》（Sherman Antitrust Act）并滥用垄断权力扼杀市场竞争。这一次，案件的范畴彻底跳出了“Windows 能否捆绑 IE 浏览器”的局部细节，而是上升到了对微软整个商业帝国的总清算： - 审查微软在整个 PC 操作系统领域是否属于非法垄断者； - 深入调查微软是否在其跨越操作系统、应用程序以及渠道合作的全部业务版图中，全面滥用这种垄断力量来排挤对手并损害长远的消费者利益。一场将深刻改变微软历史乃至整个科技行业发展轨迹的世纪反垄断世纪审判就此全面拉开帷幕。 ### [01:04:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3849s) 反垄断危机的全面爆发与微软的抗争心态在美国反垄断法体系下，单纯拥有垄断地位本身并不违法，真正违法的是“滥用垄断地位”。因此，审理微软案的法庭聚焦于两大核心问题：第一，微软究竟是不是垄断者？第二，微软是否存在滥用权力的行为？对微软而言，这场官司的走向极为严峻。在早期诉讼阶段，微软采取了极其强硬的法律策略——不仅立即对不利裁定提起上诉，还坚称自己“绝对没有做错任何事”，不肯退让半步。这种寸步不让的态度激怒了美国司法部（DOJ）。司法部的反应是：既然你连将 Internet Explorer（IE）与 Windows 操作系统捆绑这单单一件事都不愿妥协，那我们就对你的所有商业行为进行全面清算，直接发起规模庞大的世纪反垄断诉讼。然而站在微软内部的视角，公司上下弥漫着一种被自己国家政府背叛的委屈与愤怒。在员工们看来，这已经是第三次因为近乎相同的“罪名”被拉上法庭。大家忍不住抱怨：“之前不是说没有一事不再理（Double Jeopardy）吗？现在难道变成一事三理（Triple Jeopardy）了？这难道不是一个鼓励自由创业、建立商业帝国的国度吗？”此外，微软团队敏锐地意识到背后有深层政治博弈——消费者并未表现出强烈的不满，真正推波助澜的，是以 Netscape 为首的竞争对手在幕后进行的大规模政治游说。虽然借助反垄断武器攻击商业对手在科技史上屡见不鲜，但这种局面让微软下定决心：既然退无可退，那就硬碰硬，拼死抵抗到底。 ```mermaid flowchart TD accTitle: 司法部诉微软反垄断案件升级机制 accDescr: 展示微软的强硬策略与对手游说如何促使司法部提起全面反垄断诉讼并最终导致拆分裁决。 subgraph Conflict ["冲突诱因与升级"] A["微软早期强硬法务策略 （分毫不让、全盘否认、坚决上诉）"] --> B["司法部扩大调查范畴 （全面清算微软全部商业行为）"] C["竞对（Netscape等）幕后游说 指控微软不正当竞争"] --> B end subgraph Trial ["诉讼核心考量"] B --> D{"法庭双重审理核心"} D -->|"第一项"| E["是否具备垄断地位？ （PC OS 市场份额超90%）"] D -->|"第二项"| F["是否存在滥用行为？ （排他性捆绑与阻击替代品）"] end E & F --> G["2000年6月：最终裁决分拆微软 （拆分为 OS 公司与应用软件公司）"] ``` ### [01:06:25](https://www.youtube.com/watch?v=6oER1bdQa_s&t=3985s) 比尔·盖茨的作证录像与灾难性的法庭交锋 1998 年秋季，这起历史性的大审判正式拉开帷幕。在垄断地位的界定上，事实几乎无可争辩：微软在个人电脑操作系统市场的份额超过 90%，在任何法官眼中都拥有绝对的市场主导权。1998 年 8 月，负责审理此案的托马斯·彭菲尔德·杰克逊法官（Judge Thomas Penfield Jackson）发布了一项审前裁决：审判期间提交的所有庭前宣誓证词（deposition），必须仅以文字笔录形式呈交法庭。 1998 年 8 月 27 日起，司法部指定的特别检察官戴维·博伊斯（David Boies）对微软董事长兼 CEO 比尔·盖茨展开了长达 20 个小时的宣誓作证。基于“证词只会被录制为文字材料、不会公开播放录像”的前提，盖茨与微软法务团队制定了一套量身定做的作证策略：在字句上咬文嚼字，彻底避重就轻，绝不给出任何能被对方抓住把柄的实质性陈述。在这场历时数天的交锋中（如今在 YouTube 上仍能看到其中 12 个小时的完整录像），盖茨展现出对检察官和提问内容的极度蔑视： - 当检察官拿出一份文件提问备忘录（memorandum）时，盖茨死抠字眼，坚称这是电子邮件（email）而非备忘录，并以此为由拒绝回答。 - 当被问及发件人的意图时，盖茨双手一摊：“那是别人写的邮件，我怎么可能知道他在想什么？你应该去问他，我不知道。” - 甚至在辩论的高潮处，盖茨全程挂着讥诮的冷笑，直勾勾地盯着戴维·博伊斯，反问对方：“你到底如何定义‘定义’（definition）这个词？” 整场作证充斥着漫长的沉默、轻蔑的表情和对技术细节的无限钻牛角尖。在盖茨看来，只要字面上没有留下把柄，这次作证就算成功，他完全不在乎自己的态度显得多么刻薄。 ### [01:09:41](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4181s) 录像公开与公关灾难：大卫·博伊斯的致命一击然而，局势在 1998 年 10 月 9 日发生了灾难性的逆转——杰克逊法官突然推翻了自己此前的裁定，正式宣布允许在庭审中播放作证的现场录像。对于戴维·博伊斯这样的顶尖诉讼律师来说，这无异于天赐良机。博伊斯在庭审中将这套长达 20 小时的录像带运用得出神入化： - 每天法庭后排都坐满了来自各大媒体的记者。每当博伊斯需要给记者提供当天的爆炸性头条新闻时，他就会掐准时机播放一段盖茨傲慢辩驳的片段。 - 每当庭审气氛稍显沉闷，或微软证人试图博取法庭同情时，博伊斯就会抛出盖茨冷笑、翻白眼或死抠字眼的视频片段，瞬间摧毁法官和旁听者对微软残存的任何好感。 - 录像中盖茨表现出的粗鲁、学究气和目中无人，被全美媒体大肆报道，不仅让杰克逊法官在法庭上频频皱眉，更彻底重塑了公众对比尔·盖茨以及微软这家科技巨头的负面看法。对于杰克逊法官为何突然变更规则允许使用录像带，微软法务层曾产生过极深的不安与怀疑。根据记录这场审判的著作《World War 3.0》披露，微软内部一度认定杰克逊法官已成为“敌人”，甚至怀疑是司法部在私下施压促使法官改判。杰克逊法官断然否认了这一指控，但他也无法准确回忆起自己为何在 10 月 9 日改变主意。尽管微软私下怨声载道，但面对掌控生杀大权的法官，他们深怕进一步激怒对方，最终选择隐忍，并未对该程序变更发起正式的法律挑战。更关键的是，微软当时或许根本没有预料到，这几盒录像带的公开发酵会对盖茨的个人形象和整场官司带来多么毁灭性的打击。 ### [01:12:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4344s) 事实认定与拆分微软的判决落地经历了漫长而折磨的审理，1999 年 11 月，杰克逊法官正式公布了案件的“事实认定”（Findings of Fact），判定微软在 PC 操作系统领域确属绝对垄断者。判决书中详细论证了这一结论的技术与经济机制： - 微软庞大的用户安装基数（Installed Base）与极其繁荣的第三方应用软件生态，共同构成了难以逾越的“双边网络效应”。 - 这种强大的网络效应，使得任何潜在竞争对手若想开发出一款足以替代 Windows 的个人电脑操作系统，其所需投入的研发与生态迁移成本高昂到令人望而却步。这一事实认定的公布，意味着反垄断的另一只靴子即将落地。果不其然，数月之后的 2000 年 6 月 7 日，杰克逊法官正式下达了该案的最终判决（Final Judgment）：裁定微软确实滥用了其市场垄断地位。作为对其非法垄断行为的矫正惩罚，法庭下令**将微软拆分为两家完全独立的公司**——一家负责 Windows 操作系统业务，另一家专门负责 Office 等应用软件业务。这一拆分命令的震级，堪比 90 年前美国政府强行分拆洛克菲勒的标准石油公司（Standard Oil）。如今许多年轻的科技行业从业者或许已经淡忘，当年微软在司法部反垄断案中迎来的不仅是一场败诉，而是被联邦法院白纸黑字下达了勒令解体的拆分判决。 ### [01:14:34](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4474s) 拆分微软的疯狂判决与市值的腰斩按照美国联邦政府当时的裁决意向，微软将被一分为二，并且判决中包含了一系列令人匪夷所思的严苛条款：史蒂夫·鲍尔默（Steve Ballmer）必须去其中一家公司，比尔·盖茨（Bill Gates）则必须去另一家，两人绝不能在同一家公司共事。不仅如此，两人选定各自效力的公司后，还必须彻底剥离抛售所任职之外另一家公司的全部股份，以杜绝任何利益冲突。尽管这一切最终并未真正变成现实，但在那一特定历史时刻，这确实是美国政府的正式立场与官方诉求。主播们打趣道，这相当于政府试图凭空捏造出一个“盖茨公司”和一个“鲍尔默公司”——某种意义上，后来微软的演变确实走向了这两派风格的分野，只是形式完全不同。这一判决对微软市场价值与公众预期的打击是毁灭性的。从 1999 年末反垄断案“事实认定”（Findings of Fact）出炉后的 12 个月内，微软的市值从 6000 亿美元高位断崖式暴跌至 2700 亿美元，跌幅高达 55%。虽然这一暴跌与同期互联网泡沫破裂以及即将发生的 CEO 权力交接存在重合，但这项判决彻底击碎了微软作为“超级明星股”的无敌光环。主播们用当今的苹果公司作对比：如果法庭明天突然下令拆分苹果，强制将 iOS 操作系统与硬件设备剥离，要求用户能买到不带 iOS 的裸机手机，苹果的市值又会遭遇何等惨烈的雪崩。 - **拟定拆分方案核心要求**：将微软分割为两家独立实体；比尔·盖茨与史蒂夫·鲍尔默各自掌舵一家，禁止共事，且必须全部抛售非任职公司的股票。 - **市值崩塌数据**：12 个月内市值自 6000 亿美元骤降至 2700 亿美元，缩水 55%。 --- ### [01:16:16](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4576s) 杰克逊法官的“法官密谈门”与和解转机就在微软面临分崩离析的绝境之际，案件突然迎来了戏剧性的法律转折。2000 年 6 月，外界爆出主审法官托马斯·彭菲尔德·杰克逊（Thomas Penfield Jackson）在判决正式宣布前，曾长期在自己的法官办公室（Chambers）秘密会见多家媒体记者，并向他们透露案情与个人倾向。这导致裁决刚一落地，各家记者就能瞬间发出早已写就的独家爆料，这种严重违反司法程序与公正原则的行为彻底葬送了判决的合法性。上诉法院随即直接将杰克逊法官从该案中除名，并指派新法官重新审理。即便如此，在长达 15 到 16 个月的时间窗口里，全美乃至全世界所接收到的官方认知，依然是“微软必须遵照法令启动拆分准备”。随后，多股历史巨浪交织推动了事态向和解演变：上诉法院责令各方积极推动和解，而 2001 年发生的“9·11”恐怖袭击成为关键催化剂，国家危机促使各方意识到旷日持久的商业内耗必须尽早平息；加之联邦政权完成更迭，从小布什政府取代了克林顿政府，司法部的政策基调发生实质转变。具有讽刺意味的是这场反垄断马拉松的时间跨度：案件最初是针对 Windows 95 发起的，期间经历了 Windows 98 的发售（主播调侃这纯粹是一个为了赶开学季 PC 促销的营销更新），直到 Windows XP 正式发售数周后的 2001 年 11 月，美国司法部与微软才终于达成正式和解方案。 ```mermaid flowchart TD accTitle: 微软反垄断案诉讼与和解演变流程 accDescr: 描述杰克逊法官违规被撤换到最终达成和解的几个关键节点。 subgraph S1["司法危机与转折"] A["杰克逊法官秘密会见记者 （违规泄密遭撤职）"] --> B["上诉法院撤销拆分令 委派新法官重审"] end subgraph S2["外部催化因素"] C["9·11 事件爆发 （国家精力转向安全）"] D["克林顿政府交棒布什政府 （司法监管态度缓和）"] end subgraph S3["最终妥协"] B & C & D --> E["2001年11月达成和解 2002年批准5年期同意令"] end ``` --- ### [01:18:19](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4699s) 谁都不满意的和解协议与形式主义处罚 2001 年 11 月 2 日和解协议正式提出时，Internet Explorer（IE）在浏览器市场的占有率已经高达约 90%。这就让这场耗资巨大的世纪诉讼陷入了极其荒谬的尴尬境地：如果站在坚信“微软滥用垄断地位”的阵营来看，整个司法程序是一场彻头彻尾的溃败，因为伤害已经造成，网景（Netscape）已被彻底挤死；而如果站在微软与科技创新支持者的角度看，司法部的所作所为同样不可理喻——一家为个人电脑革命打造出最核心产品的创新企业，正面临互联网时代剧烈的新技术转型与生死存亡威胁，却被迫在背后绑着双手的状态下与政府打官司。 2002 年和解协议获得最终批准，不仅彻底推翻了拆分微软的禁令，其附加的 5 年期同意令（Consent Decree）实质约束力也微乎其微： 1. **禁止签订排他协议**：微软不得与 PC 原始设备制造商（OEM）签订排除竞争对手软件的排他性合同； 2. **保证系统互操作性与 API 开放**：微软必须编写详尽的 API 文档，保证 Windows 与非微软软件兼容互通，方便第三方开发者构建应用——主播嘲讽道，微软本质上就是个开发者平台，这本来就是微软赖以立身的日常业务； 3. **设立独立技术委员会**：成立一个独立的第三方技术委员会负责受理竞争对手的投诉举报——主播戏称“他们折腾了半天，其实就只是建了一条投诉热线”。 - **浏览器垄断结果**：达成和解时，IE 浏览器的市场份额已攀升至近 90%。 - **诉讼荒唐现实**：耗费数年、牵扯极广的官司，最终在法律条文层面上只换来了几乎不伤筋骨的合规清单与一条投诉热线。 --- ### [01:20:32](https://www.youtube.com/watch?v=6oER1bdQa_s&t=4832s) 真正的代价：企业文化坍塌与长达二十一年的法律枷锁虽然从表面协议上看微软侥幸过关，Windows XP 也成为一代传奇——终身售出逾 5 亿套，装机量跨越十亿台，统合了 NT 架构并留下经典的“Bliss”蓝天白云壁纸——但这场诉讼给微软造成的实际损失却是灾难性的。这不仅仅是巨额财务层面的耗损，而是彻底动摇了微软作为 20 世纪 90 年代全球最强大科技公司的灵魂与文化根基。整个公司经历了长达 16 个月“行将被肢解”的心理折磨。在公众视野中，比尔·盖茨在法庭作证录像中回避闪躲、神情尴尬的画面被媒体反复公开播放，微软的叙事形象从“全美最聪明的创新领袖”断崖式滑落为“贪婪邪恶的反竞争恶霸”，甚至连普通工程师都会被外界质疑“你为什么要在这种邪恶的公司上班”。在法庭上，著名律师戴维·博伊斯（David Boies）与司法部的策略极为纯粹——通过电子邮件和证词矛盾彻底摧毁微软高管的公信力，让外界觉得微软满嘴谎言、拒不认错；而微软却教条化地逐条反驳，甚至自欺欺人地坚称“网景从未构成生死存亡的威胁”，而盖茨此前公开撰写的备忘录里明明白白写着“网景是诞生于互联网的致命威胁”。这种全面被动的公关与道德指责反噬了微软内部，员工们开始动摇甚至自我怀疑：“我们真的只是在做好软件吗？还是背地里做了违法勾当而我毫不知情？” 在执行层面，巨大的合规与协议文档编制工作给产品开发套上了沉重的枷锁，研发迭代速度严重放缓。与此同时，私营企业与地方政府发起了铺天盖地的围剿： - **民事诉讼潮与巨额和解**：Sun Microsystems、AOL、RealNetworks 等竞争对手接连发起巨额民事反垄断索赔，迫使微软私下掏出了数十亿美元的和解金； - **各州总检察长追击**：多州总检察长不愿接受联邦和解协议，继续独立起诉微软，纠缠长达五年之久； - **欧盟反垄断战火**：大西洋彼岸的欧盟反垄断审查愈演愈烈，直到 2009 年双方才达成和解； - **漫长的监管阴影**：最终同意令直到 2011 年 5 月才正式失效。从 1990 年联邦贸易委员会（FTC）首次展开调查算起，直到 2011 年，微软长达 21 年的生命历程——也是其历史上绝大部分的黄金岁月——都笼罩在反垄断调查与诉讼的沉重阴云之下。 ### [01:24:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5094s) 反垄断的代价：比尔·盖茨的改变与全公司的“精神崩溃” 即便最终司法部反垄断案没有对微软的日常业务造成根本性的强制拆分或实质重构，微软在法律层面上可以说是“打赢了这一仗”，但这场持久战对公司的摧残依然是灾难性的。在当时的美国反垄断法框架下，关键的衡量标准是“消费者福利”是否受到实质损害，而微软将浏览器捆绑进操作系统的做法，固然残酷打击了 Netscape 等现有竞争对手以及潜在的新入局者，却很难证明直接伤害了消费者。然而，为了换取这个在法理上“正确的结局”，微软付出了难以估量的隐性代价。最致命的损失，莫过于这场浩劫彻底摧毁了微软最核心的护城河——比尔·盖茨本人。在整个科技行业，微软独一无二的竞争优势就是盖茨；但在亲历了反垄断调查的轮番轰炸、残酷的庭审质询以及长达数年的公开羞辱后，盖茨彻底变了，再也没能恢复到从前的状态。甚至在托马斯·彭菲尔德·杰克逊（Thomas Penfield Jackson）法官正式下达拆分判决之前，权力的交接就已经悄然开始：1998年7月，当司法部诉讼全面升温时，史蒂夫·鲍尔默被晋升为公司总裁，成为明确的二号人物；紧接着在1999年11月法庭认定微软具备垄断事实之后，盖茨于2000年1月13日正式宣布卸任 CEO，将帅印交接给鲍尔默，自己转任新设立的“首席软件架构师”（Chief Software Architect）并保留董事长职位。几个月后，法官便正式勒令将微软一拆为二。面对这样一场旷日持久、直击个人的围剿，整家公司的精力被彻底抽空。在分秒必争的科技行业，一旦全公司最顶尖的资源无法全力专注于研发和产品，失去清晰的“北极星战略”，失败便在所难免。正如一位经历过该时期的微软老员工所言，整整五年的诉讼拉锯战，其本质就是“整家微软公司的精神崩溃”。虽然不能将微软后续在消费电子领域的所有溃败都简单归咎于反垄断案，但这确实为理解当时微软为何“神志不清、心力交瘁”提供了最关键的时代背景。 ```mermaid flowchart TD accTitle: DOJ 反垄断案对微软领导层的冲击 accDescr: 展示从1998年司法部诉讼升温到盖茨退位、鲍尔默掌权的演变过程。 subgraph Phase1["司法部风暴 (1998-1999)"] A["1998年7月：DOJ 反垄断案升温 鲍尔默晋升为微软总裁"] --> B["1999年11月：法庭裁定垄断事实成立 盖茨心力交瘁"] end subgraph Phase2["权力交接与重压 (2000)"] B --> C["2000年1月13日：盖茨辞去 CEO 转任首席软件架构师与董事长"] C --> D["2000年年中：杰克逊法官下达拆分判决 全公司陷入长达15个月的拆分阴影"] end subgraph Phase3["新核心确立"] D --> E["史蒂夫·鲍尔默正式接任 CEO 成为公司的'情感支柱'与掌舵人"] end ``` ### [01:28:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5294s) 鲍尔默的三大任务：成为全公司的“情感支柱” 史蒂夫·鲍尔默正是在这样内忧外患的极端环境下走马上任的。在大众眼中的刻板印象里，鲍尔默往往是那个在舞台上大汗淋漓、疯狂跳跃大喊“开发者！开发者！开发者！”（Developers, developers, developers!）的狂躁形象；但许多经历过那个时期的微软老员工指出，外界很少意识到，鲍尔默在当时是整家公司的“情感定海神针”。那个著名的在台上疯狂起舞、高呼“我爱这家公司！”（I love this company!）的场面，实际上发生在2000年9月——当时所有人都坚信微软即将被法院强行拆解，士气跌至谷底。鲍尔默正是用这种甚至略带疯癫的极度热情，硬生生撑住了团队的信心。在长达15个月可能面临拆分的极端焦虑期内，微软竟然没有出现核心技术人才的大规模流失，这在硅谷历史上几乎是个奇迹。鲍尔默履新后，肩负着按优先级排列的三大历史使命： 1. **稳住全公司的情绪**：首要任务就是让员工每天还能有动力来上班，维系公司的凝聚力。 2. **收拾反垄断的法律烂摊子**：结束与监管机构的全面战争，寻求与政府达成和解。 3. **保持公司增长与产品领先**：在动荡中继续扩大营收和利润，让微软重回正轨。 ### [01:30:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5403s) 布拉德·史密斯的一页 PPT：从监管公敌到政府最信任的盟友为了完成第二项任务，鲍尔默上任后做出的最关键人事任命之一，就是提拔布拉德·史密斯（Brad Smith）担任总法律顾问（General Counsel，现任微软副董事长兼总裁），并给了他极其明确的指令：“去缔结和平”（go make peace）。在史密斯参加微软董事会面试时，他的整个竞聘汇报只有极其震撼的一页 PPT，上面写着简短的一句话：“是时候缔结和平了”（It's time to make peace）。史密斯清晰地认识到，团队必须彻底抛弃受害者心态。无论微软内部上下如何认为整个诉讼不公、过程如何荒谬，都必须翻过这一页，去找到微软能够承受的妥协方案并迅速全面和解。这是一次极其艰难的心智切割，而作为新老交替代表的史密斯展现出了非凡的战略眼光，不仅终结了诉讼，更彻底扭转了微软与全球政府的关系。 - **形象的彻底反转**：微软从90年代末反垄断靶场上的头号公敌，彻底蜕变为西方自由世界各国政府“最值得信赖的技术合作伙伴”。 - **监管维度的豁免地位**：在如今苹果、谷歌、Meta、亚马逊等科技巨头深陷反垄断调查泥潭的当下，微软反而是唯一一家未被美国联邦政府列入核心反垄断调查名单的巨头，并成为美国政府及其盟友最庞大的软件与服务供应商。 ### [01:32:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5531s) 走向企业级巨头：XP的辉煌与重磅赞助商 ServiceNow 在稳住军心并逐步化解监管危机后，鲍尔默转向了第三项任务：推动公司持续盈利并开拓新市场。尽管比尔·盖茨卸任 CEO 后依然以全职员工身份担任首席软件架构师，并在重大战略决策中深度参与技术路线的制定，但微软在后反垄断时代真正发生质变的，是其业务底层逻辑的转向——微软正大踏步从一家面向消费者的广义科技公司，蜕变为全球无与伦比的“企业级科技巨头”（The Enterprise technology company）。鲍尔默在过去十余年磨砺出的企业销售能力在这一阶段全面爆发，微软日后庞大的商业帝国，绝大多数都是在后司法部时代开拓的新领地中建立起来的。在这一关键转型期，微软不仅打造出了历史上最成功的操作系统 Windows XP，同时也拉开了后续 Vista、Zune、Bing 搜索、Windows Mobile、Windows 8 等一系列波澜壮阔战役的序幕。在本段结尾，节目还特别介绍了本季赞助商 ServiceNow： - **企业级 AI 平台**：ServiceNow 专注于业务流程自动化、提升服务交付和运营效率，超过85%的财富500强企业都在运行其系统，它已与微软并列成为全球最重要的企业级技术供应商之一。 - **与微软深度生态整合**：两家公司宣布扩大合作，将其企业级 AI 助手全面打通——用户自秋季起可在 Microsoft Copilot 中直接调用 ServiceNow 的 Now Assist，并首批整合进包括 Microsoft Teams 在内的各类 Office 办公应用。 - **历史回响的峰会亮相**：萨提亚·纳德拉（Satya Nadella）亲自出席了 ServiceNow 举办的 Knowledge 年度盛会并发表主题演讲，这一场景令人不禁回想起1997年比尔·盖茨在 Macworld 上通过视频连线宣布为 Mac 开发 Office、拯救陷入绝境的苹果的历史性时刻。 ### [01:35:07](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5707s) ServiceNow 与微软 Copilot 的深度集成主持人以 ServiceNow 与微软 Copilot 的企业级双向集成为切入点展开赞助商口播。当前两者的 AI 能够在不直接共享底层敏感数据的前提下，实现天衣无缝的协同体验。例如，当企业员工在 Teams 中向 Copilot 询问公司的笔记本电脑领用政策时，Copilot 会在后台将该请求与上下文无缝传递给 ServiceNow 的 Now Assist；Now Assist 根据该员工的具体权限读取公司内部合规政策，再将答案以富卡片（Rich Card）的形式返回给 Copilot，员工可以直接在卡片中触发流程申请。未来，微软 Copilot 还将反向深度嵌入 Now Assist，允许用户直接基于 ServiceNow 平台内部沉淀的资产与知识库，自动生成包含 PowerPoint 演示文稿、Excel 表格在内的各类 Office 文档。 --- ### [01:36:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5769s) 1996 年微软的四项业务构成与“后台办公室”的雏形为了理解微软后来庞大的企业级业务究竟是如何建立起来的，需要先梳理在司法部（DOJ）反垄断案爆发前微软原有的业务版图。在 1996 年比尔·盖茨接受《连线》（Wired）杂志凯文·凯利（Kevin Kelly）的专访中，盖茨系统性地拆解了当时微软的业务架构。彼时微软超过 90% 的营收仅由三块支柱业务贡献： 1. **Windows**：桌面客户端操作系统。 2. **Office**：当时年收入已达 40 亿美元的办公套件。 3. **NT / BackOffice**：即服务器与企业级后端业务。剩余不到 10% 的长尾业务则涵盖 MSN、电子商务、电脑游戏、多媒体百科全书（Encarta）、地图、以及与梦工厂（DreamWorks）和 NBC（MSNBC）的合资项目。盖茨在当时敏锐地指出，以“NT / BackOffice”为代表的服务器业务是全公司增长最快的部分，增速甚至超越了 Windows 与 Office 本身。 ```mermaid flowchart TD accTitle: 1996年微软业务营收构成 accDescr: 展示1996年微软四大业务板块及营收占比结构，其中核心三大业务占比超90% subgraph Core["核心三大支柱业务 (占比 > 90%)"] W["Windows 客户端"] O["Office 办公套件 (~40亿美元/年)"] NT["Windows NT / BackOffice\n(增速最快的服务器业务)"] end subgraph Other["长尾实验性业务 (占比 < 10%)"] M["MSN / 电商 / 游戏 / 地图 / 百科"] JV["合资公司 (DreamWorks, MSNBC)"] end Core --> MS["微软总体营收 (1996)"] Other --> MS ``` 这一时期盖茨依然习惯性地用具体产品名“NT / BackOffice”来统称整个企业级市场，这也侧面暴露了当时盖茨的产品技术思维局限：在盖茨眼中它只是“企业采购的一堆杂七杂八的东西”，而真正对企业市场具备深刻战略热情、并将其塑造成体系化增长引擎的，正是史蒂夫·鲍尔默（Steve Ballmer）。当时微软虽然意识到了这块金矿，但尚未摸索出成熟的销售通路（Go-to-market）、企业级定价模式与配套的服务组织架构。 --- ### [01:37:55](https://www.youtube.com/watch?v=6oER1bdQa_s&t=5875s) 从个人电脑客户端走向企业 IT：三边飞轮的构建由戴夫·卡特勒（Dave Cutler）主导编写的 Windows NT 早期虽然性能强劲、专为工作组网络设计，但其本质上依然是运行在高性能 PC 上的客户端操作系统，依然受限于微软前 25 年“每个桌面一台电脑”的 PC 中心主义。进入鲍尔默时代后，微软迎来了决定性的认知跃迁：**企业级市场服务的核心不是终端用户，而是 IT 部门与系统级基础设施（Systems）**。在司法部反垄断诉讼之后，微软在客户端操作系统领域腹背受敌、处处受限，但在服务器与 IT 系统市场，微软面对的是 Sun Microsystems、IBM 和 Oracle 等巨头。当时微软在很多企业服务细分领域的份额几乎为零，意味着他们可以放开手脚主动进攻。这片战场与鲍尔默的商业化组织能力完美契合。在鲍尔默任内，微软相继打造出了一系列日后各自年营收突破数十亿美元的庞大企业级产品矩阵： - **SQL Server**（企业级数据库） - **Active Directory**（目录服务与身份认证） - **Exchange**（企业邮件与日程服务） - **Dynamics**（企业资源与客户关系管理） - **SharePoint**（企业协作平台，归属于 Office 部门但具备系统级平台属性）这些后端服务与桌面端的 Windows PC 以及 Office 办公套件形成了紧密的共生关系，铸就了科技史上极为强大的“三边飞轮”生态闭环： ```mermaid flowchart LR accTitle: 微软客户端到服务器三边协同飞轮 accDescr: 展现Windows操作系统、Office套件与企业后端服务器组件之间的闭环联动体系 A["客户端操作系统\n(Windows NT / PC)"] --> B["生产力应用套件\n(Office / Outlook)"] B --> C["企业后端IT基础设施\n(SQL Server / Exchange / Active Directory)"] C --> A ``` - **全面协同压制单点竞品**：企业 IT 部门发现，微软的服务器套件与 Windows PC 及 Office 结合得天衣无缝。采购整套系统意味着没有任何单个组件必须做到行业顶尖（Best of breed），因为“一体化协同且来自单一可靠供应商”的低摩擦成本，足以抹平与其他独立单点方案（Point solutions）的功能差距。 - **牢不可破的 IT 锁定**：一旦企业转向微软全家桶，就极难替换。无论是其它厂商的目录服务还是第三方邮件服务器，在这一套打通的生态面前都丧失了竞争力。 --- ### [01:41:44](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6104s) SQL Server、Exchange 与 Active Directory 的连环组合拳这套企业级攻势最早打响于数据库领域。1998 年微软推出经过彻底重构的 SQL Server，使其成为首款真正具备企业级性能、足以叫板 IBM 大型机数据库及 Oracle 的产品。更关键的是，不同于 IBM 昂贵的大型机架构，SQL Server 原生运行在廉价的 x86（Intel）通用硬件上。微软向企业 IT 提出了极具杀伤力的“总体拥有成本”（TCO）主张：企业无需再为 IBM 大型机支付高昂费用，只需从戴尔（Dell）等厂商采购廉价的 x86 Windows 服务器，就能搭建全套 IT 架构。随后，微软依托底层架构在应用层发起了决定性战役： 1. **Exchange 与 Outlook 攻陷企业通信**：正值电子邮件成为企业“杀手级应用”的黄金时代，市场上原本充斥着雷·奥兹（Ray Ozzie）开发、被 IBM 以 35 亿美元天价收购的 Lotus Notes。微软直接向企业客户推销 Exchange——它不仅包含邮件服务，还深度集成了日历与通讯录，更重要的是，其客户端 Outlook 直接免费打包在企业必买的 Office 办公套件中。这种“免额外决策、无缝兼容”的杀手锏让 Exchange 近乎“自我推销（sells itself）”。 2. **Active Directory 构筑终极身份护城河**：当企业的数据库、邮件、日历与 PC 终端全部运行在微软体系上时，IT 管理员急需掌控组织内每一个员工的权限、设备访问安全与邮件路由。微软顺势推出了 Active Directory（活动目录）。一旦一家企业的 IT 将 Active Directory 作为人员、权限与资产的“唯一真实数据源（Source of Truth）”，整家公司的软件选型就会彻底“倾斜”，后续不论采购任何内部系统，首要标准都变成了“是否支持 Active Directory 认证”。 3. **企业协议（Enterprise Agreement）的捆绑模式**：鲍尔默更开创了里程碑式的企业协议（EA）打包销售模式。企业一次性签订长期大额合同，即可解锁整个产品库。哪怕客户最初只使用了其中 30% 的组件，一旦未来产生新的业务场景，由于对应功能的微软产品早已包含在已付费的套餐中“免费奉送”，企业自然会毫不犹豫地优先采用微软产品，从而将竞争对手彻底挤出采购清单。 ### [01:45:52](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6352s) 企业级捆绑与商业模式的飞轮在讨论微软的发展历程时，人们往往容易将其贬低为一家“庞大且无聊的企业级公司”（big boring Enterprise company），但实际上，微软在此期间所实现的商业模式创新在整个科技史上都名列前茅。在 IBM 统治企业级市场的时代，科技系统中甚至没有真正意义上的“最终用户”（End Users）；而微软则在这个全新时代中，探索出如何向拥有海量实际用户的企业组织销售并构建庞大的软件系统。微软商业模式创新的核心在于：软件本身具有零边际成本（Zero marginal cost）的特性。这意味着微软向捆绑包中塞入更多功能和新产品时，自身几乎不承担任何边际开支。通过将丰富的产品组合打包销售，微软不仅能极大地增强客户粘性，将客户牢牢锁定在自己的生态中，还能伴随客户企业的人员扩张实现“获取并扩展”（Land and expand）。这种基于捆绑销售的打法，为微软创造了长久、高韧性且复利增长的现金流，最终转化为极其可观的企业价值。 ```mermaid flowchart TD accTitle: 微软企业级业务飞轮 accDescr: 描述微软如何利用零边际成本和企业协议构建高粘性、长周期的收入增长飞轮。 subgraph Flywheel["企业级业务复利飞轮"] A["零边际成本软件产品线 (Windows + Office + 后台软件)"] B["高性价比全家桶捆绑 (企业协议 EA)"] C["极高客户粘性 (切换成本高昂)"] D["锁定3年年金式长约 (随员工人数规模自然扩张)"] end A --> B B --> C C --> D D --> A ``` - **统一销售摊薄进入市场成本（GTM）**：微软无需针对每款软件单独派遣销售团队，而是在敲开大企业大门时将 Windows、Office 以及整套软件套件一次性推销出去，将整个 Go-to-Market 成本有效摊薄到所有产品线上。 - **由一次性买断转为年金流**：微软推出了“企业协议”（Enterprise Agreement，简称 EA），不再按单件软件售卖光盘，而是根据企业规模按每名员工（早期按设备）收取固定年费，成功将以往不稳定的单次软件销售转变为持续增长的年度年金。 --- ### [01:48:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6491s) 三年期 EA 协议与 IT 管理员的诉求博弈微软 Enterprise Agreement（EA）的一个关键机制是其标准的“三年期协议”。这一商业设计与微软核心产品的发布节奏达成了令人惊叹的完美协同：Windows 和 Office 通常大约每三年推出一个重大版本升级。这就意味着，无论客户在哪个时间点签署或续签这套三年协议，在其合同生命周期内都能确保至少获得一次重大的全套软件升级。然而，企业级市场的运作逻辑中存在着一个根本性的核心矛盾——**普通员工的需求与企业 IT 管理员的诉求完全对立**： 1. **员工（用户）的本能诉求**：普通员工在办公电脑上倾向于追求使用便利、尽可能少的系统限制、上网浏览、玩点小游戏摸鱼，甚至为了提高个人工作效率不惜牺牲系统安全性（例如运行自定义的 VBA 宏代码，乃至使用未经授权的软件）。 2. **IT 管理员的管控目标**：企业 IT 部门面对的是成千上万可能引发灾难的“流氓代理人”（Rogue agents）。他们最关心的不是员工的个人爽快度，而是安全合规、隐私保护、防范黑客入侵、避免吃官司以及防止数据泄露。IT 部门需要的是绝对的控制权：禁止随意升级、禁止自行安装软件、禁止运行宏、封锁特定网址。微软敏锐地抓住了 IT 买家的这一痛点，通过企业协议为客户提供高度定制化、可深度管控的系统环境，完美迎合了出资购买软件的 IT 决策层。 --- ### [01:50:40](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6640s) 创新的代价：向后兼容与保守的“牢笼” 面向 IT 管理者的销售模式虽然带来了巨大的商业利益，但也给微软的创新机制套上了沉重的枷锁： - **新功能的“阻力”**：普通用户渴望创新的前沿功能，但 IT 管理员却对“新事物”充满警惕——任何新的 UI 界面或功能都会带来巨大的培训成本，还会增加员工求助客服的困惑。前 Office 业务高管曾直言不讳：“当时在 Office 团队，我总觉得哪怕我们整整十年不添加任何新功能，一切也照样运转如常，根本没人会注意到；甚至人们可能还愿意为此付更多钱。” - **销售团队抵制产品迭代**：正如 Steven Sinofsky 在《Hardcore Software》一书及 Substack 专栏中所述，产品经理们辛辛苦苦根据用户研究做出了自认为用户会喜欢的新功能，结果销售人员却会跑回办公室愤怒抗议：“千万别加这个！你们是在开玩笑吗？如果非要逼我客户接受这种全新的功能界面（例如后来的功能区 Ribbon）或大幅变动，我在销售环节会遭遇铺天盖地的反对！”因此，企业端的产品改动只能被压缩为微小、渐进式的演进。 - **华尔街投行的真实日常**：主播 David 分享了自己在 2007 年刚从大学毕业、入职华尔街瑞银集团（UBS）担任投资银行分析师的真实经历。在 2007 年，他们办公用笔记本被 IT 系统锁得密不透风，依然运行着老旧的 Windows XP 和 Office 2003。内部网络布满防火墙，员工无法安装任何软件，大量外部网站被封锁。起初分析师们还能摸鱼访问 Miniclip.com 玩小游戏，但很快也遭到了 IT 部门的全面封杀。 - **向后兼容性（Backward Compatibility）的双刃剑**：企业客户最看重的是总拥有成本（TCO）与系统的向下兼容性。微软向企业许诺：无论企业基于以往系统做过何种魔改、依赖哪些历史软件，微软都保证“绝不破坏兼容性，并持续提供支持”。这构成了微软坚不可摧的护城河，但同时也埋下了巨大的隐患——这种被深度绑定的向后兼容性包袱，在日后移动化与现代化转型的大浪潮中，反过来极深地拖累了微软自身的发展步伐。 --- ### [01:53:59](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6839s) 2007 年的收入版图与 OEM 商业模式的对比在史蒂夫·鲍尔默（Steve Ballmer）出任首席执行官的最初七年内，微软的企业级业务完成了一场彻底的乾坤大挪移。到了 2007 财年，分析师估计在微软大约 510 亿美元的总收入中，收入构成发生了颠覆性的变化： | 收入来源 | 收入占比（约数） | 业务特征与客户群体 | |:--- |:--- |:--- | | **多年期企业协议（Multi-year EA）** | **40%** | 三年期企业长约，涵盖 Windows、Office 等绝大部分套件，提供稳定的现金流年金 | | **单年期企业协议（Single-year EA）** | **15%** | 针对企业客户的短期或按年签署的企业协议 | | **OEM 预装业务** | **30%** | 向戴尔（Dell）、联想（Lenovo）等 PC 制造厂商授权 Windows，由 OEM 负责消费级分发 | | **其他所有业务总和** | **15%** | 包括 Xbox 主机游戏业务、在线服务等其他边缘项目 | - **企业级年金占比过半**：到 2007 年，多年期与单年期 EA 合计已经占到微软全公司营收的 **55%** 以上。在鲍尔默掌舵的短短七年间，企业协议成功成为整个公司的第一大顶梁柱。 - **OEM 模式与零售盒装软件的毛利对比**：早在 90 年代，微软就曾算过一笔账，决定放弃向消费者单独售卖实体盒装软件的笨拙方式，转而让 PC 硬件制造商替微软完成市场分销。在 90 年代，微软通过零售渠道销售盒装 Windows（Boxed software）的毛利率（Gross margin）仅有极其凄惨的 **29%**（需承担包装、仓储、分销及零售让利等巨大成本）；而直接向 OEM 授权操作系统则在后来演变为毛利率极高的轻资产数字授权生意。 ### [01:56:13](https://www.youtube.com/watch?v=6oER1bdQa_s&t=6973s) 告别盒装零售：企业协议与 OEM 的商业飞轮在传统的盒装软件零售模式下，微软必须承担实打实的物理成本：压制软盘或光盘、包装盒制作、物流运输、与零售商分成，以及巨额的销售与市场营销支出。这绝非零分销成本、零边际成本的生意，其毛利率仅有惨淡的 29% 左右。然而，通过 PC 制造商的 OEM 渠道销售时，毛利率一跃升至 75%。微软只需向硬件厂商交付一次系统底层代码（bits），PC 厂商便会搞定后续的所有分发工作。这种模式不仅带来了 75% 对比 29% 的毛利率碾压，更具备无与伦比的规模效应。正如 Visa 的“网络之网络”（network of networks）效应一样，微软只需与每家 OEM 厂商敲定协议，各家厂商在市场上的自发扩张就会直接反哺微软，而微软自身无需额外付出扩展成本。到 2007 年，微软约 85% 的收入来自于针对大客户的企业协议（Enterprise Agreement, 简称 EA）和 OEM 预装渠道。微软果断将低效的盒装软件抛弃在一边，销售团队全面转向捕鲸式的大企业客户公关。企业协议不仅利润丰厚，更以极高确定性的“年度年金”（annual annuities）形式入账。微软彻底抛弃了最糟糕的卖软件方式，全力押注在两种最顶级的商业模式上。而在司法部反垄断诉讼之后摸索出的这套企业销售体系，到 2007 年贡献了公司超过一半的营业收入。 ```mermaid flowchart TD accTitle: 盒装零售与 OEM/企业协议商业模式对比 accDescr: 对比传统零售盒装软件的高成本低毛利与OEM企业协议的规模效应及高利润结构。 subgraph RETAIL["传统盒装零售模式（毛利率 ~29%）"] R1["物理介质制作 (软盘/光盘)"] --> R2["包装与实体分发 (物流至零售商)"] R2 --> R3["零售商渠道分成 + 高昂市场营销"] end subgraph SCALE["现代商业模式（毛利率 ~75% / 年金制）"] O1["OEM 渠道分发"] --> O2["一次交付代码镜像 硬件厂商自行预装扩展"] E1["大企业销售协议 (EA)"] --> E2["捕鲸式大单销售 高利润、持续年度年金"] end ``` --- ### [01:57:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7074s) 企业化的代价与 Windows XP 的合流壮举这种全面拥抱企业级客户的战略也潜藏着巨大的隐患：当企业决策者成为真正的买家时，软件的体验质量（尤其是面向普通终端用户的体验）便不再是公司的首要优先级。在 2007 年 iPhone 问世颠覆行业格局之前，这一缺陷尚未全面暴露。理解微软在消费级软件领域的起伏，最好的切入点是 Windows 的演进历程。Windows XP 的诞生堪称技术与商业上的双重奇迹。在 1998 至 1999 年反垄断诉讼激战正酣、2000 年法官勒令拆分、2001 年达成和解协议的狂风暴雨中，微软竟然在重压之下开发并推出了这款极其出色的操作系统。从技术架构上看，微软在之前的近十年里一直在维护两套并行的代码库： - 一套是面向企业环境的 Windows NT（如 Windows 2000），底层稳健强大，但需要高配置的企业级 PC，界面生硬且不够亲民； - 另一套则是面向普通消费者的 Windows 9x 系列（Windows 95、Windows 98），虽然拥有图形友好的交互界面，底层却缺乏真正的现代系统功能与稳定性。尽管两者对外暴露了相同的 API 供开发者编写程序，但两套系统的实现机制、底层兼容性及用户体验完全是两码事。Windows XP 完成了不可思议的统一工程：它将源自 9x 时代极具亲和力的界面体验，完全移植到了稳固强大的 Windows NT 内核之上。这宣告了源自 Windows 3.0 乃至更早 Interface Manager 时代的消费级 9x 代码库彻底终结，全线产品合流至 NT 体系。 - **产品版本细分**：XP 分为家庭版（Home）和专业版（Professional）。许多极客用户即便不用任何企业功能，自装电脑时也抱有“必须买 Pro 版”的心理。 - **拥抱数字生活**：XP 强化了数字摄影、数字音乐、数字视频以及家庭网络功能。尽管苹果后来凭借“数字枢纽”（Digital Hub）概念打响名堂，但当时无数普通用户正是通过 Windows XP 将数码相机里的照片导入电脑，或者在 XP 桌面上整日挂着 Napster 客户端下载音乐。 --- ### [02:02:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7323s) 911 阴影下的纽约发布会与“以 Windows 为中心”的旧信条 Windows XP 的发布会呈现出科技史上极为罕见而怪异的氛围。它既不同于乔布斯精心雕琢的苹果 Keynote，也不同于当下的 Google I/O。发布会开场是一支福音合唱团深情演唱《美利坚美丽》（America the Beautiful），紧接着比尔·盖茨与时任纽约市长鲁迪·朱利安尼（Rudy Giuliani）并肩走上舞台，共同发表反对恐怖主义的演说。这一切的根源在于发布时间：2001 年 10 月，地点纽约——距离“9·11”恐怖袭击发生仅仅过去了一个月。面对这场筹备三年的重磅旗舰发布，微软当时不得不面对是否应该如期举办、是否应将全场焦点转向致敬一线救援人员的巨大抉择。这场发布会还透露出微软权力结构的一丝微妙：当时史蒂夫·鲍尔默已正式出任微软 CEO，但登台与朱利安尼共同为划时代产品揭幕的依然是创始人兼公众代言人比尔·盖茨。伴随 Windows XP 一同面世的还有新版 Office。这是微软熟稔于心的看家绝技：让顶级办公套件在新系统发布首日提供最佳支持，既借此抬高操作系统的实用价值，又反过来帮助 Office 借助新操作系统的普及抢占市场份额。这一策略源于 DOS 时代的惨痛教训——当时 Lotus 1-2-3 和 WordPerfect 在 DOS 平台上把微软自家的生产力软件打得毫无还手之力。自开发 Windows 以来，微软便立下誓言决不再重蹈覆辙。然而，这种“微软因 Windows 的繁荣而繁荣”（As Windows goes, so goes the company）的理念，如同迪士尼过去坚守“动画乃立足之本”一样，逐渐演变为了微软内部不可动摇的宗教教条。在企业市场中，这一逻辑被进一步放大：所有的 Windows 设备构成了企业协议的核心，而所有企业级服务器技术的卖点都在于与本地 Windows 网络无缝协同。与此同时，Windows XP 还享受到了 PC 行业历史上最惊人的一波时代红利。从 1975 年微软创立到 1986 年上市的 11 年间，全球 PC 出货量以每年约 98% 的复合速度近乎翻倍狂飙。即便步入 2001 年，全美乃至全球家庭普及 PC 的巨大时代顺风依然在强力托举着 Windows XP。 ### [02:06:47](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7607s) 历史性的 PC 浪潮与绝对垄断在探讨微软的业务扩张之前，回看个人电脑（PC）在美国的渗透率数据令人震撼。在微软 IPO 之前，美国家庭的 PC 普及率仅有 8% 左右；即使经历了一年又一年翻倍式的爆炸增长，到微软 1997 年上市十三年后，全美家庭的 PC 渗透率也仅仅达到 37%。而在 Windows XP 发布两年后的 2003 年，这一数字才爬升至 62%。2003 年听起来已经是相当现代的年份，但当时依然有近四成的美国家庭甚至没有一台电脑。对于当时的微软而言，个人电脑浪潮是人类商业史上最宏大的长期趋势之一。更关键的是，微软在这个市场中拥有被美国司法部认定的绝对垄断地位。在这一时期，微软根本不需要操心如何拓展市场份额——只要 PC 市场的总盘子在扩张，微软的业绩就会亦步亦趋地水涨船高，它就像是整个个人电脑行业的一支杠杆指数基金。 - 微软 IPO 前：全美 PC 家庭渗透率约 8% - 微软 IPO 节点：渗透率仅约 10% - 1997 年：渗透率攀升至 37% - 2003 年（Windows XP 周期）：渗透率达到 62% --- ### [02:08:06](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7686s) DirectX 的技术奇迹与主机战场的开端许多长期关注微软的听众一直期盼节目能深入探讨 Xbox。事实上，微软在游戏领域的野心远早于 Xbox 主机本身的诞生。在 Windows 95 时代，微软推出了意义深远的 DirectX API，这套极其精妙的技术允许开发者完全绕开 Windows 繁琐的底层系统直接调度硬件加速。在此之前，像约翰·卡马克（John Carmack）开发《毁灭战士》（Doom）完全是依靠天纵奇才的算法硬生生在没有硬件加速的 PC 架构上跑出第一人称射击效果；而 DirectX 的出现直接催生了 90 年代末 PC 游戏的黄金时代，《雷神之锤》（Quake）、《反恐精英》（Counter-Strike）以及后来的《半条命》（Half-Life）皆受惠于此。这股在 PC 游戏领域的绝对优势顺理成章地延展到了家庭游戏机市场。2001 年 11 月，就在 Windows XP 推出仅仅几周之后，初代 Xbox 正式发售。那正是微软历史上最魔幻的时刻：在华盛顿特区，微软正面临着险些被美国反垄断部门强制肢解的生死判决；而在雷德蒙德，工程师们不仅推出了凝聚八年心血的划时代操作系统 Windows XP，还同时推出了一台硬核的游戏主机。微软凭借在反垄断审判前积攒的庞大势能与顶尖工程实力，依然在消费科技领域展现出无人可及的统治力。 --- ### [02:10:02](https://www.youtube.com/watch?v=6oER1bdQa_s&t=7802s) Xbox 的财务底色与人才资源的大挪移从商业角度审视，Xbox 虽然在主流文化中大获成功，但对微软整个帝国的财务贡献却十分尴尬。翻阅微软旧财务报表中的“娱乐与设备事业部”（Entertainment and Devices）可以发现，Xbox 本质上是一项长期处于“全寿命保本”边缘的业务： | 财年 | 营收规模 | 营业利润/亏损 | 阶段背景与表现 | |:--- |:--- |:--- |:--- | | **2006** | 40 亿美元 | **亏损 14 亿美元** | 初代 Xbox 推出 5 年后仍严重亏损 | | **2008** | 80 亿美元 | 利润 4 亿美元 | Xbox 360 步入成熟期，但利润率极低（约 5%） | | **2009** | 80 亿美元 | 利润 1 亿美元 | 营收停滞，利润大幅缩水 | | **2010** | 80 亿美元 | 利润 7 亿美元 | 处于阶段性高点，但同期微软整体利润高达约 200 亿美元 | 当两位主持人在筹备这期节目与史蒂夫·鲍尔默（Steve Ballmer）对谈时，鲍尔默曾轻描淡写地回应过一些未达预期的投资与收购：“我们不过是亏了点钱而已（We only lost money）。”在微软这种年利润数百亿美元的现金印钞机面前，资本从不是稀缺资源。正如比尔·盖茨（Bill Gates）曾言，微软自创立第三年起就从未缺过资金，管理层唯一面临的硬约束是“公司是否有足够聪明的人才去把事情办成”。 ```mermaid flowchart TD accTitle: Xbox 的得失权衡 accDescr: 展示 Xbox 消耗顶级人才与它在在线服务及品牌年轻化方面带来的战略溢出 subgraph Cost["代价与消耗"] A["顶尖产品与软硬件设计人才"] --> B["Xbox 事业部 (高强度研发)"] B --> C["微薄的财务回报 (长期盈亏平衡)"] end subgraph Gains["战略溢出效应"] B --> D["Xbox Live (2012年达4000万订阅)"] D --> E["大规模云端与订阅服务经验 (反哺 Azure)"] B --> F["捕获年轻消费者心智 (品牌免于边缘化)"] end ``` 正是因为人才才是最稀缺的资源，Xbox 在某种程度上成了微软的“人才黑洞”。为了研发主机和 Xbox Live（到 2012 年已拥有 4000 万付费订阅用户），微软集中了一批对审美、物理硬件、软硬件结合与工程架构有着极高要求的产品天才和拼命三郎。这些人才如果被投放回微软其他核心业务，原本可能会产生更大的协同效益。但不可否认的是，Xbox 留下了两笔无价的战略遗产：第一，它是人类科技史上最早的大规模在线订阅服务之一，其运维海量互联网服务的技术基因直接启发并奠定了后来的 Azure 云计算；第二，它让微软在年轻一代消费者的精神生活中保持了强大的存在感。 --- ### [02:14:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8097s) 从 Whistler 到 Longhorn：滑雪胜地代号与 Vista 的梦魇前奏谈及 Windows XP 之后的演进，便绕不开微软标志性的滑雪胜地开发代号文化。温哥华附近著名的双子滑雪名山是西雅图人与微软员工最爱的度假地。Windows XP 的开发代号正是其中的一座山峰——惠斯勒（Whistler）；而团队原本雄心勃勃规划紧随其后、在两三年内推出的宏大换代系统，则顺理成章地以相邻的黑梳山（Blackcomb）命名。然而，Blackcomb 的架构设想极其庞大复杂，工程难度超乎想象。更要命的是，Windows XP 发布后暴露出极为严重的安全漏洞。原本管理层以为只需花费三个月发布一个补丁包即可应付，结果全公司最顶尖的工程师被迫花了将近两年时间全力以赴打磨安全补丁，最终才产出了被誉为拯救 XP 口碑的传奇更新——Service Pack 2（SP2）。安全危机消耗掉了本该投入下一代系统的精锐工程兵力，Blackcomb 的发布窗口不得不大幅延后。面对进退维谷的局面，微软决定在 Whistler 和 Blackcomb 之间插入一个温和的过渡小版本。去过那两座雪山滑雪的人都知道，两山之间有一家极具盛名的滑雪酒吧叫做“长角牛沙龙”（Longhorn Saloon）。管理层于是轻松地将这个过渡项目定名为“Longhorn”——没人能料到，这个本意只是小幅迭代的代号，最终会演变成微软历史上最惨烈的开发灾难之一（即后来的 Windows Vista）。大卫不禁回忆起自己高中时混迹各类极客技术博客、狂热追踪 Longhorn 各类泄露界面的日子，当时他甚至特地在自己的 Windows XP 上安装第三方桌面 Shell 美化包，只为了提前体验 Longhorn 那带有半透明侧边栏和大时钟的前卫界面。 ### [02:17:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8235s) Longhorn 的无尽预热与期待落空在 Windows XP 发布之后，微软对代号为“Longhorn”（即后来的 Windows Vista）展开了一场前所未有的宣传造势。与现代科技公司严守发布秘密的风格截然不同，微软通过各大开发者博客和粉丝社区，事无巨细地提前透露 Longhorn 的各项前沿特性——而此时该系统甚至连一个确定的交付日期都没有。这种持续“挤牙膏”式的信息释放，让技术圈与公众对新系统的期待值被拉到极致，所有人都在翘首以盼：“Longhorn 到底什么时候才能发布？” 这种等待最终演变成了一场长达五年的漫长消耗战。节目主播调侃道，这就像苹果如果突然在开发者官网上公开列出 iOS 23 的全套蓝图，然后让用户苦等数年一样荒谬（尽管苹果在公布 Apple Intelligence 时也罕见地采用了“未来一年逐步推出”的策略，但微软当年的做法更为失控）。对于苦等五年的技术爱好者而言，微软画出的宏大饼图始终未能落地，背后其实潜藏着严重的研发灾难与组织失控。 --- ### [02:18:44](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8324s) Longhorn 的“三大支柱”与技术死局 Longhorn 构想的最初雏形源自一个更为庞大的操作系统代号“Blackcomb”。微软原本计划先削减范围打造 Longhorn，但各种野心勃勃的功能最终又被全数塞了回来，形成了支撑整个系统的“三大技术支柱”： ```mermaid flowchart TD accTitle: Longhorn 的三大支柱与重构危机 accDescr: 展示 Longhorn 的三大架构支柱以及最终无法整合导致代码库重构的过程 subgraph Pillars ["Longhorn 规划的三大支柱"] A["Avalon (图形引擎) DirectX 硬件加速 / Aero 界面"] B["Indigo (通信框架) 面向 Web 服务的网络架构"] C["WinFS (存储革新) 面向对象的文件系统数据库"] end subgraph Crash ["整合崩溃与推倒重来"] D["架构膨胀与.NET 深度绑定 各团队子项目无法编译为统一构建版"] E["推倒重来 (Reset) 基于 Windows Server 代码库重新分支拉包"] end Pillars --> D D --> E ``` - **第一支柱：Avalon（后来的 WPF）**。这是一个借助 GPU 硬件加速的全新图形引擎，微软试图将 DirectX 直接融进操作系统底层，从而带来华丽流畅的视觉动效。然而现实却十分尴尬：当时整个 PC 供应链正掀起“上网本（Netbook）”低配浪潮，硬件厂商们迫切需要新系统来提振 PC 销量，却发现这套全新的“Aero”玻璃特效界面极其吃配置，要求电脑配备近乎游戏级别的独立显卡。相比软硬件一体的苹果 macOS Aqua 界面，微软在依赖 OEM 合作伙伴的生态下搞硬件加速，出现了极其严重的沟通脱节与定位误判。 - **第二支柱：Indigo（后来的 WCF）**。一套试图将整个互联网架构完全缝合进 Windows 的 Web 服务框架，但其概念模糊抽象，在研发体系中更像一场狂热的技术空想。 - **第三支柱：WinFS**。这是比尔·盖茨最为推崇、技术野心最大的一项架构革新。微软企图复活当年在 Cairo 项目中未竟的梦想——打造一个“面向对象的文件系统”。系统不仅存储文件，更能原生识别日历邀请、联系人等各种数据类型。然而这套方案完全脱离了实际需求，第三方应用开发者根本毫无调用意愿；更致命的是，底层团队陷入了“WinFS 规范没敲定，其他上层模块就无法推进”的无限内耗之中。雪上加霜的是，开发团队还执意将.NET 框架直接编入系统底层核心。各个部门各自为战，虽然能在自己本地编译运行子模块，但当需要产出一个完整可运行的操作系统构建版本（Build）时，整个 Longhorn 代码库根本无法完成编译。历经无数次战略闭门会（Offsite）与架构评审后，项目在第四年全面崩盘。微软不得不放弃原有开发分支，退回旧版的 Windows Server 代码库重新拉分支（Re-fork），以拼积木的方式勉强拼凑出一个功能大幅缩水的交付物。 --- ### [02:24:02](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8642s) Windows Vista 发布：一场灾难与企业文化的失败 2006 年，在史蒂夫·鲍尔默接任 CEO 六年后，比尔·盖茨亲自登台发布了代号为 Vista 的新系统，打出了 "The wow starts now"（震撼，由此开始）的营销口号。然而发布会的表现却暴露了系统的苍白：除了一套层层推进的“3D Alt-Tab”窗口切换（Flip 3D）和侧边栏小组件外，根本拿不出任何具有真正说服力的杀手级特性。更糟糕的是，系统引入了备受诟病的用户账户控制（UAC），本意是为了安全防护，实际使用中却以狂轰滥炸般的弹窗弹到用户崩溃，导致用户都在想方设法把它彻底关闭。市场给出的反馈极为惨烈： - **企业市场全面抗拒**：即使到了 Vista 发布三年半后的 2009 年，全球仍有近四分之三的企业 PC 坚守在 Windows XP，坚决拒绝升级到 Vista。 - **OEM 厂商反水**：由于消费者完全不买账，微软甚至不得不破例允许 OEM 厂商在 Vista 发布后的整整两年内，继续销售预装 Windows XP 的电脑。曾供职于微软 Office 部门的主播指出，这场灾难本质上是当年 Windows 团队文化的终极失败。与 Office 部门严谨的工程管理体系（提前三年定死交付日期、严格的 Dev/Test/PM 三人组架构、日常故障分流及零 Bug 反弹原则）相比，Windows 团队往往以“我们在搞操作系统级的底层炼金术，常规软件工程原则不适用”为借口，陷入自负的精英主义泥潭。整个团队“沉迷于吸食自己的迷幻药（Smoking their own supply）”，深信自己比所有人都聪明，将用户的真实诉求抛之脑后，耗费数年在一堆凭空捏造的复杂技术里相互扯皮，最终换来了一场被市场全盘唾弃的惨败。 ### [02:27:07](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8827s) Windows Vista 重启与人才流失：Brian Valentine 转投亚马逊 Windows Vista 的最终发布在流程上经历了一次彻头彻尾的重构与重启（reset）。为了打破开发僵局并设法将产品推向市场，微软调来了负责 Exchange 和 Windows Server 企业级业务的 Brian Valentine 主持大局。他的任务非常明确：大刀阔斧地砍掉功能，削减掉原本规划中被称为“三大支柱”的 Longhorn 宏大愿景。即便如此，在全面缩编之后，整个交付过程依然耗费了整整两年时间。更为惨痛的是，在 Vista 终于面世之后，Brian Valentine 随即离开了微软。大批顶尖工程师也随之纷纷离职，穿过街道跳槽去了亚马逊（Amazon）。Brian Valentine 随后出任了 amazon.com 整个工程平台团队的负责人，并在极高层的执行管理团队中任职多年。 - Brian Valentine 原本在微软主导战绩卓越的 Exchange 与企业服务器部门，被临时抽调去处理 Vista 烂摊子。 - 经过无情削减 Longhorn 愿景中的核心支柱功能后，交付仍拖延了两年才得以完成。 - Vista 发布后，包括 Brian Valentine 在内的大量关键系统级工程人才直接流向亚马逊。 --- ### [02:28:01](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8881s) 盖茨角色转变的技术黑洞反思在反思反垄断案给处于消费级统治巅峰时期的微软带来的深远冲击时，播客主持提出了一种个人推论：比尔·盖茨不再担任 CEO、转任“首席软件架构师”（Chief Software Architect），对这一困境起到了关键的负面作用。CEO 的职责要求其必须与 OEM 硬件伙伴、企业客户和普通用户深度接触并保持反馈循环；而当盖茨仅将全部精力沉浸在技术决策中时，微软内部陷入了一种“沉迷于自我供给（getting high on your own supply）”的黑洞状态。这种技术视角的脱节导致微软缺少一个“调节阀（governor）”，没人去踩刹车制止为了重构架构而重构 Windows 的盲目冲动。比尔·盖茨不仅是顶级工程师，更是受过极致商业训练的谈判大师与法务天才。然而，盖茨缺乏像史蒂夫·鲍尔默那样深耕企业客户成功、建立稳固企业伙伴关系的热情。让盖茨退居单一的技术架构角色，反而放大了技术架构的失控风险。 ```mermaid flowchart TD accTitle: 比尔·盖茨角色转变带来的架构脱节 accDescr: 展示盖茨从全功能CEO转变为首席软件架构师后，失去外部反馈机制导致技术过度重构的过程。 subgraph CEO_Role["全功能 CEO 时期（平衡态）"] A["CEO 职责"] --> B["外部反馈机制 (OEM、企业伙伴、终端客户)"] A --> C["商业博弈与法务把控"] B & C --> D["约束技术野心，保证交付"] end subgraph CSA_Role["首席软件架构师时期（失控态）"] E["仅专注技术决策"] --> F["缺乏外部反馈阀门 (失去现实业务制约)"] F --> G["过度追求架构重写 (Longhorn 黑洞)"] end ``` --- ### [02:29:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=8986s) 苹果“Get a Mac”广告与“It Just Works”的品牌建立 2006 年末，在 Vista 步履蹒跚推向市场的同时，苹果公司推出了极具杀伤力的“Mac versus PC”（Get a Mac）系列电视广告。彼时苹果 Mac 在个人电脑市场的份额微不足道（仅占 2% 到 3% 左右，即便到 2024 年也仅有约 8%），但这些广告直击要害，精准刺痛了消费者、企业界以及微软内部最脆弱的神经——因为广告中所描绘的 PC 崩溃、臃肿等问题，全都是不争的事实。 ``` “嗨，我是 Mac。” “我是 PC……我又死机了。” ``` 这一时期，苹果内部正处于全速运转的巅峰状态：史蒂夫·乔布斯重掌大权，iPod 风靡全球，“数字枢纽（Digital Hub）”战略大获成功，带动 Mac 销量快速增长，甚至成为大学生选购电脑的首选。Vista 的疲软成了苹果绝佳的突破口。苹果借此向消费者灌输并确立了“开箱即用、无缝顺畅（It Just Works）”的品牌承诺。Mac 能做到，iPod 能做到，这为随后 2007 年 iPhone 的诞生铺平了道路，大众对苹果打造一款高品质手机的极度期待在此刻达到了沸点。 --- ### [02:32:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9134s) 微软消费级霸权的终结与开发者的流失 Vista 的灾难标志着微软作为一家具有引领性的“消费级科技巨头”时代的终结，其地位从此再未恢复。不仅消费市场流失殆尽，更致命的是，微软失去了作为“开发者创新平台”的号召力与凝聚力： 1. **Win32 API 失去魅力**：几乎没有开发者愿意专门为 Vista 开发原生应用；Win32 API 不再是全新创新应用的首选目标，业界主流迅速转向了 Web 网页应用。 2. **唯一坚守的阵营**：还在为 Windows 编写桌面软件的，只剩下极少数为企业编写定制化内部工具的企业级开发者，或者像 Adobe 这种必须维持跨平台历史大作的传统软件厂商。新一代颠覆性科技公司（例如 Facebook）绝不会开发专用的 Windows Vista 客户端。 3. **顶层技术战力被全面吸干**：Vista 如同一个不断膨胀的黑洞，不仅吸纳了大批普通研发队伍，还从全公司抽调了大量执行副总裁和资深特级工程师（Distinguished Engineers）前去灭火，导致微软无暇兼顾移动端（Mobile）、社交网络（Social）、网页搜索（Search）以及正在面临转折点的浏览器大战（IE 市场份额即将跳水）。 --- ### [02:34:56](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9296s) 鲍尔默的悔恨与核心技术人才的流向在与播客主持人交流时，史蒂夫·鲍尔默曾坦言，尽管人们常批评他的收购决策（他自嘲那些糟糕的收购“仅仅是赔了钱而已”，金钱对当时的微软来说并不是稀缺资源），但 Vista 才是他在任 CEO 期间最糟糕的时刻。因为 Vista 让微软最优秀的一批工程师全部“脱离赛场、失去战斗力”，这种人才被深陷泥潭所消耗的机会成本才是致命的。一位微软前高管曾深感痛心地总结道，最令人沮丧的事实是：微软大批最顶尖的系统级底层工程专家选择辞职，驾车穿过华盛顿湖前往西雅图市区的亚马逊，先是帮一家“在线书商”搭建系统，并最终将其打造成全球领先的企业云计算巨头（AWS）。这种戏剧性的人才外流，成为了微软战略史上最沉重的一记耳光，但也为后续微软痛定思痛、走向 Azure 的云端救赎埋下了伏线。 - 资金并非稀缺资源，并购失败亏损无伤大雅；将顶尖核心人才长期套牢在失败架构上才是真正的“杀手级”失误。 - 原本在企业级大杀四方的技术骨干转战西雅图，反向助力竞争对手构建出主导市场的云计算平台。 - 主持人预告随后的发展脉络将贯穿搜索、移动系统、Windows 8、Zune 等产品线，并最终汇聚向 Azure 这一核心支柱。 --- ### [02:36:33](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9393s) 赞助商赞助：Pilot 财务与财税服务在节目进入 Azure 与云战略的全面讨论之前，主持人播插了长期合作伙伴 Pilot.com 的赞助介绍： - **业务定位**：Pilot 专注于为各类初创企业与成长型公司提供全套记账（bookkeeping）、财务会计与税务服务，是目前全美规模最大的面向初创企业的会计服务机构。 - **特别致敬**：主持人特别提到了 Pilot 联合创始人兼 CEO Wasim Daher，称其是节目赞助商中独一无二的存在——既是节目的赞助商 CEO，同时也是本期微软专题深度调研的重要事实来源与行业受访者之一。 ### [02:37:05](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9425s) 赞助商插播与初创企业的专注哲学本段内容以一段幕后花絮开场：Wasim 在麻省理工学院（MIT）念书时曾为校报采访过比尔·盖茨（Bill Gates）。他把当年的采访稿找出来发给主播们，盖茨在访谈中提到了微软即将推出的 40GB 便携式媒体播放器（Portable Media Center），这成了一份极具时代感的“时间胶囊”，相关 PDF 资料也将放在节目的参考来源中。随后话题转到赞助商 Pilot。主播提到比尔·盖茨的西雅图邻居杰夫·贝佐斯（Jeff Bezos）以及源自 AWS 的经典准则——初创公司应该将精力放在“让啤酒更好喝”（what makes your beer taste better）的事情上。换句话说，初创企业有限的时间和资源只应投在能为公司、产品和客户带来质变的核心领域，其余非核心业务一律外包。财务与记账就是最典型的例子：每家公司都必不可少，必须由专业人员把关以规避风险，但做好它本身并不会直接提升产品的核心竞争力。Pilot 为企业提供全套财务解决方案，从总账、记账、税务，到高阶的 CFO 职能（如投资者报告、预算编制、董事会汇报材料中的财务板块），服务对象包括 OpenAI、Airtable、Scale 等知名公司及大型电商企业，能够陪伴公司从初创期一直扩展到高成长期。 --- ### [02:38:48](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9528s) 浏览器与搜索：微软错失的互联网流量入口接下来进入本章的核心主题——搜索业务。这一章也可以被冠以另一个副标题：“一桩未竟的收购”。在科技商业史上，有一段险些以 470 亿美元达成的天价交易如今已少有人提及，那就是微软在 2008 年对雅虎（Yahoo!）的世纪收购案。要理解这笔未竟的收购，需要先回顾当时在线广告与搜索领域的背景格局： - **程序化广告技术的竞争**：当时能提供搜索和在线广告定向投放核心技术的有两家先驱，一家是市场领头羊 DoubleClick，另一家是 aQuantive。微软在 DoubleClick 的竞标中输给了 Google，随后仓促以约 60 亿美元（最终近 70 亿美元）收购了 aQuantive。但这次收购以巨大灾难告终，微软最终几乎全额计提了商誉减值。 - **搜索自研的停滞**：微软极其渴望获得搜索市场份额，但其内部推动的 MSN Search 以及随后的 Windows Live Search 进展缓慢，始终难以打开局面。 - **浏览器的全面失守**：在十年前打赢第一次浏览器大战后，微软在浏览器领域彻底松懈，Internet Explorer（IE）严重退化。网页开发者怨声载道，因为 IE 充斥着非标准的定制规范，经常莫名其妙出现兼容 bug；普通用户也极为厌倦，操作系统每一次升级，IE 看起来都毫无新意，连标签页（tabs）功能都不支持。 - **Firefox 的崛起与 Google 的前端大门**：大约在 2007 年前后，Firefox 异军突起，凭借创新的多标签页浏览体验迅速抢占市场，而 Google 则是 Firefox 的默认搜索引擎。这一现象揭示了浏览器与搜索之间密不可分的共生关系：浏览器是通往搜索的“正门”，而搜索则是浏览器最核心的变现机器。Google 凭借竞价广告模式可以直接将搜索变现；而如果一家公司像 Firefox 或苹果（Apple）那样拥有前端用户流量，即便自身不运营复杂的搜索竞价引擎，也可以通过向 Google 导流获取高昂的“流量获取成本”（TAC）。时至 2024 年，业内传闻苹果每年仅靠将 Safari 的默认搜索流量导向 Google，就能进账高达 200 亿美元。 ```mermaid flowchart LR accTitle: 浏览器与搜索变现逻辑 accDescr: 展示浏览器作为前端入口向搜索引擎输送流量并获得商业回报的流程 subgraph FrontEnd["前端流量入口"] Browser["浏览器 (如 Safari / Firefox)"] end subgraph Monetization["搜索引擎与商业变现"] SearchEngine["搜索引擎 (如 Google)"] AdMarket["关键词竞价广告系统"] end Browser -- "默认搜索流量" --> SearchEngine SearchEngine -- "产生巨额广告收入" --> AdMarket SearchEngine -- "分成 / 流量获取成本 (TAC)" --> Browser ``` --- ### [02:42:18](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9738s) 搜索的飞轮效应与微软的后知后觉微软早年在搜索上的迟钝，根源在于从上到下没有真正意识到搜索商业模式的恐怖威力。1998 年 Google 成立时，外界普遍怀疑竞价排名的广告模式能否跑通；随后又怀疑它能否具备规模效应；直到 2004 年 Google 上市，其惊人的利润率才震惊了整个华尔街。更关键的是，许多人很晚才领悟到：**搜索是一个赢家通吃、规模效应极强的超级双边市场**。搜索业务的飞轮机制体现在： 1. **数据飞轮**：搜索查询量最大的引擎能积累最多的行为数据，从而训练并返回最精准的搜索结果。 2. **变现与竞价流动性**：规模最大的引擎吸引了最多的广告主进场竞价，推高关键词价格，形成极高的单位查询收入。 3. **资本再投资壁垒**：赚取最丰厚利润的领头羊能够投入远超对手的资本开支（CapEx）建设数据中心，并在研发（R&D）上持续下注，让搜索速度更快、体验更好，进一步拉大与第二名的差距。 Google 这种结合了搜索技术与竞价拍卖的广告模式，被播客主播称作“资本主义最伟大的商业发明之一”。而在微软内部，搜索在很长时间里仅仅被视作 MSN 门户的一个附属功能，或是 Windows Live 软件套件里的一环，管理层总觉得靠现有班底修修补补就能做出来。彼时的微软急需像当年 J Allard 写下《Windows: 互联网上的杀手级应用》（The Next Killer Application on the Internet）或 Steven Sinofsky 写给盖茨的备忘录那样的当头棒喝，向高层敲响警钟：未来五年内没有任何事情比拿下搜索收入更重要。 --- ### [02:44:49](https://www.youtube.com/watch?v=6oER1bdQa_s&t=9889s) 470 亿美元的雅虎豪赌与机场密谈直到 2008 年，面对 Google 势不可挡的扩张，微软终于坐不住了。为了在搜索领域获得哪怕一线生机，微软向雅虎发起了全面收购要约，报价一路攀升到了令人咋舌的 470 亿美元。这堪称微软试图在搜索领域挽回局面的“最后一记万福玛利亚长传”（Hail Mary pass）。要知道，Google 早在 1998 年创立、2004 年上市，而微软直到 2009 年才正式推出自主搜索品牌 Bing。在经历了漫长且火药味十足的拉锯战与多次往返差旅后，雅虎联合创始人杨致远（Jerry Yang）和大卫·费罗（David Filo）飞抵西雅图。史蒂夫·鲍尔默（Steve Ballmer）亲自赶往波音机场（Boeing Field），双方在机场展开了极具戏剧性的闭门谈判。关于那场会谈的具体细节存在不同版本的传闻，但核心结果是明确的：比尔·盖茨和鲍尔默看着雅虎创始人的眼睛，敏锐地意识到雅虎团队根本没有被收购的诚意，只是在虚与委蛇、拖延拉扯。如果继续硬吞，不仅代价过于昂贵，甚至若被雅虎继续抬价，合并后也必将引发剧烈的企业文化“器官排斥”（organ rejection）。微软最终决定放弃收购，这笔 470 亿美元的世纪并购彻底告吹。但主播随即抛出了一个深具历史回溯感的反转问题：即便以当年看似疯狂的 470 亿美元天价来看，买下雅虎对微软来说真的算一笔亏本买卖吗？两位主播相视一笑，心照不宣地暗示了这笔交易背后那个与“阿里巴巴（Alibaba）”密切相关的隐藏宝藏。 ### [02:46:53](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10013s) 拆解雅虎收购案：阿里巴巴与雅虎日本的隐形财富主播们进一步深入剖析微软当年开出 470 亿美元收购雅虎（Yahoo!）的这笔交易。表面上看，花 470 亿美元买下雅虎在搜索市场上约 15% 的份额似乎代价过高（当时谷歌遥遥领先排第一，雅虎位列第二）。但如果仔细拆解雅虎当时的资产，会发现这笔交易里其实潜藏着两个极为庞大的宝藏：雅虎日本（Yahoo Japan），以及雅虎持有的 40% 阿里巴巴（Alibaba）股份。阿里巴巴的这笔投资堪称商业史上最成功的风险投资之一。仅这两项资产在当时合计就价值超过 300 亿美元。如果把这 300 亿美元从 470 亿美元中扣除，微软实际上仅需花费约 150 亿美元，就能买下整个搜索市场 15% 的份额。对比 2023 年 Alphabet（谷歌母公司）超过 3000 亿美元的年收入，每 1% 的搜索市场份额相当于每年几十亿美元的潜在产出；花费区区 150 亿美元去换取 15% 的搜索市场，从数学模型来看划算到了极点。人们总是习惯性低估搜索市场的巨大体量与惊人盈利能力。尽管后来的事实是雅虎走向没落，最终以仅仅 50 亿美元被威瑞森（Verizon）收购，如今由阿波罗全球管理（Apollo Global Management）等机构持有，人们不禁假设：若微软当时买下雅虎的流量，真能打造出抗衡谷歌的引擎，还是会让其随雅虎一起凋零？但即便不谈业务协同，单看阿里巴巴的股权，这笔交易本也可让微软血赚：2014 年阿里巴巴在美上市时，这 40% 的股权价值高达 920 亿美元。虽然微软不是对冲基金，未必能拿稳至上市，但错失这笔历史级投资依旧令人唏嘘。 - 雅虎当时占据搜索市场约 15% 的份额，稳居行业第二。 - 雅虎持有的 40% 阿里巴巴股份在 2014 年 IPO 时价值高达 920 亿美元。 - 扣除雅虎日本与阿里股权价值后，微软拿下雅虎核心业务与搜索份额的隐形成本仅约 150 亿美元。 --- ### [02:48:52](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10132s) 必应的诞生、雅虎搜索合作与萨提亚·纳德拉的早年履历微软在收购未果后于 2009 年正式推出了必应（Bing）。为了将搜索业务认真做大，微软广纳顶尖技术人才，成功打造出一款极具可行性的搜索引擎，年利润可达约 10 亿美元。与此同时，微软与雅虎达成商业合作：由必应为雅虎后台提供全套搜索支持。雅虎保留自身流量的商业化变现权益，但微软获得了执行所有搜索查询背后的海量数据资产。主持人们在此插播了一段鲜为人知的高管履历轶事：在必应早期，掌舵该业务的正是萨提亚·纳德拉（Satya Nadella）。纳德拉于 1992 年从 Sun Microsystems 加入微软，最初担任 Windows NT 的技术布道师（evangelist）。随后，他在微软接手了第一个产品岗位——负责从未正式发售的“老虎服务器”（Tiger Server，当年微软针对“信息高速公路”与有线电视机顶盒设想的狂热产物）。在微软，纳德拉的职业路径极为丰富且不走寻常路：他曾在 Dynamics（微软收购并对标 Salesforce 的 CRM 部门）工作，随后负责企业级中间件 BizTalk Server，接着接手 Bing 搜索引擎，再之后被提拔掌管“服务器与工具部门”（Server and Tools Business），最终孵化并带领 Azure。在微软的传统体系中，产品经理（Product Manager）往往隶属于营销部门而非纯工程团队，而纳德拉正是在这一跨越前沿探索、企业服务与大规模互联网服务的历练中崭露头角。 - 纳德拉 1992 年加入微软，首份工作是 Windows NT 技术布道师。 - 纳德拉负责的首个产品研发项目是针对交互式电视与信息高速公路的“Tiger Server”。 - 在接管 Azure 之前，纳德拉曾先后历练于 Dynamics CRM、BizTalk Server 以及必应搜索部门。 --- ### [02:50:55](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10255s) 云计算的隐形基因：必应、Xbox Live 与 Hotmail 的分布式淬炼在微软与雅虎的合作架构中，雅虎负责前端流量变现并在前五年拿走 88% 的收入，但微软承担了全套后端广告销售和搜索匹配。这种安排让微软不仅通过必应汲取了海量查询数据来训练算法，还成功搭建起一套完整的广告交易流动性市场（如今该广告平台已扩展到包括 Netflix 带广告订阅层级在内的众多外部合作伙伴）。更关键的影响在于底层架构的蜕变。一旦接入雅虎的庞大流量，必应就必须成为一个必须保持 24/7 全天候运行、具备极低延迟、极快响应能力的大规模分布式计算系统——这恰恰是整个计算机领域最硬核的技术挑战。而这套能力，在本质上完全就是“云计算（Cloud Computing）”。微软当时已经悄然积累了发展云服务所需的三大关键分布式基石： 1. **Xbox Live**：服务数千万实时在线联机游戏用户的平台； 2. **必应（Bing）**：全球顶级规模的高并发分布式搜索网络； 3. **Hotmail**（后演变为 Outlook.com）：自 1997 年收购以来持续大规模稳定运营数十年的消费级 Web 邮件服务。这三大业务为微软锤炼出了一批擅长大流量、高可用 Web 服务的核心工程人才与底层架构 DNA，为日后 Azure 云计算的爆发埋下了至关重要的技术伏笔。 ```mermaid flowchart TD accTitle: 微软云计算技术基因的演进 accDescr: 展示必应、Xbox Live和Hotmail如何共同为微软Azure云计算平台奠定高并发分布式系统基石。 subgraph Infrastructure["大规模分布式系统能力沉淀"] A["Hotmail (1997收购) 数十年消费级Web服务运维"] B["Xbox Live 4000万实时在线联机用户"] C["Bing + 雅虎搜索后台 海量查询与低延迟分布式系统"] end subgraph Result["技术演进与战略成果"] D["分布式计算与超大规模基础架构DNA"] E["Azure 云计算服务平台"] F["微软统一广告变现生态"] end A --> D B --> D C --> D D --> E C --> F ``` --- ### [02:53:24](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10404s) 商业范式转移：从“卖软件许可”到“卖整个经济” 为什么微软对搜索引擎如此执念，甚至屡次掀起百亿美元级别的并购攻势？核心在于商业模式的根本性降维打击。作为传统科技巨头，微软的基石逻辑是“软件许可模式”——开发一套 Windows 或 Office，按份直接卖给客户。但谷歌开辟的搜索广告模式展示了一种几乎可无限扩张的新范式：如果世界上有 70 亿人口，微软只能向他们推销固定价格的操作系统许可证；但如果向他们提供免费搜索，再通过商家的广告竞价变现，从每个用户身上榨取的价值可能是软件许可证的 3 到 5 倍。广告驱动的软件变现效率彻底超越了软件买断销售。更深层的一层逻辑在于“市场天花板的差异”： - **软件经济学**：企业或个人的“软件支出预算”在整体财务中只占很小一部分。 - **搜索与广告经济学**：广告不仅服务于软件公司，还涵盖机票、帐篷、零售、汽车等所有实体经济行业。搜索广告的本质不是在“卖软件”，而是在为整个实体消费经济销售获客渠道。这一差异在两家公司的浏览器战略中暴露无遗： - **微软开发 IE 浏览器**：在赢得浏览器大战后占领绝大部分市场份额，却完全不知如何利用它，仅仅将它视作保护 Windows 操作系统许可证销售的“防御型工具”。 - **谷歌开发 Chrome 浏览器**：将浏览器作为吸纳全球互联网流量的绝对入口，借此将所有流量转化为商业变现广告，将其用作争夺全局商业价值的“进攻型利刃”。意识到在搜索广告时代落后且面临被颠覆的危险后，到了 2007 年 10 月，随着社交网络浪潮（如 Facebook）的兴起，微软下定决心：决不能在下一波基于网络流量与广告的生态革命中再次缺席。由于微软缺乏打造原生消费级社交产品的基因（不愿重蹈 Google+ 式的内部强行研发弯路），但坐拥充裕的资本，他们迅速将目光投向了对初创明星 Facebook 的资本结盟与大规模竞逐。 ### [02:57:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10651s) 微软对 Facebook 的巨额报价与 2.4 亿美元战略投资在互联网广告和社交领域的焦虑驱动下，微软曾试图全资收购 Facebook。微软当时开出了一套极为复杂的交易架构，其实质是允许 Facebook 股东随着公司未来价值的增长而逐步套现，而不是在当下一次性把所有筹码拿走。这笔交易在媒体报道中的估值高达惊人的 240 亿美元——要知道这发生在 2007 年，距离 Facebook 创立仅仅过去三年，且不久前雅虎才开出过一份著名的 10 亿美元收购要约。然而马克·扎克伯格（Mark Zuckerberg）对此毫无兴趣，甚至连正式回复都没有给。虽然部分高管曾私下同微软接洽并暗示过估值区间，但最终收购依然告吹。收购未能成行，双方转而达成了一笔投资与商业合作协议。2007 年 10 月，微软向 Facebook 投资 2.4 亿美元，换取其 1.6% 的股权，直接确立了 Facebook 达到 150 亿美元的估值基准。作为商业协议的一部分，微软获得了截至 2011 年在 Facebook 国际市场上独家销售横幅广告（banner ads）的权利。微软此举意在借助 Facebook 庞大的海外流量库存，冷启动（bootstrap）自身的广告主双边市场。从纯财务投资的角度来看，如果微软在 Facebook 2012 年 IPO 时全部抛售，将在 5 年内获得 7 倍的回报；若持有至 2014 年卖出，则约为 14 倍回报。但这笔财务收益对微软而言微不足道，因为微软从不是一家对冲基金，钱也是它最不缺的资源。微软真正面临的瓶颈是人才、执行力、专注力以及能否成事的企业基因。这笔交易真正深远的影响，在于促成了两家公司的蜜月期：Facebook 在商家主页等场景中深度集成了 Bing 地图，双方展开了大量业务互惠，并且大批微软骨干人才跳槽流向早期 Facebook（例如后来的 Statsig 首席运营官 Vijaye Raji）。 ```mermaid flowchart TD accTitle: 微软 2007 年对 Facebook 的投资与商业合作架构 accDescr: 微软出资 2.4 亿美元获取 Facebook 1.6 股权，同时获得四年国际横幅广告独家销售权，推动双方技术合作与人才流动。 subgraph Deal ["2007年10月交易核心"] MS["微软 (Microsoft)"] FB["Facebook (估值150亿美元)"] MS -- "注资 2.4 亿美元 (持股 1.6%)" --> FB end subgraph Value ["战略与商业协同"] FB -- "独家海外 Banner 广告销售权 (至2011年)" --> MS MS -. "提供广告主网络与变现冷启动".-> FB FB -. "采用 Bing Maps 等微软技术生态".-> MS MS -. "早期核心技术人才输送".-> FB end ``` ### [03:00:35](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10835s) 比尔·盖茨全职退休与《时代》周刊的时代注脚 2008 年 6 月，比尔·盖茨正式从微软全职岗位退休。他卸任了首席软件架构师一职，仅保留董事会主席身份，将精力完全转移至比尔与梅琳达·盖茨基金会（Bill & Melinda Gates Foundation）。此时恰逢移动技术大转折的前夜：苹果刚刚在 2007 年发布了初代 iPhone，而 2008 年 iOS 2.0 携 App Store 与开放 SDK 破空出世，彻底颠覆了整个软件生态。在比尔·盖茨退休之际，《时代》（Time）杂志刊登了一段极其刻薄的评价：“盖茨在合适的时间离开了科技界。说来有趣，这已经不再是属于书呆子（nerds）的行业了。盖茨曾处在个人电脑革命与互联网革命的绝对中心，但现在的大创新恰恰发生在他所不擅长的领域：iPod 是一场审美革命，MySpace 是一场社交革命，YouTube 是一场娱乐革命——这都不是盖茨擅长做的事，科技界不再需要他了。” 节目主播指出，尽管这种论调过于片面，且将问题过分归咎于盖茨个人，但它极其精准地反映了当时舆论界对微软的普遍看法——在经历了 Windows 95 的绝对辉煌后，彼时的微软已被大众打上了“落伍与无关紧要”的标签。 - 微软被外界轻视的根源，在于其近乎偏执地聚焦企业级市场，同时在消费级市场屡遭惨败。 - 当时的开发者焦点已经彻底转向开源世界和 Web 平台（如 Linux、Apache、MySQL、PHP 构成的 LAMP 技术栈，也是构建 Facebook 的底层），这与微软服务的大型企业开发者生态完全处于两个不同的平行宇宙。 - 缺乏开发者的狂热追随，就注定无法催生消费级应用的活力；平台若想繁荣，不能仅靠自己去开发所有软件。 ### [03:02:55](https://www.youtube.com/watch?v=6oER1bdQa_s&t=10975s) 消费化浪潮冲击下的“Windows 中心主义” 随着企业级霸业的建立，微软建立起由锁定效应构筑的高墙，但 iPhone 的降临彻底粉碎了这一计算公式。iPhone 掀起了“影子 IT”（Shadow IT）和“自带设备办公”（BYOD, Bring Your Own Device）的浪潮，引发了一场普通职场人反抗企业 IT 部门专权的技术革命。用户开始追求更优质的软件交互与硬件体验，倒逼企业接纳消费级设备进入内网。而在这一时期，微软内部原有的消费级产品嗅觉与设计基因早已荡然无存。即便偶尔有像 Courier 双屏平板电脑或 Kin 社交手机这样极具创新潜力的试验性项目冒头，它们最终也难逃被内部扼杀的命运。微软内部根深蒂固的教条是：Windows 必须是宇宙的中心，任何新产品如果不能让 Windows 看起来更耀眼，或者可能会威胁、分流 Windows 的地位，就会被直接叫停。这种企业基因的桎梏，彻底堵死了微软通过颠覆式创新实现自我重塑的路径。 ### [03:04:39](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11079s) 移动计算的先行者为何沦为“零件供应商” 尽管在大众叙事中微软错失了移动互联网，但事实上微软涉足移动领域极早，且投入巨大。比尔·盖茨对移动计算的痴迷远超时代，早在 2001 年 Windows XP 的发布演示幻灯片中，移动计算（mobile computing）就被列为核心特性之一；盖茨多年来持续押注多点触控、自然用户界面、平板电脑（Tablet PC）和手写笔计算。早在现代智能手机普及前，微软便推出了 Windows Mobile 操作系统。然而，Windows Mobile 的本质与后来的 iOS 截然不同： - **产品形态与定位**：Windows Mobile 最初搭载于类似全键盘黑莓（BlackBerry）形态的设备上，屏幕较小，界面生硬地照搬桌面端体验（甚至带有一个迷你的“开始”菜单）。它本质上依然是企业级战略的附庸，核心目标是与桌面 Windows PC、Exchange 邮件服务器和企业内网无缝同步，完全是为企业雇员配发使用的商务机，并未考虑普通家庭消费者。 - **产业分工的角色偏差**：微软在生态中将自身摆在手机厂商“原料供应商”（ingredient）的被动位置。在传统 PC 阵营中，虽然戴尔等 OEM 厂商会预装软件，但安装的核心依然是标准化的 Windows XP；而在 Windows Mobile 阵营，手机硬件厂商获得了修改操作系统底层代码的权限。 - **体验碎片化与控制权缺失**：硬件厂商的核心优势在于设计外形、配合电信运营商搞定基站频段互联，它们只把 Windows Mobile 当作补齐邮件和计算功能的底座。厂商拿到微软的代码后会自行魔改，导致最终体验完全取决于各大手机厂商良莠不齐的研发水准，微软既没有掌控端到端的软硬件体验，也未能建立起对生态的绝对控制。 ### [03:07:59](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11279s) 智能手机革命前夕与收购 Danger 微软此前虽然拥有 Windows Mobile，但那时的产品与后来的现代智能手机完全是两码事。iPhone 的横空出世彻底重置了行业的一切规则。在 iPhone 问世之后、Windows Mobile 仍处于尴尬过渡期的那段旅程中，微软收购了 Danger 公司。Danger 正是风靡一时的 T-Mobile Sidekick 的缔造者——用户只需按一下按钮，屏幕就会旋转滑开，露出一整块横向全键盘。这台标志性设备不仅在当时酷炫无比，还频繁出现在热门美剧《明星伙伴》（*Entourage*）中，主角 Turtle 用的就是它。这桩收购案背后还有一个极具戏剧性的巧合：Danger 的联合创始人正是安迪·鲁宾（Andy Rubin）。不过在微软收购 Danger 之前，鲁宾早已离职并创办了 Android，而正是 Android 在未来彻底击碎了微软的移动业务梦。 - **T-Mobile Sidekick 的时代符号**：凭借独特的滑盖旋转键盘设计成为千禧年代中期的流行文化标志。 - **安迪·鲁宾的轨迹**：联合创立 Danger $\rightarrow$ 离职并创立 Android $\rightarrow$ 被 Google 收购 $\rightarrow$ 最终成为扼杀 Windows 移动生态的终极对手。 --- ### [03:09:01](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11341s) iPhone 震撼登场与鲍尔默的“500 美元”论断 2007 年 1 月，史蒂夫·乔布斯正式发布第一代 iPhone（同年 7 月发售）。这场演讲被誉为自道格·恩格尔巴特（Doug Engelbart）的“所有演示之母”（The Mother of All Demos）以来最壮丽的技术演示。消费者为之惊叹，但既有的移动通信巨头们却难以置信：黑莓（BlackBerry）创始人最初惊呼“他们是怎么做到的？”，随后又自我安慰说黑莓会安然无恙；Palm 首席执行官此前更是放话，声称“做 PC 的家伙不可能轻易摸清门道直接跨界杀进来”。而史蒂夫·鲍尔默在发布会后的回应尤为出名——他公开断言：“在 500 美元这个价位上，它绝不可能成功。”这一言论常被后人当作缺乏远见的笑柄，但如果回到当时的历史语境，其背后其实有两个深层逻辑： 1. **首席销售员的本能**：作为微软 CEO，面对竞争对手扔下的重磅炸弹，鲍尔默在面对无数企业级客户时绝不可能长他人志气。他的职责是永远保持推销状态（“Always selling”），向外界传递微软阵营依然强大、竞品极其昂贵的信号。 2. **硬件范式与商业模式的颠覆**：当时的移动行业本质上是“嵌入式系统”逻辑，一切围绕压低成本运转，主流旗舰机在运营商合约补贴后仅售 100 美元左右。而初代 iPhone 是一台完整 Mac 电脑的微缩版，完全不惜硬件代价以确保极致体验。虽然初代 iPhone 尚未完善合约机模式，但苹果后来通过 iPhone 3G 等机型开创了高额运营商补贴模式，成功在商业上闭环。 --- ### [03:11:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11469s) 从 Windows Mobile 到 Windows Phone：水土不服的新范式随着 2008 年 App Store 推出以及 2009 年 iPhone 销量迎来爆发，微软终于意识到必须全力求变。他们决定利用原有的 Windows Mobile 资产进行爆改，推出全新的 Windows Phone。然而，这彻底颠覆了微软数十年来维系硬件生态的根本方式： - **旧时代（类 Roku 模式）**：微软向 HTC、摩托罗拉等代工厂承诺，无论厂商制造多么廉价、低端的硬件，微软的底层软件都能兼容运转，大家在几美分的极薄利润上抠成本。 - **新时代（Windows Phone 模式）**：为了正面对抗 iPhone，微软必须提供高品质的用户体验。微软将自身品牌置于首位，制定了极其严苛的硬件规格要求，并从便携播放器 Zune 身上移植了开创性的 Metro UI 动态磁贴界面。但在商业模式上，微软依然顽固地坚守 PC 时代的授权制：向 OEM 厂商按台收取操作系统授权许可费（哪怕每台只收 5 美元），让硬件厂商自行承担研发制造成本去市场上搏取利润。 --- ### [03:13:08](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11588s) Android 的“零元杀招”与降维打击微软试图延续授权费模式的最大盲区在于：战场上的对手已经不再是旧势力，而是 Google 与其开源的 Android。Google 将 Android 从最初模仿黑莓的全键盘形态迅速转向全触屏的“iPhone 复刻版”，并带着三记重拳直接切入微软原有的 OEM 客户群： 1. **完全免费**：系统授权费直接为 0 美元。 2. **开源代码**：向所有 OEM 厂商开放底层源代码供其自由定制。 3. **分化挑拨**：Google 径直对 HTC、摩托罗拉等硬件厂商指出：“看看微软当年对 PC 厂商做了什么！整个 PC 价值链的所有利润都被操作系统厂商吸走，硬件厂商沦为打工仔、利润为零。你们怎么能让微软在手机上重演一次？” 在硬件物料清单（BOM）每一美分都关乎生死的高风险、低毛利制造行业中，微软哪怕只收取 5 美元的授权费，与 Google 的 0 美元相比也是一道无法逾越的天堑。Google 的交换条件仅仅是预装其全套云服务，且随着 Google Play 应用商店的成熟，硬件厂商只要想要庞大的第三方 App 生态，就必须全盘接受 Google 的服务捆绑。 --- ### [03:15:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11710s) 商业模式的代差：微软真正输给的是搜索广告深入剖析这场商战的核心本质会发现：微软在移动领域的真正对手从来都不是苹果的 iPhone，而是 Google 的 Android；而 Android 的根基，其实早在 2002 年 Google 确立搜索广告变现模型时就已埋下伏笔。这是一条跨越十年的商业模式因果链： ```mermaid flowchart LR accTitle: Google 从搜索广告到推出免费 Android 的商业飞轮 accDescr: 展示 Google 如何利用搜索广告利润支撑 Android 免费，最终通过垄断流量入口扼杀微软授权模式。 subgraph Monetization ["核心变现引擎"] A["高利润搜索广告 (2002年确立模式)"] end subgraph MobileStrategy ["移动战略布局"] B["争夺下一代平台 搜索入口"] C["收购并开源 Android (系统授权费 = $0)"] end subgraph Outcome ["竞争格局终局"] D["锁定全部移动搜索流量"] E["降维打击微软 传统 OS 授权收费模式"] end A --> B B --> C C --> D C -.-> E ``` - **第一步**：Google 的核心收入来自搜索广告，利润极其丰厚且单用户价值极高。 - **第二步**：为了让利润最大化，Google 必须垄断所有搜索请求的“第一道门面”（入口）。 - **第三步**：面对从 PC 向移动互联网的范式转移，Google 必须掌控下一代计算平台，确保所有移动端搜索流量无条件导向自己。 - **第四步**：只要能保住搜索广告这一超级现金牛，操作系统本身完全可以白送。这就导致了一场不对称战争：除非微软也能全面转向 Google 的商业模式（完全放弃 OS 授权费并依赖广告变现），否则根本毫无胜算。致命转折点随后到来：在 AT&T 独家拿下 iPhone 销售权后，陷入恐慌的运营商巨头 Verizon 急需反制王牌，最终选择全力押注摩托罗拉 Droid，并砸下天文数字级的营销预算。这一举措瞬间将市场格局收窄为“iOS 对决 Android”的双雄争霸，彻底关上了微软移动端翻盘的大门。尽管微软当时已经拥有搜索引擎 Bing，具备做广告变现的潜在土壤，但作为一家流淌着 PC 软件基因的企业，微软当时的组织文化根本无法接受把核心营收赌在“做一家广告公司”之上。历史的惯性最终让微软错失了移动互联网的整张船票。 ### [03:18:01](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11881s) 谷歌的降维打击：用免费工具侵蚀微软的核心利润池微软的立身之本一直是向 PC 用户与企业销售软件。无论这些销售是通过大型企业许可协议（Enterprise Agreements）打包还是零售，其本质始终是一家不折不扣的商业软件公司。然而，面对微软庞大的利润帝国，谷歌采取了风险投资人比尔·格利（Bill Gurley）所描述的“比免费还便宜”的颠覆性商业模式（Less Than Free）：谷歌逐一审视微软各项高额收费的核心价值产品，计算出哪些产品的单用户运营成本足够低廉，随后直接将其包装为免费服务推向市场。微软有 Outlook 和 Exchange，谷歌就推出免费的 Gmail；微软有昂贵的 Word、Excel 和 PowerPoint 办公套件，谷歌就针锋相对地拿出免费的 Google Docs 及 G Suite（即后来的 Google Workspace）。 ```mermaid flowchart TD accTitle: Google's Disruption Strategy against Microsoft accDescr: 谷歌通过免费云办公软件侵蚀微软软件许可利润并喂养自身广告引擎的飞轮图 subgraph MSFT["微软传统模式"] M1["核心收费软件 (Office / Exchange)"] --> M2["企业授权协议收费"] M2 --> M3["高额软件利润池"] end subgraph GOOG["谷歌颠覆飞轮"] G1["低边际成本替代品 (Gmail / Docs / Sheets)"] --> G2["向大众完全免费开放"] G2 --> G3["掠夺用户关系与数据"] G3 --> G4["喂养搜索与平台广告主业"] end G2 -.->|"降维打击 / 侵蚀边际定价权"| M1 ``` 从商业逻辑上看，这种策略对谷歌来说稳赚不赔。一方面，免费服务能为谷歌带来源源不断的用户数据和更紧密的连接，用户在生态内发起更多搜索查询，即便在站外，谷歌也能凭借沉淀的数据进行更精准的广告变现；另一方面，即便这些生产力工具本身不直接盈利，由于搜索核心业务的利润极其丰厚，只要这一举措能够沉重打击或牵制微软，对谷歌而言就是巨大的战略胜利。不过，这种进攻在消费级市场和企业级市场呈现出截然不同的结局：在个人消费端，Google Docs 和 Sheets 确实迅速普及；但在企业级市场，微软的生态锁定（lock-in）依然牢不可破。即便全球有大量 Google Sheets 的拥趸，但在全球企业办公的真实世界里，绝大多数严肃的表格处理依然全盘被 Excel 垄断。 - 谷歌的核心策略是寻找微软收费最高但底层单用户服务成本最低的软件品类进行免费化。 - 即使免费软件不能为谷歌直接贡献营收，其带来的用户数据与广告协同效应，加上对微软护城河的削弱，也足以支撑其战略投入。 - 绝大多数普通消费者日常极少使用复杂的电子表格，而全球商业与企业内部的核心表格工作流，自始至终由 Excel 牢牢把守。 --- ### [03:19:59](https://www.youtube.com/watch?v=6oER1bdQa_s&t=11999s) 移动时代的利润池变迁与鲍尔默的硬件觉醒这场与谷歌的漫长较量，直接促使微软最高管理层意识到了一个严酷的现实：在移动时代，科技行业的利润池（profit pools）结构正在发生不可逆转的转移。比尔·盖茨执掌时期的核心信条是坚定不移地做纯软件平台，让周围所有的硬件和外设都必须围绕 Windows 运转并实现互操作。在 PC 时代，这种策略取得了惊人的成功——随着时间推移，几乎所有的硬件 OEM 制造商都在残酷的价格战中丧失了定价权和利润，而全部的超额行业利润几乎被微软一家吞噬。然而，史蒂夫·鲍尔默敏锐地察觉到，伴随谷歌 Android 阵营与苹果生态的崛起，移动硬件产业链绝不会重演 PC 时代的分配逻辑。未来的硬件生态不再保证能把行业利润拱手让给上游的纯操作系统开发商。商业世界有一条铁律：任何成熟企业都极难主动切入一个毛利率显著低于自身现有业务的领域。对于习惯了纯软件授权近乎无本万利毛利率的微软来说，甚至连投资建设数据中心、运营毛利相对较低的云计算业务都曾历经犹豫，涉足低毛利、重资产的硬件业务更是违背本能。但鲍尔默意识到，如果微软在移动浪潮中既没有搜索入口，又不敢涉足硬件终端，公司将彻底无法享受到移动时代的任何利润红利。正因如此，鲍尔默顶着巨大的内部阻力力推硬件，不仅批准推出了自主品牌硬件 Surface，更开启了收购诺基亚手机业务的激进冒险。 - 比尔·盖茨的战略核心是做高毛利的平台级纯软件公司，利用产业链将硬件厂商利润挤压殆尽。 - 鲍尔默意识到移动硬件平台不再是任由纯软件抽取利润的“工具人”，产业链价值向软硬件一体化和搜索服务倾斜。 - 微软面临跨越“利润率鸿沟”的痛苦抉择：从极高毛利的软件授权跨入毛利较低的云计算与硬件制造。 --- ### [03:22:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12123s) Windows 7：从“道奇蝰蛇”回归“丰田凯美瑞” 在全面转向硬件和移动化之前，微软必须先修补自己在桌面操作系统上的信誉裂痕。在 Windows Vista 惨遭滑铁卢后，原负责 Office 产品管理的史蒂芬·西诺夫斯基（Steven Sinofsky）被临危受命，调任主持 Windows 部门。西诺夫斯基将 Office 团队著名的铁血执行文化——“我们交付产品”（We ship）带到了 Windows。这种管理文化在业内令人叹为观止：提前整整三年敲定发布日期，随后组织起 6000 人的庞大团队，排除一切干扰，无论发生什么情况都必须在既定日期当天准时交付软件。西诺夫斯基与团队沟通时曾给出一个极其生动的汽车比喻：此时的 Windows 并不需要多么宏大的技术远见，曾经的 Longhorn/Vista 妄图打造一辆性能狂暴但极难驾驭的“道奇蝰蛇”（Dodge Viper），结果车毁人亡；而 Windows 此时真正需要的，是一辆绝对皮实、省心耐用的“丰田凯美瑞”（Toyota Camry）。西诺夫斯基在 Windows 7 上完美实现了这一点： 1. 它彻底抚平了企业和用户的创伤，没有那些花哨而不稳定的 Vista 特性与性能倒退，做到了“设备不崩溃、企业网络不瘫痪”。 2. 它保留了左下角熟悉的“开始”按钮、可预测的标准菜单、飞快的文件检索与响应速度。 3. 它重构了产品研发组织，建立了高度精简的“开发（Dev）- 测试（Test）- 产品经理（PM）”三权制衡架构，恢复了严格的排期与规划流程。 Windows 7 虽然没有带来令人瞠目结舌的颠覆性功能，但它就像丰田凯美瑞一样极其可靠地满足了企业与个人用户的一切刚需，成为微软历史上口碑最好的操作系统之一，也让团队有底气腾出手来思考更长远的下一代交互形态。 - 西诺夫斯基将 Office 团队“三年排期、严守交期、万人协同准时交付”的强执行力带入 Windows 团队。 - Windows 7 放弃了 Vista 追求“道奇蝰蛇”般的激进构想，选择成为稳健实用、不宕机、体验丝滑的操作系统界“丰田凯美瑞”。 - 伴随组织结构的扁平化与规范化，微软不仅修复了市场信任，更为后续下一代操作系统的规划打下了地基。 --- ### [03:24:14](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12254s) iPad 冲击与 Windows 8 的豪赌：触摸屏与开发者平台重构 Windows 7 大获成功之后，团队迅速开启了下一代操作系统 Windows 8 的筹划。西诺夫斯基带领团队走到了一个极其激进的十字路口：他们断定未来计算的终极范式是触控与平板电脑。虽然比尔·盖茨早在多年前就鼓吹平板电脑概念，但在市场上始终缺乏实质性的成功范本。正当 Windows 8 规划进行了一两年之际，2010 年苹果正式发布了 iPad。iPad 展现的多点触控交互犹如一记惊雷，既彻底验证了西诺夫斯基团队对于“大屏平板加多点触控是未来计算主流范式”的技术预判，也让鲍尔默和微软高管层陷入了空前的恐慌。在初代 iPad 发布会上，史蒂夫·乔布斯（Steve Jobs）公开提出了著名的论断：在后 PC 时代，传统的 PC 将沦落为用途专一的“皮卡车”（pickup truck），而像 iPad 这样的平板电脑将变成普罗大众日常使用的“家用小轿车”（car）。刚刚在智能手机市场被 iPhone 击溃的微软高层，眼睁睁看着苹果企图用 iPad 复制移动端的胜利，进而直接抄底微软最核心的 PC 领地。尽管站在今天复盘，事实证明将 iPhone 交互放大而来的 iPad 并没有真正取代 PC（反而是智能手机在很多轻量场景取代了 PC），但在 2010 年的节骨眼上，这种生存威胁对微软而言无比迫近。此外，当时 Windows 虽然坐拥惊人的营收和企业渗透率，却在核心技术心智上严重掉队——全世界最有激情的开发者都在涌向无需受制于底层的 Web 应用和移动应用，Windows 正在丧失引领行业下一个前沿的吸引力。为了同时解决这两大生存危机，微软在 Windows 8 上发动了一场一箭双雕的豪赌： 1. **产品与形态抗击**：在 PC 和平板之间全力押注触控模式，试图在微软的后花园里“比苹果还苹果”，坚决阻击 iPad 蚕食核心 PC 市场； 2. **开发者生态重塑**：彻底推翻旧模式，推出全新的触控模式 Metro UI，并将底层开发环境全面押注在 HTML 5 等现代 Web 标准上，试图通过拥抱前端开发者已经熟练掌握的技术栈，将流失到 Web 世界的开发者重新抢回 Windows 生态。 - 苹果发布 iPad 验证了触控大屏的潜力，乔布斯关于“PC 是皮卡，iPad 是轿车”的论调令微软高层极其忌惮。 - Windows 8 肩负双重战略使命：其一是用全触控体验反击 iPad 对 PC 领地的侵蚀；其二是重构开发者生态。 - 为吸引日益脱离 Windows 的 Web 开发者，微软在 Windows 8 的 Metro UI 环境中深度集成了 HTML 5 研发栈。 ### [03:28:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12483s) Windows 8 的分裂：强推 Metro UI 与“凯美瑞变滑板车” 微软为 Windows 8 勾勒的野心是一场全方位的激进重构。他们不仅要全面支持触控，还要构建一套崭新的工具链，让开发者基于 HTML5 打造的应用能够在 ARM 处理器上顺畅运行，以配合即将到来的平板浪潮——微软甚至亲自下场造硬件，推出了搭载 ARM 架构的 Surface RT。Windows 8 就此变成了“新开发者平台”加“触控优先”的双头怪兽。然而，这种触控优先理念在操作系统层面的落地方式，却引发了一场灾难性的用户认知混乱：传统的“桌面”竟然被降级成了一个普通应用；用户开机进入系统后，迎面而来的是铺满屏幕的动态磁贴（Live Tiles），想要找传统的桌面，还得专门去点磁贴或者在左下角摸索切换。主播 Ben 谈到，他在微软工作期间长期使用 Windows 8，那种双重交互隐喻并存的撕裂感令人发疯——在 Metro UI（后称 Modern UI）与传统 Win32 应用之间来回切换，一会儿在桌面运行，一会儿又掉入平板触控优化的应用切换器，完全摸不清逻辑。大卫用了一个极具画面感的比喻来形容普通消费者的崩溃心理：“这就好比我买了一辆丰田凯美瑞（Toyota Camry），满心以为它就是辆凯美瑞，结果开出来发现是个滑板车！” 这种拧巴的产物背后，其实源于微软内部高层的战略专断。最初的设计设想中，触控磁贴界面仅专属于平板设备，而针对桌面 PC 则有一版延续 Windows 7 经典体验的传统系统。然而，高层最终降下一道不可违抗的旨意：“Windows 就是 Windows，必须在所有设备上统一步调（Windows is Windows, we need to ship Windows across all devices）。”当时团队已经投入了海量的研发资源、政治资本与职业生涯押注在 HTML5 社区和 Metro UI 上，既然“只能有一个 Windows”，两个截然相反的交互系统便被强行缝合在一起推向了市场。 ```mermaid flowchart TD accTitle: Windows 8 策略冲突与系统缝合架构 accDescr: 展现微软将平板触控与传统桌面强行合并为一个系统的内部决策与市场后果 subgraph Origin["最初规划（分道扬镳）"] Tablet["平板设备分支 Metro UI / 动态磁贴 HTML5 / ARM 架构"] Desktop["桌面 PC 分支 延续 Windows 7 经典体验 Win32 / x86 架构"] end subgraph Mandate["高层一统意志"] OneWindows{"高层指令： 'Windows 就是 Windows' 必须全设备统一"} end subgraph Product["交付产物：Windows 8"] Hybrid["双头缝合怪 开机全屏磁贴 / 桌面沦为二级应用 撕裂的交互隐喻与多任务体验"] end subgraph Reaction["市场惨烈反馈"] Backlash["极客与 OEM 抵触 / 开发者冷眼旁观 / 商业溃败"] end Tablet --> OneWindows Desktop --> OneWindows OneWindows --> Hybrid Hybrid --> Reaction ``` --- ### [03:30:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12636s) 时代错配：触控灾难与“放大的手机”逻辑 Windows 8 的推行虽然不像当年的 Windows Vista 那样存在致命的基础性能缺陷，但整体宣发与市场落地却极其糟糕。专业科技媒体和评测人在花时间摸索熟悉后尚能理解其逻辑，但早在评测解禁之前，网络舆论就已被点燃——微软内部将这群声音极大、情绪极端的 0.001% 硬核老用户戏称为“地下室里的人”（the basement）。他们的愤怒迅速席卷全网，将整个产品的口碑彻底拉垮。同时，OEM 硬件厂商更是苦不堪言：硬件厂商根本没有做好全面转型生产触控设备的准备，而微软却将全部火力押在触控优化上，导致市面上充斥着夹生产品，消费者感到极度困惑：“我难道要站起来用手指戳台式机显示器吗？” 更致命的问题在于硬件技术的时代错配。2012 年的移动芯片性能极其孱弱，让低功耗的 ARM 处理器去同时承载完整的 PC 笔记本工作流，在物理上根本无法成立。iPad 当年之所以能大获成功，是因为乔布斯将其设计为“放大的 iPhone”，而不是“缩小的 Mac”。Windows 8 却试图在十多年前做出一台“缩小的 PC”，而现实的技术条件根本撑不起这样的野心。两位主播感叹，这种理念如果是放在今天或许能取得完全不同的结果。今天的硬件性能（特别是 Apple Silicon）早已过剩，很多用户都梦寐以求拥有一台像联想 Yoga 那样可以 360 度翻转的完美机器——比如一台 13 英寸 M3 MacBook Air，翻转过来后通过统一二进制机制无缝切换为运行 iOS 界面与触控应用的 iPad，共享底层数据与缓存，免去外出同时携带手机、平板与笔记本三件套的累赘。但在 2012 年，Windows 8 确实生不逢时，是一场彻头彻尾的商业滑铁卢。 - 开发者生态在 2013 至 2014 年间不仅没有被激活，反而跌落回冰点，没有任何人愿意针对 Windows 8 专属平台进行开发。 - 软件工程界所有的核心精力全部分流走向了两个阵地：Web 网页端应用（Web App）以及基于 iOS / Android 的移动原生应用（Mobile App），微软的生态被彻底绕开。 --- ### [03:33:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12795s) 消费级至暗时刻与企业级闷声发大财这标志着微软作为一家“消费级科技巨头”的彻底死亡。盘点当时的战场，微软几乎交出了全败的答卷：音乐播放器 Zune 夭折；Bing 搜索引擎市场份额微乎其微；Windows Phone 在移动操作系统大战中被 Android 和 iOS 彻底碾压；连统治核心地位的 Windows 8 也遭到了前所未有的冷遇。反映在资本市场上，微软的股价长达十年死死钉在 30 美元上下止步不前。2012 到 2013 年间的微软，陷入了外界眼中暗无天日的至暗时刻。然而，极具戏剧性的是，就在外界一片唱衰、消费业务节节败退的同时，微软财务报表上的光芒却刺眼得令人震惊——公司的营收和利润依然保持着极其恐怖的增长。除了 2008 年全球金融危机曾造成微小波动之外，微软庞大的企业级商业引擎（Enterprise）几乎从未出现过业绩下滑。即使操作系统接连经历烂版本，企业软件与销售体系依然如同永不减速的重型火车一样轰鸣前行。更深层的蜕变正在水面之下发生：云计算与 Azure 已经悄然成型并高效运转。微软并没有像外界所认为的那样故步自封，它确确实实完成了自我重构，并再次牢牢卡住了下一代计算技术的最前沿。只是这一切变革，完全发生在大众视线之外的企业级基础设施维度之中。 --- ### [03:35:10](https://www.youtube.com/watch?v=6oER1bdQa_s&t=12910s) 寻找“比尔·盖茨接班人”：Ray Ozzie 登场与云计算的真正源头公众对微软的转型普遍存在一个巨大的叙事误区：很多人误以为史蒂夫·鲍尔默只会抱着传统的 Windows 摇钱树不放、靠销售企业授权度日，直到后来萨提亚·纳德拉（Satya Nadella）接任 CEO 才力挽狂澜发明了 Azure。然而历史的真相并非如此，Azure 的起源要深远得多。早在 2004 至 2005 年间，比尔·盖茨开始筹划从微软彻底退休，准备辞去首席软件架构师（Chief Software Architect）的职务。鲍尔默非常清楚，自己本质上是个商业组织奇才而非前沿技术远见者，他必须找到一位能统领技术架构的“比尔”。然而盖茨在技术界的影响力无可替代，一个人填补不了这个空缺，微软的策略是物色“两个比尔”：对内提拔了克雷格·芒迪（Craig Mundy），而对外，他们锁定了最理想的天才人选——Lotus Notes 的缔造者雷·奥齐（Ray Ozzie）。 Ray Ozzie 在软件工程界享有传奇声誉，且与微软渊源极深。极少有人知道，早年著名的 Lotus 1-2-3 和 Lotus Notes 并不是同一个团队的产品：Lotus 1-2-3 是 Lotus 自研的，而 Lotus Notes 则是 Ray Ozzie 自己的创业软件工作室独立开发的，Lotus 仅充当发行商角色。这种独特的独立架构让 Ray 得以像游戏工作室周旋于发行商之间一样，不仅拥有自主权，还能私下与微软保持密切的技术互通与合作协议，甚至作为外包承包商深度参与过微软 LAN Manager 等早期核心产品的研发。到了 2004 年前后，Ray 创办了 Groove Networks，微软顺水推舟将整家公司连同 Ray 本人一并收购。 2006 年 6 月，当盖茨正式宣布退休计划时，Ray Ozzie 被官方任命为盖茨在微软首席软件架构师职位上的正式接班人。盖茨和鲍尔默交给他一项终极使命：“由你来擘画微软未来的技术愿景——微软拥有顶尖的资产、无敌的商业机器和满营的天才，在下一代计算技术浪潮中，请为微软指明出路。” 随后，Ray Ozzie 于 2005 年 10 月亲手撰写了那份奠定微软后二十年命运的纲领性文件——《互联网服务颠覆》（Internet Services Disruption Memo）。他在备忘录中深刻写道： > “时代环境再次发生了剧变。这一次，服务计算与通信技术经历了巨大而持续的进步，使得基于服务的商业模式成为现实。宽带与无线网络的普及彻底重塑了人与人之间的交互方式，人们正日益被那些体验简便、开箱即用且由服务所驱动的软件生态所吸引。” ### [03:38:34](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13114s) 颠覆旧世界的火种：Ray Ozzie 与“红狗”秘密孵化当企业开始重新审视 IT 基础设施的经济学时，一股潜在的巨变正在酝酿：如果企业不再需要为购买服务器、机架安装、运维数据中心以及处理隐私合规等琐事承担沉重的资本支出（CapEx），而是按需以订阅方式向大型数据中心租用计算能力，甚至直接购买托管的端到端解决方案（SaaS），商业形态将会怎样？2005 年，Ray Ozzie 敏锐地洞察到了这一趋势。2006 年 1 月，在首席执行官 Steve Ballmer 的全力支持下，Ozzie 开始在微软内部悄悄招募人才，筹备一个高度机密的孵化项目。值得玩味的是，这个后来孕育出云计算帝国 Azure 的项目，居然完全被隔离在微软现有的任何事业部之外——它没有被放在理所应当承接它的“服务器与工具事业部”（Server and Tools）。作为 Ballmer 时代微软企业级战略的核心支柱，Server and Tools 掌管着 Windows Server，这是一款卖给其他公司安装在自家服务器上的操作系统。按理说云计算必须从中诞生，但现实情况是，云模式对原有的业务模式和市场通路（GTM）构成了致命颠覆： - **渠道伙伴利益冲突**：微软原有的企业市场由埃森哲（Accenture）等咨询巨头、增值分销商（VAR）以及 OEM 硬件厂商（如 Dell）共同筑造。戴尔把服务器卖给最终客户，埃森哲负责到现场实施部署 Windows Server。如果微软突然跳出来说“我们自己把基础设施做成云服务直接租给你”，这无异于砸了所有生态伙伴的饭碗。 - **内部 KPI 与考核激励的断崖式错配**：Windows 与企业业务团队唯一的考核指标就是“卖出了多少份 Windows 许可证以获取授权收入”。而云计算是一项需要微软先砸数十亿巨资采购硬件、自建巨型数据中心、初期不产生任何软件授权许可、只能祈祷未来有人来按量付费租用的生意。“在当时的考核和薪酬体系里，Windows 根本不在这个等式之中”，没有任何部门愿意自己革自己的命。 ```mermaid flowchart TD accTitle: Azure 孵化背景与传统业务模式的冲突 accDescr: 对比传统 Windows Server 模式与新一代云服务模式在产品形态、渠道伙伴和核心考核指标上的根本差异。 subgraph Legacy["传统模式：Server & Tools 体系"] A["交付形态：Windows Server 软件许可"] --> B["通路：Dell/HP 硬件 + 埃森哲上门部署"] B --> C["核心 KPI：售出的 Windows 软件 License 数量"] end subgraph Cloud["新生模式：Red Dog 秘密孵化项目"] D["交付形态：自建机房提供基础设施 / 云服务"] --> E["通路：微软直接按需出租与订阅计费"] E --> F["核心指标：数据中心利用率与长期计算用量"] end Legacy -.->|"利益颠覆 & 组织排异"| Cloud ``` --- ### [03:41:53](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13313s) 传奇出山：Dave Cutler、手写 Hypervisor 与云端豪赌为了推进这个代号为“红狗”（Project Red Dog）的绝密计划，Ray Ozzie 请出了微软技术殿堂中最具威望的硬核传奇：Windows NT 的首席架构师 Dave Cutler，以及同样来自 DEC 的工程巨擘、曾与 Brian Valentine 一道把 Windows Vista 推向终点线的 Amitabh Srivastava。在底层系统与靠近裸机的代码开发领域，无人能出 Dave Cutler 其右。两位技术大师招兵买马组建团队，从零开始打造 Azure。最令技术圈叹为观止的是，Dave Cutler 甚至没有借助任何现存的开源软件，凭借一己之力从底层硬生生手写了一个全新的 Hypervisor（虚拟机监视器），作为支撑 Azure 虚拟化与多租户运行的基石。在当时，VMware 是一家市值庞大、倾举国之力研发虚拟化技术的独立软件巨头，而 Dave Cutler 却云淡风轻地表示“这活我一个人就能搞定”。 Steve Ballmer 顶住了来自 Windows 部门、渠道合作伙伴以及整个企业销售体系的巨大阻力与利益反扑。尽管 Ballmer 最初并未立即理解云的全部深意，但随着他与越来越多的企业级客户面对面交流，他清晰地意识到这是企业计算的必然未来。Ballmer 就像突然“扳动了开关”一样全面转向，毫不犹豫地下注：自建遍布全球的巨型数据中心需要动辄数十亿、上百亿美元的真金白银投入，这绝非无关痛痒的边缘试水，而是一场关乎公司命运的超级博弈。 --- ### [03:44:00](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13440s) PaaS vs. IaaS：保守的 Windows Azure 与开源宿怨在回顾这段历史时，主播们对“这是否是一场赌上公司命运的举动（Bet-the-company move）”展开了趣味辩论。Ben 认为，微软向来富得流油，花几十亿资本开支谈不上生死存亡，真正称得上“豪赌”的是把 Ray Ozzie 和 Dave Cutler 这样的国宝级技术元老全押在了这个项目上。此外，Azure 在诞生之初其实做了一定的妥协，并未在第一时间打破 Windows 中心主义： - **PaaS 定位而非 IaaS**：最初产品被命名为“Windows Azure”，底层只跑 Windows Server，并且主打 Platform as a Service（PaaS），这与亚马逊 AWS 从一开始就坚持的 Infrastructure as a Service（IaaS）有着鲜明差异。 - **对开源社区的长期防备**：彼时微软仍将开源（Open Source）视作洪水猛兽甚至“毒瘤”。微软的核心商业逻辑是为操作系统、编程语言、集成开发环境（IDE）和服务器软件收费，而开源世界却把这些全部免费奉送。虽然 Azure 最初没有直接拥抱 Linux 和 LAMP 架构（直到 2014-2015 年 Nadella 时代才彻底转变），但团队心知肚明——为了防止组织产生剧烈排异反应，他们只能采取渐进策略，日后必定要走向与 AWS 全面正面对决、提供企业信赖的一切服务。与此同时，SaaS 化的探索也悄然启动。微软与知名电池制造商金霸王（Energizer）等客户展开早期试点，开始尝试将 SharePoint 等协作服务以托管软件的形式对外交付，拉开了应用层云端化的序幕。 --- ### [03:46:30](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13590s) 易帅换将：Satya Nadella 挂帅与智能云的崛起到了 2010 年，Ray Ozzie 准备离开微软。在离职之际，他和 Steve Ballmer 将代号 Red Dog 的项目正式更名为 Azure，并整体并回了服务器与工具事业部（Server and Tools）。随后，Ballmer 亲赴华盛顿大学发表了一场极具历史意义的演讲，高调宣布微软“All In”云业务，正在“将公司的未来押注在云与 Azure 之上”——这番话表面是对公众表态，实则是说给微软内部各业务山头、尤其是服务器部门听的军令状。紧接着，Ballmer 做出了一个雷厉风行的决定：撤换服务器与工具事业部的总裁 Bob Muglia。尽管 Bob 当时极其出色，带领该部门做到年收入高达 120 亿美元、年增长率达 30% 至 40%（Bob 后来加盟 Snowflake 担任 CEO，战绩同样辉煌），但 Ballmer 坚信，微软需要一位完全没有历史包袱、能彻底将整个组织重塑为“云优先”（Cloud-First）的新统帅。Ballmer 从 Bing 搜索引擎部门抽调出了一位冉冉升起的高管——Satya Nadella。 - **培养未来领袖的伏笔**：调任 Satya Nadella 既是因为 Azure 亟需破局者，也是管理层为这位顶级潜质管理者量身定制的战略历练。Bing 的舞台已经不足以施展其抱负，唯有统领核心转型业务，才能为他日后顺理成章接任首席执行官奠定无可替代的功勋。 - **历史性的胜利**：如今微软财报划分为三大板块——生产力与业务流程（Productivity and Business Processes，涵盖 Office 365 及 LinkedIn）、更多个人计算（More Personal Computing，包含 Windows 与 Surface）、以及智能云（Intelligent Cloud，以 Azure 为核心）。当年这个被边缘孵化、饱受内部质疑的秘密项目，如今让“智能云”毫无悬念地成为了整个微软帝国体量最大、利润最丰厚的超级引擎。 ### [03:49:15](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13755s) 智能云的崛起与“混合云”护城河在微软的业务版图中，Windows 曾经是不可撼动的绝对核心，但如今智能云（Intelligent Cloud）已取而代之，不仅成为微软规模最大、利润最丰厚的业务，更是增长最迅猛的引擎。在 2023 财年，智能云部门创造了惊人的 880 亿美元营收。不过，需要指出的一个关键背景是，智能云板块中包含了 SQL Server 和 Windows Server 等庞大的传统企业级资产。在萨提亚·纳德拉（Satya Nadella）执掌初期大肆宣传云业务营收时，将这些利润丰厚的传统服务器业务打包计入云端数据，确实起到了极大的拉动与包装作用。但从战略竞争的角度来看，这绝非单纯的“财报财技”，而恰恰构成了微软抗衡 AWS 最核心的竞争壁垒——“混合云”（Hybrid Cloud）。在 Azure 崛起的初期阶段，微软敲开大企业客户大门时底气十足地宣称：企业客户无需在本地部署与公有云之间做非此即彼的痛苦抉择，微软既拥有世界级顶尖的公有云基础设施，又与客户机房里多年沿用的本地 Windows Server 架构无缝协同。这种混合形态成为了微软企业级市场的王牌。 ```mermaid flowchart TD accTitle: 微软混合云战略架构 accDescr: 展示微软如何利用本地部署的服务器遗产与公共云 Azure 结合，形成抗衡 AWS 的混合云优势。 subgraph OnPrem["本地传统企业基础设施"] WS["Windows Server"] SQL["SQL Server"] end subgraph PublicCloud["公有云创新"] Azure["Microsoft Azure 公共云"] end subgraph HybridAdvantage["企业混合云解决方案 (Hybrid Cloud)"] EnterpriseClient["跨国政企客户"] CompetitorDiff["对比 AWS 的独特优势：\n无缝迁移现有架构，保护历史软硬件投资"] end OnPrem --> HybridAdvantage Azure --> HybridAdvantage HybridAdvantage --> EnterpriseClient HybridAdvantage --> CompetitorDiff ``` - **2023 财年智能云营收**：达到 880 亿美元，超越 Windows 成为微软最大、最具盈利能力且增速最快的部门。 - **业务构成**：除了公共云 Azure 之外，还深度整合了 SQL Server 和 Windows Server 等传统企业级服务器资产。 - **混合云战略定位**：面对亚马逊 AWS 的纯公有云模式，微软为传统政企客户提供了一条过渡平滑的混合云路径。 --- ### [03:50:38](https://www.youtube.com/watch?v=6oER1bdQa_s&t=13838s) 云计算的长坡厚雪与技术基因沉淀回望这段历史，一个不争的事实是：云计算的市场空间实在太过浩瀚，庞大到微软在此期间所有的其他失误与惨败——无论是移动端的落后、硬件上的亏损还是消费级产品的碰壁——在云的大潮面前都显得微不足道。整个现代科技生态与消费级应用均运转在云端；哪怕不是微软、Meta、Google 或亚马逊自营的消费级服务，也必须租用这几大巨头的云基础设施，部分收入最终都会流向微软的口袋。更为关键的是，即便在物理世界的线下经济中，从联网汽车到餐厅刷卡收银系统（POS），一切现实商业动作都在加速依赖云端驱动。尽管微软在 PC 时代曾享有令人称羡的 90% 市场垄断份额，而在云市场上所占的份额远低于此，但云计算是一场长达三四十年的庞大长期顺风车（Secular Tailwind）。追根溯源，微软能够奇迹般赶上这班列车，关键的种子是在 2006 年播下的： - **组织孵化与高层庇护**：2006 年由雷·奥兹（Ray Ozzie）在核心业务部门之外设立孵化小组，并获得了首席执行官史蒂夫·鲍尔默（Steve Ballmer）全力提供的政治掩护与资源支持。 - **Bing 搜索引擎团队**：为公司内部磨练并沉淀了驾驭大规模分布式系统的底层工程实力。 - **Xbox Live 服务平台**：在极早期就运营着拥有 4000 万订阅者、全天候在线、实时且零延迟的全球网络基础设施。 - **超大规模互联网产品运营**：MSN 坐拥 7.5 亿注册用户的高并发流量处理经验，Hotmail 拥有数亿级用户依赖的高可靠 Web 应用程序架构。 - **企业协作套件经验**：SharePoint 和 Exchange 让微软深谙如何为大型企业构建基于服务器的应用软件。 - **底层系统级硬核功底**：Windows Server 团队长期积累了硬核服务器操作系统开发能力，为构建虚拟机监控器（Hypervisor）和底层虚拟化架构奠定了无可替代的工程基石。 --- ### [03:53:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14026s) 鲍尔默常被忽视的功绩与“创新者的窘境” 公众舆论通常将鲍尔默的执政末期视为微软错失时代的“溃败期”，而将所有的荣誉、转型叙事与云端胜利全部归功于萨提亚·纳德拉。然而剥开时间线来看，在鲍尔默交棒给纳德拉之前的整整七年里，Azure 一直在鲍尔默治下潜心研发；更是鲍尔默本人亲自挑选了纳德拉来统领这项业务，最终促成了纳德拉登上 CEO 之位并收获所有赞誉。在当时做出拥抱云端的决策极其违背本能。如果站在当时微软高层的视角，Windows 和 Office 是人类商业史上最赚钱的生意之一：边际成本为零、拥有强大的自我强化网络效应，以及极高的企业迁移壁垒。而云计算却是一个看似糟糕得多的商业模式： - **重资产投入与持续运营成本**：必须斥巨资购买土地、建设庞大数据中心、采购昂贵机柜，而且电力能源消耗是极其昂贵且无法抹去的真金白银成本。 - **亲手瓦解软件授权护城河**：客户在 Azure 服务器上运行的可能是自研软件甚至开源软件（如 Linux），企业可能不再购买 Windows 授权，也不再采购昂贵的 Exchange 或 Windows Server 许可。顶着如此巨大的“创新者的窘境”，微软依然敢于在 2006 年自我革命、押注一个商业模式可能远劣于主业的新生事物，鲍尔默与雷·奥兹在战略远见和保驾护航上功不可没。 --- ### [03:56:18](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14178s) Office for iPad 的“封存派对”与守成者博弈 Ben Gilbert 在节目中分享了自己亲历的内部切肤之痛。2011 年至 2014 年他在微软工作，2011 年实习时参与 Office 365 体系内的 Word Web App，而重返微软后全职参与的项目正是绝密研发的 **Office for iPad**。在 2012 年前后的现实世界中，用户迫切希望在自己的任何设备上随时访问文档，然而微软深陷“Windows 唯我独尊”的重力场。当时整个 200 人的跨部门团队紧锣密鼓奋战数年，终于迎来了原定的发布日。令人啼笑皆非而又心酸的是，团队最终迎来的不是“发布派对”（Ship party），而是一场“封存派对”（Shelf party）——产品被公司高层直接叫停、无限期封存。这起发生在 2013 年的产品雪藏事件，根本原因在于高层的防守诉求：微软当时刚刚推出了 Windows 8 和自研的 Surface 平板电脑，管理层制定的核心市场营销口号是“Office 在 Windows 上表现最好”（Office is first and best on Windows），必须向市场证明“全球唯独 Surface 平板能运行正宗的完整 Office”。在当时的普通工程师眼中，这种为了强推 Surface 和 Windows 8 硬件而拒绝满足亿万 iPad 用户刚需的做法，无疑是极其短视、把头埋在沙子里的“鸵鸟政策”。直到一年后萨提亚·纳德拉出任 CEO，Office for iPad 才被迅速解冻并正式推向市场，成为纳德拉新官上任的核心标志性事件之一。然而随着时间推移，Ben 坦言跳出年轻工程师的局部视角后，自己理解了巨头守成时的商业逻辑： - **上行收益有限**：即便立刻推出 Office for iPad，很难因此多签署几份巨额的企业续约协议（Enterprise Agreements）。 - **潜在下行毁灭性巨大**：在平板浪潮初现时，iPad 极有可能颠覆传统 PC 甚至从根基上杀死 Windows。在无法预判其是否会埋葬自己核心支柱前，过早将王牌产品双手奉上、为潜在竞争对手赋能，在战略上具有极大的自毁风险。 - **守成者思维（Thinking like an incumbent）**：推迟一年发布并未让微软损失任何真金白银，在当时战局未明的关口，选择等待与防守反而符合一个庞大商业帝国的核心利益考量。 ### [03:59:40](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14380s) 纳德拉时代的文化重塑与移动战略的落幕微软向移动时代的转型充满戏剧性。纳德拉在出任微软 CEO 的第一年里，把在 iPad 上发布 Office 作为一项显而易见的快速胜利（easy win），以此向外界释放明确信号：微软的企业文化已经改变。公司正从原本以“设备与服务”（Devices and Services）为主导的路线，彻底转向“云优先、移动优先”（Cloud first, mobile first）。在 iPad 上正式发布 Office，成为了这套全新战略叙事中最有力的现实佐证。伴随这一转折，微软在鲍尔默（Steve Ballmer）治下的移动端战役也走到了终章。为了理清这段历史，话题回到了最扑朔迷离的一桩公案——到底是谁决定收购诺基亚（Nokia）的手机业务？ ```mermaid flowchart LR accTitle: 战略转型与诺基亚收购决策 accDescr: 展示从鲍尔默时代的设备与服务转向纳德拉的云优先策略，以及诺基亚被收购的流程。 subgraph S1["鲍尔默晚期困境"] A["Windows Phone 遇冷"] --> B["诺基亚威胁转向 Android"] B --> C["70亿美元发起收购"] end subgraph S2["权力交接与新战略"] D["鲍尔默/盖茨退居幕后"] --> E["纳德拉接任 CEO"] E --> F["Office 登录 iPad"] F --> G["转向'云优先、移动优先'"] end C --> D ``` ### [04:00:25](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14425s) Android 的“负成本”模式与诺基亚的逼宫回顾 2011 年，微软发布 Windows Phone 之际，该系统在面对 Android 时几乎注定处于下风。著名投资人比尔·格利（Bill Gurley）曾撰文剖析 Android 的“低于免费”（Less Than Free）商业模式——Android 不仅是免费提供给厂商，Google 甚至愿意掏钱补贴推广： - 为“Droid”大型营销活动注资； - 向电信运营商及其门店销售人员发放激励佣金，引导消费者放弃 iPhone 转买 Android 设备。在各大手机厂商纷纷投向 Android 阵营时，唯有诺基亚愿意与微软站在一起。诺基亚自主的 Symbian 系统已近黄昏，开发新系统的尝试又受挫，面对无平台可用的绝境，时任诺基亚 CEO 斯蒂芬·埃洛普（Stephen Elop，前微软高管）在 2011 年 2 月宣布全面拥抱 Windows Phone。然而到了 2012 至 2013 年，Windows Phone 的颓势无可挽回，Android 成为绝对主流。2013 年诺基亚向微软下了最后通牒：“我们要转投 Android 了，除非你们能开出让我们留下的条件。” 在鲍尔默看来，这笔交易“只是钱的问题”（It was only money）。当时微软市值约 3000 亿美元，诺基亚手机业务收购价约为 70 亿美元，占市值的 2.3%。从现金流的角度看更为夸张：2013 年微软的营业利润高达 270 亿美元，账户上躺着巨额不知如何处置的现金，而且这些现金并没有在低迷的股价中得到华尔街应有的估值溢价。在鲍尔默眼里，这就像一笔风险投资（VC bet）——花掉市场根本没计入估值的现金，去博取保留移动端生态核心席位的胜率，况且背后还有被诺基亚倒戈背刺的直接业务下行风险。 ### [04:04:55](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14695s) 谁买下了诺基亚？高层换届与四十载历史重置这笔收购提议在微软内部和董事会引发了轩然大波： - 高管团队展开了一次非正式摸底投票（straw poll），绝大多数业务线主管投下反对票，其中就包括纳德拉。 - 董事会最初直接向鲍尔默表明：这项交易缺乏内部支持。 - 随后博弈急转直下。2013 年 8 月 23 日，微软与鲍尔默联合宣布，他将在未来 12 个月内退休，公司正式物色下任 CEO 人选。 - 仅仅 10 天之后的 9 月 3 日，微软正式宣布以超 70 亿美元收购诺基亚手机部门。究竟是鲍尔默以去职换取了这笔交易的通过，还是董事会主导了妥协，外界不得而知，留下了“到底是谁买了诺基亚”的谜题。 2014 年 2 月 4 日，萨提亚·纳德拉正式挂帅成为微软新任 CEO。同日，史蒂夫·鲍尔默卸任，比尔·盖茨也卸下董事会主席一职，由约翰·汤普森（John Thompson）接任董事长。盖茨与鲍尔默双双退居幕后，微软完成了彻底的权力交接。公司需要这次彻底的“文化重启”（Cultural reset），正如法务总长布拉德·史密斯（Brad Smith）早年向董事会建言“是时候讲和了”一样，微软也需要平息长达 40 年积攒下来的内部包袱：旧日战争的伤痕、反垄断的阴霾以及 Windows 迭代的技术泥潭。而在交接这一天，所有人都清楚微软内部孕育着一张足以翻盘的王牌——Azure。 - 2014 年 2 月 4 日权力交接当天，微软股价仅为 30.50 美元，总市值略低于 3000 亿美元。 - 十年之后的 2024 年，微软股价攀升至 465 美元左右，市值突破 3.5 万亿美元，重回全球市值第一宝座。其中至少有一半的市值直接受云计算和 Azure 业务的强力托举。 ### [04:09:20](https://www.youtube.com/watch?v=6oER1bdQa_s&t=14960s) 鲍尔默 14 年 CEO 任期的大盘点在深入复盘分析之前，节目对鲍尔默 2000 年初出任 CEO 至 2014 年初退位这整整 14 年的商业成绩进行了量化复盘。尽管公众舆论常将其任期视作停滞甚至失落的十四年，但财报核心指标却呈现出完全相反的商业基本面： | 指标 | 2000 年出任 CEO | 2014 年卸任 CEO | 14 年增幅 | |:--- |:--- |:--- |:--- | | **营业收入 (Revenue)** | 230 亿美元 | 840 亿美元 | 约 **3.65x** (超 3.5 倍) | | **营业利润 (Operating Income)** | 120 亿美元 | 300 亿美元 | 约 **2.5x** (接近 3 倍) | | **市盈率 (P/E Ratio)** | **75x** (峰值达 80x) | 约 15x 谷底水平 | 估值遭遇全面压缩杀跌 | 鲍尔默接手公司时，正值互联网泡沫顶峰，微软市盈率高达 75 倍（前一个月曾冲到 80 倍历史极值），即便是 2024 年在全面 AI 热潮席卷、市值超过 3.5 万亿的今天，微软的市盈率也仅在 40 倍上下浮动。鲍尔默是在估值倍数历史绝顶处掌舵，在随后的十四年里，他用近乎三倍的营收和利润爆发式增长，去消化了那场世纪泡沫留下的巨大估值溢价。 ### [04:10:36](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15036s) 鲍尔默接盘的“烂摊子”与惨烈的估值压缩本段讨论从对史蒂夫·鲍尔默（Steve Ballmer）接任微软首席执行官时极其险恶的外部环境剖析切入。在鲍尔默上任之际，微软正面临美国司法部（DOJ）反垄断裁决甚至拆分公司的危机，头顶是正在破裂的互联网泡沫（Dot-com bubble），而且他接替的还是神话般的世界首富比尔·盖茨（Bill Gates）。主播们直言，在那个时刻接盘完全是“为团队扛雷”，接手了一个近乎不可能的烂摊子（garbage sandwich）。当微软市值被推到顶点时接手，注定了后续极其残酷的股价表现。当时微软的市盈率（P/E multiple）高达 75 倍。随着泡沫破裂，第一年估值倍数迅速理性化，并在其整个任期内持续缩水，最终跌至 14 倍。微软市值从他被任命时的 6000 亿美元骤降至他离职时的 3300 亿美元。即便微软的营收和利润在他任期内翻了三倍，这种持续的估值压缩（multiple compression）也彻底抹平了基本面带来的增长，导致股价几乎横盘了整个十四年。在 2009 年金融危机期间，投资者甚至能以 2.1 倍的市销率（P/S）买入微软股票；即使到了 2013 年，市销率也仅仅只有 3 倍。 - 鲍尔默接任时微软处于 75 倍 P/E，离任时被压缩至 14 倍。 - 市值从接任时的 6000 亿美元降至离职时的 3300 亿美元。 - 估值压缩最极端时，2009 年市销率跌至 2.1 倍，2013 年市销率仅为 3 倍。 --- ### [04:12:49](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15169s) 华尔街为何不买账：讲不好故事的代价两位主播深度探讨了为什么投资者在长达十余年的时间里对微软的营收和利润增长完全无动于衷。核心原因在于资本市场对微软长期相关性的极度怀疑：人们不再认为这里是让用户或开发者兴奋的地方，管理层从天才创始人比尔·盖茨换成被视为“销售和市场先生”的鲍尔默，产品线凌乱不堪，Windows 不再引领未来，搜索（Google）与社交（Facebook）浪潮接连错失，硬件与消费端产品节节败退。更根本的盲点在于，投资者错误地认为 Windows 和 Office 这两大现金牛毫无粘性，只盯住那些失败的新业务，却完全没有把现有企业客户的极高留存率和生命周期价值计入股价。但两位主播指出，这归根结底是 CEO 的沟通失败——微软在这段时期“完全没有讲好自己的故事”。消费者看不懂微软的战略，开发者摸不着头脑，华尔街自然也不会买单。相比之下，萨提亚·纳德拉（Satya Nadella）接手后最关键的一步，就是将核心叙事精简为“移动优先，云优先”（Mobile first, Cloud first），并通过反反复复重申十倍以上的沟通频率，瞬间理顺了市场预期。 - 资本市场对旧业务产生误判，忽视了 Windows 和 Office 超强的护城河与持久盈利能力。 - 缺乏连贯的叙事能力（storytelling），导致即使营收利润翻倍、悄悄孵化出 Azure，市场也拒绝给予溢价。 - 企业领导者必须反复传达核心战略，纳德拉以“移动优先，云优先”成功重构了资本市场信任。 --- ### [04:16:02](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15362s) 横向对比同期科技巨头：幸存即是前 5% 的胜利评价鲍尔默的成绩时，人们往往拿微软与同时期崛起的最具颠覆性的独角兽对比，例如搜索领域的 Google 和社交领域的 Facebook。然而主播们认为，拿相邻领域中人类商业史上资本效率最高、增长最快的历史级黑马做标杆，对现任 CEO 来说并不完全客观。如果把目光投向 2000 年代初的“老牌科技巨头同盟”——雅虎（Yahoo!）、AOL、时代华纳等媒体巨头、惠普（HP）、北电网络（Nortel），会发现其中绝大部分公司都遭遇了毁灭性崩溃或边缘化。在那代巨头中，真正活下来并维系繁荣的屈指可数：除了微软，就只有戴尔（Dell）、在微软注资援助与乔布斯回归后起死回生的苹果（Apple），以及甲骨文（Oracle）。仅仅是让一家 30 年历史的庞大企业在技术范式更迭中平稳过渡 15 年，保存核心人才资产与业务延续性，就足以跻身同类企业的“前 5%”。尽管微软错失了消费科技的领导权（Zune、Windows Phone 等未达预期），但它成功打下了向企业级服务与云基础设施（Azure）转型的底座。 | 公司类别 | 2000 年代初巨头代表 | 2014 年终局走向 | |:--- |:--- |:--- | | **严重衰落 / 瓦解** | 雅虎 (Yahoo!)、AOL、北电网络 (Nortel)、惠普 (HP) | 核心业务衰亡、分拆、破产清算或被边缘化 | | **穿越周期 / 维持繁荣** | 微软 (Microsoft)、戴尔 (Dell)、甲骨文 (Oracle)、苹果 (Apple) | 守住现金流基本盘，依托企业服务或新硬件实现重生 | --- ### [04:19:09](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15549s) 商业分析：进入“七种力量”框架节目推进到每期节目的经典分析环节——基于汉密尔顿·赫尔默（Hamilton Helmer）经典著作《七种力量》（7 Powers）的商业护城河分析，探寻微软究竟凭借何种机制建立起持久的差异化超额回报。这七种力量分别为：逆向定位（Counter-Positioning）、规模经济（Scale Economies）、转换成本（Switching Costs）、网络效应（Network Economies）、流程特质（Process Power）、品牌（Branding）以及锁定资源（Cornered Resource）。主播回顾道，在第一期微软史（比尔·盖茨时代）中，微软几乎集齐了所有七种力量；但在鲍尔默时代，微软作为既得利益垄断巨头，几乎天然丧失了“逆向定位”（Counter-Positioning）的能力，反而在各项业务上面临后来者的颠覆。唯一的局部例外是在 Azure 的早期推进中：面对亚马逊 AWS 的纯公共云打法，微软凭借传统企业软件服务商的身份，向财富 500 强客户打出“混合云”（Hybrid Cloud）战略，依靠多年累积的信任感对 AWS 形成了局部的逆向定位，成功切入了云基础设施战场。 ```mermaid flowchart TD accTitle: 微软 Azure 早期逆向定位示意图 accDescr: 展示微软在云计算市场如何利用传统企业信任与混合云定位对抗 AWS 的纯公共云架构。 subgraph AWS ["竞争对手: AWS"] AWS_Model["纯公共云模式 (Pure Public Cloud)"] AWS_Tension["难以兼顾传统私有架构，对老牌巨头缺乏历史信任"] end subgraph MSFT ["微软: Azure 策略"] MSFT_Model["混合云战略 (Hybrid Cloud)"] MSFT_Trust["财富 500 强长期合作与企业 IT 信任"] end AWS_Model -.->|"逼迫传统企业全量搬迁"| MSFT_Model MSFT_Trust -->|"利用传统 IT 绑定形成差异化防御"| MSFT_Model ``` ### [04:20:38](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15638s) 七种力量框架复盘：规模经济与不可逾越的企业端转换成本在分析商业护城河（7 Powers）时，主持人们回顾了移动领域的逆势对抗（Counter-Positioning）——谷歌当时凭借“比免费更便宜”（less than free）的商业模式直接打破了传统格局。然而，微软在鲍尔默时代最核心的超能力莫过于**规模经济（Scale Economies）**。拥有数以亿计的庞大用户与客户体量，微软无论进行何种规模的研发投入或并购，成本都能被这个天文数字般的用户基底所摊薄。只要他们能在企业协议（EA, Enterprise Agreement）上每台机器多收 1 美元，任何增量 R&D 和收购便完全站得住脚。这种无解的规模红利不仅在桌面时代牢不可破，而且在云计算（Cloud）时代展现出了更惊人的规模回报率。相比之下，微软的**流程力（Process Power）**在这一时期则出现了明显倒退。著名的“黑色代码”（Blackcomb）、Longhorn 研发崩溃以及后来的 Windows Vista，都充分暴露了软件研发体系的混乱。他们曾经凭借 Windows 95 发布了史上最受欢迎的操作系统，在反垄断泥潭中依然能够推出 Windows XP 这一经典之作；然而到了 Longhorn 和 Windows 8，尽管技术痛点各不相同，微软却彻底遗忘了这些核心王牌业务究竟为何具备经济价值。与此同时，其消费级市场的品牌价值（Branding Power）也被严重侵蚀，但戏剧性的是，他们赢得了企业级市场（Enterprise）的极高信任度——当企业面临微软和 AWS 提供完全相同的产品功能时，企业往往更倾向于把预算交给微软。微软真正的核心护城河在于无懈可击的**转换成本（Switching Costs）**。在企业端，企业签订的 Enterprise Agreement 几乎让 IT 部门根本无法换掉微软的产品堆栈：即使眼下合约即将到期，下一份续约协议也会打包更多超值服务，促使客户继续绑定。当年美国司法部（DOJ）和全球监管机构严防死守的是消费级浏览器捆绑，而实际上真正惊人的捆绑发生在企业市场——企业只要按网络内设备数量给微软交一笔固定费用，就能获得包罗万象的整套软件全家桶。即使市面上存在数百个表现优于微软的单点解决方案（Point Solutions），企业也绝无可能彻底放弃微软体系。而在网络效应（Network Economies）与独占资源（Cornered Resource）维度上，微软的优势开始被稀释：随着跨平台互操作性提升和文件格式标准化，网络效应不再专属绑定于单一操作系统。 ```mermaid flowchart TD accTitle: 微软2000年代企业端护城河分析 accDescr: 展示规模经济与转换成本如何筑起坚固防御，而流程力和消费级品牌力出现衰减。 subgraph CorePowers["核心优势 (Retained / Gained)"] Scale["规模经济 (Scale Economies) 天量用户分摊 R&D 与收购成本 云计算基础设施超高规模回报"] Switching["极高转换成本 (Switching Costs) 企业协议 (EA) 深度绑定 全家桶捆绑，单点方案无法替换"] EnterpriseBrand["企业级品牌信任 (Brand Power) CIO 与企业决策层首选供应商"] end subgraph ErodedPowers["减弱 / 丧失的优势 (Lost / Eroded)"] Process["流程力缺失 (Process Power) Longhorn 崩溃、Vista 延宕、Win8 迷失"] ConsumerBrand["消费级品牌力萎缩 在移动与消费硬件端被边缘化"] Network["网络效应稀释 文件格式标准化、跨平台互操作性削弱了 OS 壁垒"] end Scale --> EnterpriseDominance["坚如磐石的企业级帝国 （Enterprise Cash Cow）"] Switching --> EnterpriseDominance EnterpriseBrand --> EnterpriseDominance ``` --- ### [04:24:31](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15871s) 商业法则：零和博弈的企业文化与业绩停滞的恶性循环复盘微软当年的战略教训，首当其冲的是**公司文化（Cultural Shift）**的断崖式转变。在整个鲍尔默时代，微软股票价格长期处于横盘滞涨状态。一旦全公司员工普遍确信“无论怎么努力股价都不会再涨”，整个企业环境就瞬间沦为一场零和博弈（Zero-Sum Game）。既然做大蛋糕无法体现在公司市值上，个人收益与职级晋升的唯一途径，便变成了踩着内部同事上位： - 自己的产品线必须吞并隔壁团队的产品线； - 自己的团队必须兼并兄弟部门； - 个人为了晋升，不惜以让同僚出丑和失败为代价。这种零和死局被公司严苛的**末位淘汰与强制分布机制（Stack Ranking）**进一步激化。在著名的 1 到 5 级评级制度下，每位主管只能给出固定比例的优秀（1分和2分）名额。每隔半年或一年的绩效评估，员工实际上是在与身边的密友和并肩作战的同事进行残酷的内卷角斗。微软内部那张著名的组织结构漫画——各个事业群将枪口一致对准彼此——正是这一机制下的真实写照。蛋糕不再增长，内部恶性竞争就成为必然的副产品。 --- ### [04:26:28](https://www.youtube.com/watch?v=6oER1bdQa_s&t=15988s) 顺应用户浪潮：从“逆潮流抗争”到萨提亚时代的“拥抱与扩展” 微软在 2000 年代犯下的另一个核心战略错误是：**固执地逆流而上，试图与科技历史的趋势对抗**。无论是开源软件（Open Source）的崛起，还是基于 Web 的轻量级开发生态，微软都本能地建立高墙，阻碍潮流，试图把所有人强行拉回自家专有的生态护城河中。然而事实证明，任何公司都无法长期抗拒用户真正的意愿。具有讽刺意味的是，萨提亚·纳德拉（Satya Nadella）上台后带领微软重回巅峰的策略，恰恰是复兴了微软历史上经典的“拥抱与扩展”（Embrace and Extend）战术： 1. **彻底拥抱既有用户诉求**：用户想使用开源软件就支持开源，想在 iPhone 上办公就适配 iOS，想搭建 Web 应用就提供云基础设施； 2. **构建新型商业模式进行扩展**：依托用户已经形成的使用习惯，发明能够盈利的产品与生态系统。这种转变同样体现在微软企业业务的生死蜕变上。早期微软对待企业 IT 的态度是控制与防御——封锁网络以防员工“乱来”。但 iPhone 引发的移动大潮彻底冲垮了这一藩篱，企业员工对移动设备和灵活软件的渴望无法压制。从本地软件转向**云计算（Cloud）**是微软的拯救之举，它让微软从对用户发号施令、处处提防的“对抗者”，真正转变为服务用户的共赢合作伙伴。 --- ### [04:28:57](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16137s) 战略超前与战术失误：时机、工程实现与审美品味的脱节回看鲍尔默时代的诸多尝试，主持人们给出了一个辛辣但中肯的定论：微软在许多领域的战略方向完全正确，但**时机（Timing）、工程落地（Implementation）与审美品味（Taste）**却极度脱轨。换言之，微软是“战略上正确，战术上失调”（Strategically correct, but tactically misguided）。比尔·盖茨的远瞻洞察几乎全部被时间验证，却在微软手中惨遭滑铁卢： - **触控计算（Touch Computing）与自然用户界面（NUI）**：盖茨极早就断言触控是未来，但微软押注了需要手写笔的电阻屏和强行缩小版的 Windows PC 桌面；而后来苹果证明了电容多点触控与基于手机生态放大的平板才是正解； - **交互式电视（Interactive TV）与数字媒体**：微软在千禧年前后大规模布局机顶盒与网络电视，但这项技术过早启动了 5 到 20 年——真正的流媒体革命必须建立在宽带互联网普及、带宽爆炸的成熟网络基础之上，而不是在互联网黎明期去硬造交互电视； - **移动计算（Mobile Computing）**：盖茨很早就笃定移动设备会重塑计算版图，但微软却在底层架构尚未成熟时过早开局，并一次次押错了技术范式和标准。 ### [04:30:11](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16211s) 预测未来与产品落地的脱节：“包夹策略”的失效微软对科技未来走向的预测常常极其精准，但在将预见转化为正确产品的执行力上却频频碰壁。互动电视最终的形态是 Netflix 与 YouTube 这样的点播流媒体平台，而不是微软早期设想的、架构在 Comcast 等传统有线电视网络之上的交互层；微软布局移动生态整整早了五年，但当时他们的思维更偏向于嵌入式设备，而非将 PC 架构完整缩小。这种洞察极准但形态跑偏的现象令人费解，因为在历史上，微软曾极其擅长把控技术转型的节奏。微软内部曾有一种经典的研发哲学，被称为“包夹策略”（Bracketing）：面对一项新技术时，微软会同时并行启动两个产品线——一个瞄准当前技术边界之下的低端实用方案，另一个则激进地瞄准未来边界之上的高端方案。随着发布窗口临近或市场格局明朗，团队再视情况选择是将低端方案增强，还是将高端方案削减功能并落地。例如在 IBM 时代有 Windows 对决 OS/2；在互联网初期有 Web 浏览器对决互动电视架构；而在操作系统演进中，则有小步迭代的 Longhorn 对决野心庞大到最终流产的 Blackcomb。然而到了鲍尔默执掌的这一时期，“包夹策略”所带来的多向期权价值突然瓦解了。公司不再能维持多路径探索并在中途游刃有余地收敛，而是僵化为在每个方向上押下单点注码，内部的创新尝试也无法继续茁壮成长。这种“把握住好点子并在适当时机以正确方式执行”的核心能力，在这十年间悄然失灵。 ```mermaid flowchart TD accTitle: 微软传统的包夹策略模型 accDescr: 展示微软过去如何通过高低端两条产品线并行研发，根据市场成熟度动态调整并最终收敛交付产品。 subgraph Inputs["研发启动期"] Low["低端产品线 （当前技术能力之下 / 稳健迭代）"] High["高端产品线 （当前技术能力之上 / 激进远景）"] end subgraph Convergence["市场演进与收敛期"] Decision{"市场与技术明朗化"} OptA["增强低端方案功能"] OptB["削减高端方案复杂度"] end subgraph Output["最终交付"] Product["推向市场的杀手级产品"] end Low --> Decision High --> Decision Decision -->|"技术未熟 / 强调务实"| OptA Decision -->|"技术就绪 / 追求突破"| OptB OptA --> Product OptB --> Product ``` --- ### [04:32:03](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16323s) 正和领导力：比尔·盖茨与史蒂夫·鲍尔默的“1 + 1 = 5” 从领导力模型的角度剖析，比尔·盖茨与史蒂夫·鲍尔默在分工明确、彼此高度尊重的黄金期，形成了一种极为罕见的“正和领导力”（Positive Sum Leadership）。两人的组合远胜过任何单独一人的领导效能，实现了“1 + 1 = 5”的化学反应。这种互相成就的默契，恰似顶尖团队或创业搭档之间不可替代的协同效应。在此期间，单打独斗的盖茨极易沉迷于像 WinFS 那样宏大却脱离现实的理论技术体系；而鲍尔默则需要一位能够赢得数万名顶尖工程师绝对忠诚的“技术灯塔”，引领全员向同一愿景靠拢。更重要的是，鲍尔默需要一位敢于直面新数据而瞬间推翻既有假设的搭档。盖茨拥有海量信息处理能力，一旦察觉新事实（例如互联网浪潮的兴起），他会毫不犹豫地宣布“全盘推翻重来”；而鲍尔默则擅长驱动大象起舞，把企业内外如航空母舰般的庞大组织校准并坚定执行。两人的结合，恰到好处地在“战略韧性”与“敏捷应变”之间取得了平衡。 --- ### [04:34:08](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16448s) 生态绑定的双刃剑：伙伴优先与无法按下的“重启键” 微软是一家极度以合作伙伴为导向（Partner-centric）的企业。在其生态体系中，独立软件开发商（ISV）、增值转售商以及零售渠道所赚取的商业利润，甚至远远超过了微软自身。然而，这种深度共生也成了最大的负债：它让微软彻底丧失了按下全局“重启键”的能力。对比之下，乔布斯重返苹果时能够破釜沉舟，全面重构开发工具、硬件产品线与底层操作系统；但微软背负着无数依赖其生态生存的外部利益相关方，为了维护庞大的向下兼容与生态繁荣，根本无法割席重来。外界常有一种浅层论调，认为“苹果能在移动端获胜仅仅是因为他们在 PC 桌面端彻底输了”，但更深刻的商业本质在于：企业外部依附的利益链条越长、切换成本越高，进行跨时代代际切换的阻力就越大，而颠覆性的下一代技术往往必须依赖彻底的重启。 --- ### [04:35:22](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16522s) 鲍尔默时代的最大败笔：不会讲故事与被低估的辉煌在梳理完这段庞杂的历史后，两位主持人提炼出了本期节目最核心的终局洞见（Landing the plane）：外界对所谓“微软失落的十年”存在根深蒂固的误解与偏见，而造成这种局面的罪魁祸首，正是微软自身叙事能力（Storytelling）的严重崩盘。鲍尔默在科技泡沫的最顶峰接任 CEO，面临着全行业估值倍数严重透支的历史包袱。但为什么在其整个任期内，微软的股价始终在 30 美元附近徘徊？核心原因在于公司无法向市场清晰传达自身的战略全貌。这导致舆论形成了一种“微软干啥啥不行、在消费级市场一无是处、正走向边缘化”的自证预言式负面叙事。反观亚马逊，贝索斯成功向公众和投资者植入了“我们是一家勇于发明与漫步探索（invent and wander）的公司”的认知框架。在这套叙事包装下，哪怕亚马逊在 Fire Phone、各种线下零售探索（如 Amazon Go、本地餐饮外卖业务）甚至早期智能硬件上遭遇惨重失败，公众也只会将其赞美为“拥抱失败的伟大创新精神”；而微软遭遇挫折时，公众舆论却是一边倒地认定其“毫无作为、彻底过气”。 ```mermaid flowchart LR accTitle: 叙事框架对企业公众形象与估值的影响对比 accDescr: 对比微软与亚马逊在面对业务失败时，因不同叙事框架而导致的完全相反的公众舆论与估值认知。 subgraph Amazon["亚马逊模式"] A1["业务重挫 (Fire Phone / 本地生活业务)"] --> A2["核心叙事： '发明与漫步探索、拥抱失败'"] A2 --> A3["资本市场评价： 具有远见的持续创新者"] end subgraph Microsoft["微软鲍尔默时期模式"] M1["业务重挫 (Windows Phone / Zune 等)"] --> M2["叙事缺位 / 僵化解释： 缺乏一致性的全局故事"] M2 --> M3["资本市场评价： 丧失创新力的迟钝垄断者"] end ``` 实际上，鲍尔默时期的微软绝非昔日走下坡路的 IBM。即便在此期间错失了移动端（在面对 Google 凭借搜索盈利而将 Android 以“低于免费”的方式倾销时，传统软件授权模式几乎注定无法招架），微软依然展现出了极其罕见的战略生命力： - 将公司的整体营业收入与利润足足翻了三倍。 - 孵化并重注了 Azure 与企业级云计算基础设施。 - 为微软换取了从容进入云原生与 AI 时代的绝佳入场券，最终支撑其重新登上全球市值第一的宝座。 ### [04:40:37](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16837s) 好物推荐环节：Meta Ray-Ban 智能眼镜与 Ozlo Sleepbuds 睡眠耳机在总结完本期的宏大商业复盘后，节目进入了标志性的“Carve Outs（好物推荐）”环节。David 率先拿出了两款硬件科技产品。第一款是他对 Ben 过去推荐的“二次安利”——Ray-Ban Meta 智能音频眼镜。David 表示自己终于入手了一副，体验极佳。由于不需要在耳朵里塞入耳塞就能收听到环境音频，这一特性在家庭场景中堪称神器，尤其是当他需要扮演“婴儿监视器”的角色时，既能实时听清婴儿床里的任何动静，又能毫无障碍地与妻子、家人或朋友交谈。第二款硬件产品来自一家名为 Ozlo 的初创公司所推出的 Ozlo Sleepbuds 睡眠耳机。David 分享道，近几年来他几乎每天晚上都需要借助音频助眠，或者在半夜醒来时听点声音重新入睡。过去几年他一直戴着 AirPods 睡觉，但侧卧时 AirPods 会狠狠硌住耳朵，非常痛苦。而 Ozlo 的这款耳机专为睡眠设计，体积小巧且极为平整，即使侧躺把耳朵压在枕头上也完全不会向外凸出。David 补充道，这个团队其实正是此前在 Bose 负责 Bose Sleepbuds 的原班工程团队；在 Bose 决定砍掉该产品线后，他们选择离开并创立了这家初创公司，完整保留了 Bose 级别的精湛工程设计。Ben 听闻当场被种草，直呼“马上下单发链接”。 --- ### [04:42:23](https://www.youtube.com/watch?v=6oER1bdQa_s&t=16943s) Ben 的极致基础之选：M3 MacBook Air 与特斯拉 Model Y 轮到 Ben 分享时，他笑着自嘲自己的推荐是“有史以来最基础、最没有悬念的日常产品”。第一款是他前面顺带提到的 M3 MacBook Air。虽然他承认自己每次换新 Mac 都会感慨“这是我拥有过的最棒电脑”，但他家里主力机是一台 16 英寸的 M1 MacBook Pro，出差坐飞机携带时沉重得宛如一块巨石。随着两人近期出差和旅行频繁，这台极其轻薄、飞速且精致的 MacBook Air 成了他差旅途中的救星。第二款推荐同样“极其可预测”——特斯拉 Model Y。在 David 调侃他“终于加入特斯拉大军”后，Ben 分享了自己前不久自驾前往奥卡斯岛（Orcas Island）的周末旅行。整趟行程开了好几个小时的车、搭乘轮渡、在人口稀少的岛上四处转悠并返回，期间一次电都没有充过，回到家时电池依然剩余 18%。除了特斯拉强大的 Supercharger 超充网络兜底外，那种难以置信的加速感与驾驶乐趣也让他彻底折服。Ben 感叹自己终于理解了为什么大家把特斯拉叫作“装了轮子的 iPhone”，如今只要换回燃油车开，就会觉得浑身别扭且油腻，而 Model Y 给人的感觉则是干净、纯粹。 --- ### [04:43:54](https://www.youtube.com/watch?v=6oER1bdQa_s&t=17034s) 幕后致谢：协助梳理微软三十年兴衰的幕后功臣们随后，节目进入了漫长而真诚的致谢名单。在感谢了赞助商 J.P. Morgan Payments、ServiceNow 和 Pilot 之后，Ben 和 David 表示，为了制作本期微软史诗级节目，他们前后深入访谈了超过 20 位核心当事人与行业重磅嘉宾。Ben 率先列出了协助他的各位行业传奇： - **Brad Silverberg**：曾长期领导微软 Windows 部门，主导了里程碑式的 Windows 95 的研发与团队组建。 - **Thomas Reardon**：当年 Internet Explorer（IE 浏览器）的创始团队核心成员。后来他创办了神经接口技术初创公司 CTRL-labs，被 Meta 收购后正致力于打造通过手部微小肌肉神经颤动即可控制界面的前沿交互。 - **Steven Sinofsky**：曾先后执掌微软 Office 以及 Windows（负责 Windows 7 与 Windows 8）。David 为此完整研读了 Steven 撰写的上千页巨著《Hardcore Software》（该书脱胎于其博客，厚如案头教材）。Steven 目前担任 Andreessen Horowitz 的董事会合伙人，与两人深入对谈了数小时。 - **Julie Larson-Green**：曾与 Steven Sinofsky 紧密合作，深度掌管 Office 与 Windows 产品的设计与体验。 - **Anand Rajeswaran**：Ben 在微软开发“iPad 版 Office”时期的老同事，帮助 Ben 细致复盘了当年产品临近发售又被搁置、随后新 CEO 萨提亚·纳德拉上任后迅速放行发布的全过程内幕。 - **Fritz Lanman**：曾在微软担任战略与企业发展（Corp Dev）要职，如今是 ClassPass 与 Mindbody 合并公司的 CEO。 - **Charlie Songhurst**：曾在微软企业战略部门与 Fritz 紧密共事，被公认为极具远见、思维广阔的投资与战略大师。 - **Jon Rubinstein**：曾领导苹果硬件工程部门、后出任 Palm 公司掌门人。他从苹果竞争对手的角度，生动还原了与微软激烈交锋的岁月。 - **Ray Ozzie**：微软前首席软件架构师，如今正在创办物联网初创公司 Blues Wireless。David 与 Ben 称赞他为人极为随和且极富传奇色彩。在 Zoom 通话中，Ray 还热情地展示了他办公室里收藏的 70、80、90 年代的老式计算机硬件。 - **Rob Glaser**：90 年代任职于微软，后创立了流媒体巨头 RealNetworks。 - **Joe Belfiore**：Windows 及 Windows Phone 的灵魂人物，曾在 2001 年的 Windows XP 纽约发布会上与知名主持人 Regis Philbin 同台演示系统功能（正如当年 Windows 95 发布会邀请 Jay Leno 一样）。 - **Vivek Varma**：在微软反垄断风暴时期深涉公关与法律事务的核心高管。 - **Steve Ballmer（史蒂夫·鲍尔默）**：微软前 CEO。两人特别感激鲍尔默在百忙之中（尤其正值洛杉矶快船队的休赛期自由球员交易期）慷慨抽出时间接受访谈，极大地帮助他们厘清和打磨了对这一时期的思考逻辑。 - **The Science of Hitting**：知名投资分析 Substack 创作者，向节目组无私分享了一份详尽的微软历年财务指标大表格，使得播客录制时能随时进行精确的数据速查。 - **Todd Bishop**：GeekWire 联合创始人。Todd 调出了自己在 2000 年代初担任《西雅图邮讯报》（Seattle P-I）记者时从未公开发布过的原始现场录音，让两人得以聆听当年比尔·盖茨和史蒂夫·鲍尔默在非公开场合的最原汁原味的采访原声。 --- ### [04:48:38](https://www.youtube.com/watch?v=6oER1bdQa_s&t=17318s) 资深观察家与董事会元老，以及 Acquired 线下盛会预告接着，David 补充了对他本期调研起到决定性帮助的几位重要人物： - **Terry Myerson**：现任 Truveta 公司 CEO，曾长期统管微软 Windows 与 Windows Phone 业务。 - **S. Somasegar（Soma）**：现为 Madrona 风险投资机构的常务董事，曾是微软 Server & Tools（服务器与开发工具）部门的奠基级传奇人物。 - **Mary Jo Foley**：专注报道微软超过 20 年的权威科技记者，被誉为业内最佳。她现任独立研究机构 *Directions on Microsoft*（类似专门针对微软的 Gartner）的总编辑，分享了极其通透的行业观察。 - **Dave Marquardt**：微软历史上任期长达 33 年的独立董事。他所代表的 TVI（Technology Venture Investors）是微软在 IPO 前引入的**唯一一家**外部风险投资机构，随后他又共同创立了 August Capital。除了比尔·盖茨本人，Dave 是微软董事会任期最长的成员。 - **Arvind**（来自 Worldly Partners）：为本期节目的深入研究提供了大量的辅助资料支持。在节目尾声，两人向听众推荐了相关的往期关联节目，包括《微软卷一（Microsoft Volume I）》、《AWS 深度史》以及《英伟达系列（NVIDIA Part 1）》（其中包含了从 GPU 视角看 PC 游戏与微软发展历史的交织节点）。最后，他们向听众发出了即将到来的 Chase Center 线下巨型活动的狂热邀请，透露马克·扎克伯格（Mark Zuckerberg）将作为重磅嘉宾登台，称这不仅是 Acquired 社区的世纪盛事，更是一场专属于商业与科技信徒的“奥马哈伯克希尔年会式”的集体狂欢。 ### [04:50:46](https://www.youtube.com/watch?v=6oER1bdQa_s&t=17446s) 节目的尾声与大通中心线下聚会预告节目的最后，两位主持人向听众热情推介即将到来的重磅线下活动。对于所有热爱商业、科技以及《Acquired》节目的朋友来说，这将是一场超乎想象的科技极客聚会。本鲍尔和戴维·罗森塔尔充满激情地向听众呼吁，如果你还在犹豫 2024 年 9 月 10 日该去往何方，那么唯一的标准答案就是来到旧金山的大通中心（Chase Center），与整个社区共同庆祝这一盛事。在饱满的情绪与期待中，主持人留下一句“It's going to be awesome”，向所有听众道别，标志着长达近五个小时的“微软第二卷：鲍尔默时代”波澜壮阔的商业史诗正式落下帷幕，背景随后响起了节目经典的标志性片尾曲与旋律。 - **活动时间**：2024 年 9 月 10 日 - **活动地点**：旧金山大通中心（Chase Center, San Francisco） - **活动定位**：面向科技极客与《Acquired》忠实听众的盛大线下聚会 ## Video description In 1999, Microsoft became the most valuable company in the world. And in 2019, Microsoft became the most valuable company in the world, *again*. But… what happened in the twenty years in between? The answer, as we discovered in our research, is probably not what you think. In this episode we explore and analyze the browser wars and the DOJ case, Windows XP through 8, Surface, Xbox, search, Yahoo!, Bing, the iPhone, Nokia, mobile, social, Facebook… and oh yeah, a little thing called Azure and the enterprise — which ended up becoming so big that no failures mattered. Tune in for Microsoft, Volume II. *Links:* - Bill Gurley on Android’s “Less Than Free” business model: https://abovethecrowd.com/2009/10/29/google-redefines-disruption-the-less-than-free-business-model/ - All episode sources: https://www.acquired.fm/episodes/microsoft-volume-ii#sources *Carve Outs:* - Meta Ray-Bans: https://www.meta.com/smart-glasses/shop-all/ - Ozlo Sleepbuds: https://ozlosleep.com - M3 Macbook Air: https://www.apple.com/macbook-air/ - Model Y: https://www.tesla.com/modely *More Acquired:* - Get email updates https://www.acquired.fm/email and vote on future episodes! - Join the Slack http://acquired.fm/slack - Check out the latest swag in the ACQ Merch Store https://www.acquired.fm/store! _Note: Acquired hosts and guests may hold assets discussed in this episode. This podcast is not investment advice, and is intended for informational and entertainment purposes only. You should do your own research and make your own independent decisions when considering any financial transactions._ [Source](https://www.youtube.com/watch?v=6oER1bdQa_s)

Source quotes

[17:41] Bill Gates: When I heard Steven talk about what was happening at Cornell, I began to take the internet quite seriously. [23:51] Marc Andreessen: You go to school, you do your research, you leave and then they try and sue your business. Had I known this would happen, I would have gone to Stanford. [35:15] Bill Gates: Perhaps you have already seen memos from me or others here about the importance of the internet. I have gone through several stages of increasing my views of its importance. Now I assign the internet the highest level of importance. In this memo I want to make clear that our focus on the... [35:48] Marc Andreessen: reduce Windows to a poorly debugged set of device drivers [88:31] Steve Ballmer: developers, developers, developers [90:03] Brad Smith: It's time to make peace. [123:24] Walt Disney: with animation goes the company. [127:54] Steve Ballmer: We only lost money. [153:36] Steve Ballmer: my Acquisitions, my mistakes—we just lost money on the bad ones. [153:36] Jeff Bezos: what makes your beer taste better [153:36] Steve Ballmer: It's never going to work at $500. [218:48] Ray Ozzie: The environment has changed yet again... businesses are increasingly considering what services-based economics of scale might do to help them reduce infrastructure costs or deploy Solutions as needed and on a subscription basis. [236:14] Ben: Oh this company has its head in the sand, this is ridiculous. What users want is we have a good version of Word, Excel, PowerPoint that people can run on their iPads and we've decided not to ship it to try and Advantage surface... [275:21] This was the biggest failure of the company during this period: they did not tell their story right. [277:14] we are a company who invents and wanders

Original transcript

[0:00] I'm a little horse today so hopefully we [0:01] don't have to do a lot of [0:05] talking good luck with [0:08] that all right let's do this who got the [0:12] truth is it is it is it got the truth [0:17] now is it you is it you is it you sit me [0:21] down say it straight another story on [0:24] the [0:26] way welcome to season 14 episode 6 the [0:30] season finale of acquired the podcast [0:32] about great companies and the stories [0:34] and playbooks behind them I'm Ben [0:36] Gilbert I'm David renthal and we are [0:39] your hosts well listeners here we are [0:42] Microsoft Volume 2 at long last after [0:46] the ancient history of volume 1 we now [0:49] get to the stuff that you grew up with [0:51] the internet Windows XP Xbox the browser [0:55] search and mobile and in this era [0:58] Microsoft had a lot lot of the right [1:00] ideas with a lot of the wrong timing and [1:03] execution on everything from the zoon to [1:06] Bing but despite that from 1995 where we [1:11] start our story to 2014 where we will [1:13] end this episode Microsoft grew their [1:16] annual revenue from 6 billion to 80 [1:19] billion they became a phenomenally [1:22] successful company and really cracked [1:24] the code on selling enterprise software [1:26] I began the research thinking our part [1:29] one episode would be about the rise and [1:31] this episode would be about the fall [1:33] cultural problems failed consumer [1:35] products antitrust but it's really not [1:38] that straightforward and after spending [1:40] months unpacking it all I actually don't [1:42] think that's the right framing anyway [1:44] and on Microsoft's 1998 antitrust suit [1:47] against the Department of Justice [1:49] everyone knows of this case but most [1:51] people really have no idea what actually [1:53] happened did Microsoft lose well not [1:57] really but the answer is nuanced finally [2:00] today we dive into it all oh and [2:04] listeners we have just one announcement [2:06] for you here today yes we told you [2:09] before that September 10th we are doing [2:10] the biggest thing in acquires history [2:12] and we're doing it in the city of San [2:14] Francisco we're doing a live acquired [2:16] show at the chase Center which is the [2:19] brand new basketball arena here in San [2:21] Francisco where the Warriors play we're [2:23] putting it on with our good friends at [2:25] JP Morgan payments and as you can [2:27] imagine they know a few people at the [2:28] chase Center yeah it will be a night to [2:30] remember with a few different phases of [2:32] the evening there's going to be lots of [2:34] opportunities to meet other required [2:35] listeners from around the world and a [2:38] big show like this deserves a big [2:41] special guest and that special guest is [2:44] the one and only Mark Zuckerberg so in [2:47] addition to being the central figure in [2:49] some of the greatest Acquisitions of all [2:51] time that we have covered right here on [2:52] acquired Mark and meta are also playing [2:55] a big role in defining the next decade [2:57] of computing with AI too so it's shaping [3:00] up to be a total blast we really hope [3:02] you can join us yeah tickets will be [3:04] available soon and you can sign up at [3:06] acquired. fmsf to get emailed as soon as [3:09] they go live we're pumped we'll see you [3:12] there this show is not investment advice [3:14] David and I may have investments in the [3:16] companies that we discuss and so do all [3:18] of you if you own index funds and this [3:19] show is for informational and [3:21] entertainment purposes only okay David [3:25] the middle chapter of Microsoft the [3:27] middle chapter indeed and and boy is [3:30] there a lot to discuss so Ben you [3:33] covered this in your intro but I think [3:36] everybody kind of knows the narrative [3:38] about what happened to Microsoft between [3:40] call it 1995 and 2014 when SAA took over [3:44] there's even a quote from SAA himself in [3:48] the very first paragraph of the book [3:50] that he wrote in 2017 called Hit refresh [3:53] which I mean that title kind of gives it [3:55] away right there he writes I joined [3:57] Microsoft in 1992 because I wanted to [4:00] work for a company filled with people [4:02] who believed they were on a mission to [4:04] change the world but after years of [4:06] outdistancing all our competitors [4:09] something was changing and not for the [4:11] better Innovation was being replaced by [4:13] bureaucracy teamwork was being replaced [4:16] by internal politics and we were falling [4:19] behind and then he references the famous [4:22] gun pointing org chart by cartoonist and [4:25] software engineer Manu coret that [4:27] probably listeners many of you are [4:28] familiar with we will link to that in [4:30] the show notes and you can sum this kind [4:32] of whole narrative up as Microsoft was [4:34] winning and then it sucked for a long [4:36] time and then it is now winning again [4:38] and that's all thanks to [4:39] SAA and the question we sort of asked as [4:42] we were doing our research was is this [4:44] true and what we ended up learning from [4:47] the literally dozens and dozens of [4:49] people that we talked to surprised us a [4:51] lot and I think we'll probably surprise [4:52] listeners too yeah David you're burying [4:54] the lead here a little bit we talked to [4:56] probably four to five times as many [4:58] people as the next highest episode I'm [5:02] looking at our little thank you list [5:03] it's like 20 something people long all [5:06] right so on the last episode we left off [5:08] with the Ben as you put it unabashed [5:11] celebration of software that was the [5:14] Windows 95 consumer launch in August of [5:18] 1995 and it was perfect it had [5:22] everything had jeno had the Rolling [5:25] Stones it had the start button or [5:29] actually it had almost everything there [5:32] was one thing that was missing from [5:35] Windows 95 at launch that if you were a [5:39] consumer user of Technology of software [5:44] of products of operating systems maybe [5:47] you kind of wanted to have and that was [5:50] an internet browser yes it's so funny [5:54] cuz we sort of intentionally left all [5:56] the internet components out of Windows [5:59] 95 [6:00] in the previous episode because once you [6:02] start talking about the internet you're [6:03] really talking about the next chapter of [6:04] Microsoft and you can't help but dive [6:06] into it all but in retrospect the thing [6:09] that mattered about Windows 95 all these [6:11] years later is that's the platform that [6:14] everyone started using the internet on [6:16] and everything that we talked about in [6:17] the last episode yeah it's all important [6:19] but it's not nearly as important as it [6:21] being the internet operating system so [6:23] how did this come to be at the time [6:25] things were changing so fast there was [6:27] this phrase called internet time things [6:29] happened in weeks versus years but if [6:33] you rewind just a little bit back to [6:35] like [6:37] 9293 94 even into early 95 going online [6:42] for consumers meant using a service like [6:45] Compu serve or prodigy or of course the [6:47] big one AOL and these Services were not [6:52] what we think of today as the internet [6:54] but they were more like walled Gardens [6:57] with proprietary services that were [6:59] bundled with access via dialup modems [7:03] yeah for consumers it was kind of a [7:05] similar experience you could get content [7:07] on your computer but the main difference [7:11] was how to put content on that Network [7:13] it wasn't like anyone could just plug in [7:15] a server and then boom you have a [7:17] website it was like you had to have some [7:19] negotiating power and know someone at [7:21] AOL to go do a deal to get your content [7:24] on their platform yep I think the best [7:27] way to sum all this up is do you know [7:29] who owned the compu serve service at the [7:33] time no but I know it was a Columbus [7:35] based company oh interesting it was [7:38] owned by H&R Block the tax prep compy [7:40] really yeah whoa crazy that's what [7:44] online was like just a few years or [7:47] months before the Windows 95 launch so [7:50] Microsoft of course as you know [7:52] inheritor of the earth and all things [7:54] technology they want to play in this [7:55] online services Arena too so in 1993 [7:58] they start sniffing around AOL and see [8:01] if maybe Microsoft could acquire AOL [8:04] Steve case the founder of AOL isn't [8:06] interested in selling but there's this [8:08] whole thing where Paul Allen goes off by [8:10] himself and he buys a large stake and [8:12] that creates all sorts of headaches [8:14] because Microsoft is like well if we [8:15] can't buy them we're going to compete [8:17] with them so they start an internal [8:19] project called project Marvel to build [8:21] their own online service that becomes [8:23] MSN so there's a little slide of hand [8:26] that you just did there you said it [8:27] becomes MSN Marvel when it initially was [8:31] conceived was a proprietary online [8:33] service eventually when that completely [8:36] failed which you're about to get to they [8:38] repurpose the name MSN for their [8:41] internet-based media property U complete [8:44] shift and strategy at the same time many [8:47] people in technology especially at [8:48] Microsoft itself and lots and lots of [8:51] investors on Wall Street believe that [8:53] these wal Garden online services were [8:56] just temporary they were just a bridge [8:59] to a more utopian networked consumer [9:02] culture and economy that they called the [9:05] information Super Highway and the [9:07] specific vision of how this information [9:10] Super Highway Utopia was going to work [9:12] was interactive television all mediated [9:16] by the pay television providers so like [9:18] the cable and satellite companies out [9:20] there you know the Comcast the charters [9:22] the Time Warner Cables the direct TVs on [9:24] the satellite side these were going to [9:27] be the big consumer technology companies [9:31] and this wasn't crazy this actually made [9:32] a ton of sense because television and in [9:35] particular cable television at the time [9:37] was the primary existing consumer medium [9:41] the internet was not a thing well think [9:43] about the number of things required to [9:45] create some sort of networked [9:48] entertainment interactive thing you [9:50] would need screens you would need some [9:54] way to control those screens to create a [9:55] feedback mechanism you would need [9:57] content you would need infrastructure [10:00] connecting people's homes all of those [10:03] already existed by the cable companies [10:06] and their endpoints the televisions and [10:08] if you pitched me on the idea that [10:11] actually everyone's going to go buy a [10:12] brand new device like a PC like a [10:15] computer and we're going to have a [10:17] different set of wires that actually [10:20] bring all of that to the home or maybe [10:22] we'll repurpose some of the same wires [10:24] but gosh we need to like bring in new [10:25] networking equipment everywhere along [10:27] the way oh and there's going to be [10:29] completely different content companies [10:30] that figure out how to create the [10:32] content for their it's like all of that [10:34] falls flat of course you're going to use [10:35] all the existing infrastructure and [10:37] content you're not going to bank on [10:38] standing it all up new from Whole cloth [10:41] totally and Microsoft just like they had [10:43] done in entering the PC software Market [10:45] in partnership with IBM they're going to [10:47] partner with these big consumer cable [10:51] companies and so starting in the summer [10:52] of 1993 there all these rumors flying [10:55] around that Microsoft is working on a [10:57] big JV with the cable companies [10:59] dubbed cable soft yep you can't make [11:03] this up and the idea is that Ben like [11:05] you're saying the cable companies will [11:07] control the pipes and the customer [11:09] relationships and probably a lot of the [11:11] content Microsoft will write the [11:13] software both for the set toop boxes in [11:16] consumer homes and for the servers on [11:19] the back end and this software project [11:21] is code named tiger and then there's a [11:23] third company a third piece of this sort [11:26] of Unholy Alliance for the information [11:28] Super Highway and that was a company [11:30] called silicon Graphics that would make [11:34] all the hardware cable companies are [11:37] going to make the hardware themselves [11:38] you're going to need pretty powerful [11:39] Hardware here both at the home and on [11:41] the server side and SGI as silicon [11:44] Graphics was referred to was legendary [11:46] they are the graphics company that [11:48] enabled the CGI in Jurassic Park and of [11:51] course their founder and chairman was [11:54] legendary in Silicon Valley W Jim Clark [11:58] put a pin in that name so pinned so Wall [12:01] Street of course like nuts over all this [12:03] you know the hype is out of control it's [12:05] a trillion dollar opportunity there's [12:07] all these spy shots of Bill meeting with [12:10] John Malone at TCI and Gerald levid at [12:12] Time Warner and Bill starts spending [12:15] time with Michael Ovitz talking about [12:17] how Microsoft can get in on the content [12:19] game too either through the MSN project [12:21] or through other things they're going to [12:22] do this leads to MSNBC the cable network [12:27] that people are probably familiar with [12:29] well this is so interesting because [12:30] we're talking about this general idea of [12:32] interactive Computing involving other [12:35] people and Microsoft so far has two [12:36] initiatives Marvel and the information [12:39] Super Highway neither of which are the [12:41] internet or the web browser yes correct [12:44] you're already getting this picture of [12:46] Microsoft's business strategy which is [12:48] until we know exactly what the future [12:50] looks like start placing bets that [12:52] approximate so that we're sort of in the [12:54] mix even though we don't know exactly [12:57] what the future is which we talked about [12:59] in part one had always worked so well [13:02] for the company yep and it's going to [13:04] work really well here too so Bill [13:06] actually decides at this point that he [13:08] needs to write a book for the public to [13:11] evangelize this information Super [13:14] Highway thing kind of embarrassingly [13:16] given how long the book World takes to [13:18] you know actually publish a book it [13:20] doesn't come out until November 1995 [13:23] after the Netscape IPO has already [13:25] happened and Windows 95 has shipped but [13:28] in this book called The Road head I have [13:30] two copies of it here on my desk the [13:31] hard cover copy and the soft cover copy [13:34] which was revised and came out in [13:36] 1996 the hard cover copy is all about [13:39] the information super highway or as Bill [13:41] likes to call it information at your [13:43] fingertips and then when the soft cover [13:45] version comes out later basically they [13:47] control left every instance of [13:50] information Super Highway and replaced [13:51] it with the internet and the web browser [13:54] yeah this is one of these moments on an [13:56] acquired episode where we have just a [13:58] delightful [13:59] concrete illustration of this year it [14:03] was unclear the next year it was [14:05] extremely clear and David look up in the [14:07] indexes of both of those books the [14:09] number of references to the internet in [14:12] the hard cover version there are three [14:15] portions of the book where it is [14:17] discussed in the soft cover version the [14:20] index for the internet takes up an [14:22] entire [14:24] page there needs to be an index for all [14:26] the sub indexes of the internet in the [14:28] soft C version amazing so the hard cover [14:32] version is the state of play in January [14:35] 1994 when a young Windows networking [14:38] engineer named James Allard or Jay as he [14:42] goes by writes a memo to Bill Gates and [14:45] to the senior leadership at Microsoft [14:48] entitled Windows the next killer [14:51] application for the internet and in this [14:55] memo he points to a new piece of [14:57] software coming com out of the national [14:59] Center for supercomputing applications [15:01] at the University of Illinois that is [15:03] spreading like wildfire and appears to [15:06] be written by some like kid programmer [15:10] there by the name of Mark Andre in his [15:13] free time it's not even like his real [15:15] job yes and it is called Mosaic and in [15:19] this memo Jay argues that the internet [15:21] and this software instantiation of it in [15:24] the Mosaic web [15:25] browser kind of looks like it is going [15:28] to become exponential phenomenon given [15:30] the rate at which it is growing and that [15:33] it represents an enormous opportunity [15:35] for Microsoft to quote Embrace and [15:39] extend the internet into Windows itself [15:43] and this is the origin of the Embrace [15:46] and extend Mantra and the exact words he [15:49] uses are Embrace extend innovate in [15:53] popular press and public opinion of [15:55] Microsoft that would of course get [15:56] changed to embrace extend extinguish [15:59] by their effectively competitors and [16:01] political enemies but the Embrace and [16:03] extend thing is actually a brilliant [16:05] business strategy there's already a [16:07] whole bunch of people who love this [16:08] thing we want to embrace that new [16:10] Behavior there's sort of no product [16:11] Market fit risk because we can clearly [16:14] already see it happening people want to [16:15] use this browser thing to access [16:17] hypertext on the internet we're going to [16:19] embrace that and we're going to figure [16:20] out a way to work it into our business [16:23] model to extend the functionality in a [16:25] way that we can make money on right the [16:26] business model is we sell Windows [16:28] through o and businesses and the like [16:30] and to Consumers and we can just bake [16:32] this into it honestly it's pretty [16:34] incredible that Jay lays out the whole [16:37] winning strategy for Microsoft and the [16:40] internet here in January 1994 this is a [16:43] few months before Netscape is even [16:45] founded yes yeah Netscape as a company [16:47] does not exist yet there's just the [16:48] Mosaic web browser at the University of [16:50] Illinois yep then and this sort of [16:54] exponential growth theme the very next [16:56] month in February 1994 Bill's technical [17:00] assistant/ Shadow which is a legendary [17:03] role at Microsoft now exists at Amazon [17:05] 2o a man named Steven sinowski goes on a [17:08] recruiting trip to his Alma moer at [17:10] Cornell University and while he's there [17:12] there's a big snowstorm he gets stuck on [17:14] campus he has to stay on campus for a [17:16] few extra days it's very Cornell story [17:18] yeah most Cornell thing ever the most [17:21] itha story ever he notices that all [17:24] these kids especially when the campus is [17:26] snowed in they're all using the internet [17:30] and he knows what the internet is you [17:31] know it was an academic project for [17:32] years you know he was an academic guy [17:34] before getting into commercial software [17:36] and joining Microsoft but it was this [17:38] way for scientists to basically trade [17:41] research and you're starting to get some [17:44] cool entertainment use cases but there's [17:47] certainly no business or business [17:49] interest or commerci it's all just like [17:51] the way that academics communicate with [17:53] each other and this is what absolutely [17:55] floes Stephen he's like I remember the [17:58] internet as what you're saying Ben and [18:01] now I'm here on campus and all these [18:03] kids are using it for flirting [18:05] registering for classes messaging each [18:08] other sending email that has nothing to [18:10] do with papers or work or school or [18:12] academics or anything he gets so excited [18:16] that he writes another memo to Bill and [18:18] the leadership team entitled Cornell is [18:21] wired exclamation point it's so funny [18:24] Microsoft history is told through a [18:26] series of memos every Milestone is [18:29] some executive publishing a companywide [18:32] memo well it's so funny cuz some of [18:34] these memos definitely were like [18:36] internal memos for exactly what you say [18:38] and some of them were like written for [18:41] publication to the Press yep and bill [18:44] has great quote when I heard Steven talk [18:46] about what was happening at Cornell I [18:48] began to take the internet quite [18:50] seriously so Steven and Bill organize an [18:54] internet offsite quote unquote with all [18:56] the top execs with Jay bill and [18:59] everybody who's investigating this [19:01] internet thing and it takes place on [19:03] April 5th 1994 which is the very next [19:08] day after Netscape was incorporated on [19:11] April 4th amazing and at this offsite [19:15] Bill totally gets religion that the [19:19] internet as Jay said in his initial memo [19:21] is actually an exponential phenomenon [19:24] and as Bill puts it to the team gathered [19:26] there and then the whole company later [19:28] it is a core Microsoft company value [19:31] that exponential phenomena cannot be [19:34] ignored oh wow I had no idea that was [19:37] kind of the impetus of him taking it [19:39] seriously I mean think back to [19:41] everything we talked about in the last [19:42] episode the whole concept of Microsoft [19:44] is founded on the idea that mors law is [19:47] a thing and therefore we can develop [19:51] software that people have never dreamed [19:53] of that in just a few years will be [19:55] usable So speaking of Netscape being [19:58] Incorporated the day before remember I [20:00] said to put a pin in the name Jim Clark [20:02] of course many listeners already know [20:04] where we're going here Jim Clark [20:06] legendary founder of SGI Silicon Valley [20:08] Legend well a couple months before that [20:11] in February 94 it's crazy how fast all [20:15] of this happens it's just insane Jim is [20:18] still at SGI he's really frustrated with [20:21] the board and the company though for not [20:24] pushing even harder on this information [20:26] Super Highway opportunity so what does [20:28] he do he resigns from the company kind [20:31] of like in protest the company he [20:33] founded wow and on his very last day in [20:36] February 94 at SGI he cold emails the [20:41] kid in Illinois Mark Andre and so you [20:44] get the opportunity to team up with an [20:46] industry Legend uh of course yeah so Jim [20:49] in this email writes you know I'm [20:51] impressed with Mosaic and like clearly [20:53] this seems to be getting adoption um you [20:55] know if there's any way that you and I [20:57] Mark might be able to collaborate that [20:59] would be quote of interest to me so the [21:02] two of them get together and then they [21:03] found this company on April 4th 1994 [21:06] called electric media and the initial [21:09] goal of electric media soon to be [21:12] Netscape is that oh Mark is this hot [21:15] shot programmer clearly the information [21:17] super highway is what this web is going [21:19] to turn into we're going to do what SGI [21:22] was supposed to do we're going to make [21:25] set toop boxes for the information Super [21:27] Highway yes in retrospect you got to be [21:31] looking at them thinking how dense are [21:33] you Mark andr is the person in the world [21:35] who understands what a crazy exponential [21:39] phenomena the internet is the web is [21:41] what it can be I mean Mark had I think [21:44] by this point already put the image tag [21:47] into HTML so they can now send images [21:50] that render in browsers and when Jim [21:52] Clark emails him they decide to go do [21:55] the information superhighway and not to [21:57] do the internet yeah this is amazing so [22:00] the way they're going to do this as a [22:02] startup this is so great one of the [22:05] other big things that SGI had done [22:08] besides building the graphics [22:09] workstations that Hollywood ran on and [22:11] did Jurassic Park and all that was they [22:14] were Nintendo's technology partner for [22:18] the N64 yep they made the graphics [22:21] engine for the N64 and so Jim has this [22:24] relationship with Nintendo the N64 is [22:26] going to be coming out it's going to be [22:27] this amazing [22:29] box in the living room attached to TVs [22:31] and consumers homes they're going to [22:34] team up with Nintendo and turn the [22:37] N64 into an information superhighway box [22:41] meanwhile it's hilarious that we keep [22:42] talking about the information Super [22:43] Highway because it never happened there [22:45] were these little tests done with cable [22:48] companies that would wire up 300 houses [22:50] or something but it never happened [22:52] anywhere at any sort of scale and so [22:54] when you're listening to this and you [22:56] keep trying to figure out like sorry [22:57] what exactly was the information Super [22:59] Highway and like what did it look like [23:01] nobody knows because it never happened [23:03] right it's such a classic case of way [23:05] way way too many cooks in the kitchen [23:07] and just total slide wear and the fact [23:10] that even Mark himself didn't pound the [23:13] table for no the internet's going to be [23:14] the thing that really shows you how the [23:17] human brain is not wired to understand [23:21] compounding theoretically this network [23:23] should continue to get more nodes the [23:25] technology should evolve little by [23:27] little More's is happening on the [23:29] compute side you know there's some [23:31] reason to think that bandwidth is going [23:32] to be available to homes in the same way [23:34] it's available to like universities and [23:36] companies but still it just wasn't [23:39] obvious enough to continue down that [23:41] path it was almost like great I've made [23:42] a name for myself doing this toy thing [23:44] that probably isn't the future and so [23:46] now we'll go do the big boy stuff [23:47] because that's what all the experts are [23:48] saying totally so shortly after this by [23:52] late spring 1984 Mosaic now has a [23:55] million active users clearly Bill Gates [23:58] is paying attention here so shortly [24:02] after all this Jim Clark and Mark [24:04] andrees say wait a minute let's just go [24:07] do this Mosaic thing they scrap the N64 [24:10] information Super Highway they go raise [24:13] money from Kleiner Perkins John door [24:15] invests joins the board most legendary [24:18] investor of the time and by October [24:23] 1994 the newly christened Mosaic [24:26] Communications Corporation [24:28] post the first version of its browser [24:32] Mosaic navigator for free to download on [24:36] the web and their business plan is that [24:38] they're going to give away the consumer [24:40] browser for free and they're going to [24:42] charge companies for Server software so [24:45] if you want to host a website you know [24:47] you're a corporation or whomever you [24:49] need server software to do that they're [24:51] going to charge companies for the server [24:53] software great and listeners you should [24:55] be paying attention to something David [24:57] just said there he said it's the Mosaic [24:59] Corporation and then you said it's [25:01] Mosaic Navigator even though it's called [25:05] Mosaic because Mark andreon wanted to [25:07] draft off the success of the previous [25:09] project he had done called Mosaic this [25:11] is completely new code they founded a [25:13] company they started writing code from [25:15] scratch they had the experience of [25:17] writing Mosaic the thing owned by the [25:20] university and the NCSA before this is a [25:23] new thing called Mosaic that does the [25:25] same thing architected for commercial [25:27] use yes exactly Ben so meanwhile back at [25:31] the University of Illinois even though [25:34] they're a academic and government [25:36] institution here they realize that [25:38] they've got something valuable in the [25:40] original Mosaic to pile a code millions [25:42] of users Mark doesn't work there anymore [25:45] but seems to be working so they license [25:49] their Mosaic the original one to a local [25:53] company called spy glass so the state of [25:56] play is you've got [25:58] the old Mosaic which spy glass has the [26:01] right to license for commercial use you [26:03] have a new thing that will become [26:05] Netscape called Mosaic that is totally [26:07] separate code and Mosaic Mark Andre's [26:10] new Mosaic keeps trying to go do deals [26:14] like sell their server software and [26:16] every time they find a customer spy [26:18] glass they threaten to Sue and they [26:21] basically blow up the deal because they [26:23] keep calling each customer and saying [26:24] yeah we're going to sue yep yep and so [26:28] this is obviously very frustrating and [26:31] technically illegal Mark Andre's Mosaic [26:34] realizes that this is going to be a [26:36] existential problem for them unless they [26:38] do something about it they actually Sue [26:41] spy glass you guys got to stop so there [26:43] is a settlement well the net of all this [26:46] is that Mark and Jim's company let's [26:50] call it that changes its name in the [26:53] fall of 1994 to Netscape and Mark had a [26:57] typically great mark quote about this to [26:59] the Press at the time you go to schol [27:02] you do your research you leave and then [27:04] they try and your business had I [27:06] known this would happen I would have [27:07] gone to [27:10] Stamford which of course is apocryphal [27:13] because he had no means of going to [27:15] Stanford at the time he lived in the [27:16] midwest he was going to go to a state [27:17] school no one recognized his genius at [27:20] the time so yeah yeah yeah totally it's [27:22] also the most Mark ENT quote ever which [27:24] is awesome y anyway okay meanwhile the [27:28] online services the comp you serves you [27:30] know Prodigy AOL Etc they're not blind [27:34] they see that the internet and the web [27:36] is also becoming a thing they want to go [27:40] license a web browser and incorporate it [27:43] into their platform so I think it was [27:45] late 93 or early 94 but this was a [27:47] seminal moment where the aols of the [27:50] world interconnected with the internet [27:52] and now you could not just navigate the [27:54] proprietary services but also surf the [27:56] open web so [27:58] Netscape isn't interested in licensing [28:00] because they have their own business [28:01] model selling web server software and [28:03] they want to allow free downloads of the [28:05] Navigator client spy glass though they [28:08] start licensing that original Mosaic and [28:12] they start doing deals with the online [28:14] service providers and a small startup [28:17] company called booklink goes and codes [28:20] up another browser that they start [28:21] licensing to companies as well so Bill [28:25] Gates and Steven sinowski they go and [28:27] meet with booklink in May of 1994 so [28:30] coming right off of this internet [28:31] Retreat we're going to make this a core [28:34] part of Microsoft and a core part of [28:36] Windows and they're interested in [28:39] licensing booklink as well they start [28:41] negotiating they're talking about a call [28:43] it $2 million license deal and all of a [28:46] sudden AOL comes in and buys the whole [28:49] company of booklink for $30 [28:52] million so this now leaves Microsoft [28:56] without a browser and there are [28:59] basically only three real browsers on [29:01] the market there's booklink that AOL [29:02] just bought there's Netscape Navigator [29:06] Netscape Navigator which is not [29:08] available to license and then there's [29:11] Mosaic and spy glass so Microsoft goes [29:16] to who else spy glass they license the [29:20] source code for Mark Andre's original [29:24] Mosaic browser from Spy class software [29:29] for $2 million and that code base [29:32] becomes the base upon which Microsoft [29:37] builds Internet Explorer well David I'm [29:40] glad you took the bait I am here to tell [29:43] you that that is the public narrative [29:48] uh and very close to the truth but [29:50] there's some more Nuance here you ready [29:52] to hear it well hey if you click the [29:53] about menu in the early versions of [29:55] Internet Explorer oh yeah a text box [29:58] pops up that says based on NCSA Mosaic [30:01] distributed under a licensing agreement [30:03] with spy glass Inc yes all of that [30:06] technically true it is just not quite as [30:09] meaningful as you think it is so as with [30:12] all of these things it's not just like [30:15] Bill Gates and Steven sinowski are [30:17] having a think and the rest of Microsoft [30:19] is sitting around waiting for the think [30:20] to finish and then an edict comes down [30:23] and then they go and do the work there [30:25] are a lot of people with a lot of ideas [30:27] working on a lot ofu in parallel and [30:29] that is why Microsoft's history is so [30:31] delightfully messy is there's a zillion [30:33] initiatives going on and it's never [30:35] clear if your thing is going to become [30:37] the next company strategy or not so here [30:40] is a slightly different version of this [30:42] history with different players and I [30:45] want to underscore it for one big reason [30:48] it will come up later in antitrust okay [30:52] so some of the windows95 team in late 94 [30:56] led by Thomas Reen is p off before it [30:59] ships to start thinking about what [31:01] should we do after Windows 95 ships what [31:03] would the next Marquee Investments be [31:05] for what at the time they're calling [31:07] Windows 97 which of course there was [31:09] never a Windows 97 so the group's [31:12] opinion is all internet all the time you [31:15] know how could the next iteration of [31:17] Windows be extremely internet native in [31:19] a very embedded way and there's tons of [31:22] proposals in this little group there's [31:23] virtual meeting software think Zoom type [31:26] things there's an email client [31:28] specifically built for the internet [31:30] rather than for your company's corporate [31:31] Network which at the time was novel then [31:33] there's of course a browser but the big [31:36] Vision was what if the whole Windows [31:39] shell is a browser every visual thing [31:43] that you interact within Windows what if [31:45] that actually was like a HTML rendered [31:48] server [31:49] communicating online thing and the team [31:53] technically kind of looked at it this [31:55] way we should build HTTP Direct into the [31:58] operating system since it was just [32:00] another protocol on top of the tcpip [32:03] protocol that the internet is based on [32:06] we should provide reusable UI component [32:10] to any application that wants to display [32:12] HTML that's a good engineering building [32:14] block to build is this HTML renderer [32:17] that any application can sort of frame [32:19] in and use so of course Microsoft the [32:22] strategy here is we will develop a [32:24] browser application that used the [32:26] building block that others could also [32:28] used to render HTML so they actually go [32:30] to Netscape and say hey we have this [32:34] great HTTP stack we have the HTML engine [32:38] we have these wrappers to go around it [32:40] instead of rewriting all of it just use [32:41] our off-the-shelf code that we intend to [32:43] ship with Windows famously you know [32:46] Netscape did not do that and so iie [32:49] Internet Explorer actually ends up being [32:50] kind of the only application that used [32:53] all these windows components and once [32:56] they got going on the browser they [32:57] convinced the the windows leadership [32:58] that actually we can do this fast we [33:00] should get this done as a part of [33:02] Windows 95 not wait for the next big [33:04] release we're going to get to this in a [33:06] minute but when Windows 95 launched it [33:08] had either at launch or very shortly [33:10] thereafter it was called the Plus Pack [33:12] yes and Internet Explorer was available [33:14] as part of the Plus Pack yes so anyway [33:17] how does NCSA Mosaic and spy glass come [33:19] into this well the Nuance is spy glass [33:22] had actually massively changed the [33:25] Mosaic code they were trying to create [33:27] the spy glass browser that was sort of [33:29] based on this NCSA codebase but it [33:32] wasn't very good and so that is what [33:36] Microsoft was able to get their hands on [33:38] they could not license the original NCSA [33:41] version that was gone or at least not [33:43] available for license and so they sort [33:45] of tried backing out a lot of the spy [33:48] glass stuff ultimately it wasn't that [33:51] helpful in creating Internet Explorer [33:52] and they spent just as much time trying [33:54] to undo a lot of it and then build the [33:57] that Explorer stuff on top so ultimately [34:01] did it actually accelerate their path to [34:02] Market and was it actually Mark Andre's [34:05] code some of it was in there but you [34:08] know it's not like they grabbed it off [34:09] the shelf and now it's IE oh it makes [34:11] for such a good story though it sounds [34:14] like reality is a lot like the Dos [34:19] acquisition yes Microsoft bought qos [34:22] quick and dirty operating system from [34:24] Seattle computer products was that the [34:26] same thing as mic moft [34:29] dos sort of a lot of work went into it [34:32] after the deal yeah as you would expect [34:35] same thing here but it is definitely [34:37] true that if you click that about box in [34:39] the early versions of Internet Explorer [34:41] holding to it it's just so delicious it [34:43] is delicious and the two big takeaways [34:45] here at least from this additional [34:47] version of the story is one what they [34:50] actually wanted to do was make Windows [34:53] Web enabled in a really deep integrated [34:55] way not just have this one little [34:57] application called a browser and [34:59] technically there was a lot of [35:00] co-mingling there a lot of what became [35:03] the code underpinning Internet Explorer [35:05] was actually Windows code implemented in [35:08] Windows operating system to do these [35:09] protocols and two still a lot of work to [35:11] make iie after the deal so this brings [35:15] us now to the launch preparations for [35:17] Windows 95 and in the spring leading up [35:20] to all this bill writes another memo [35:23] this one intended for publication so to [35:26] speak that is the famous internet Title [35:28] Wave memo I just want to do a big quote [35:31] from it here perhaps you have already [35:33] seen memos from me or others here about [35:36] the importance of the internet I have [35:37] gone through several stages of [35:39] increasing my views of its importance [35:41] now I assign the internet the highest [35:44] level of importance in this memo I want [35:47] to make clear that our focus on the [35:48] Internet is crucial to every part of our [35:51] business the internet is the most [35:53] important single development to come [35:56] along since the IBM PC was introduced in [35:58] 1981 it is even more important than the [36:01] arrival of the graphical user interface [36:04] can't get any more clear than that very [36:06] clear so that brings us to the August 95 [36:09] Windows 995 launch scheduled for the [36:13] 24th on August 9th couple weeks [36:16] beforehand Netscape goes public with a [36:19] market capitalization of3 [36:22] billion massive IPO massive I mean this [36:25] is like 1995 we're talking about [36:27] Netscape we should say goes from 1 [36:30] million to 15 million users in one year [36:34] I mean just instant product Market fit [36:36] it was so clear that people wanted to [36:38] browse the web a lot of the time in [36:40] technology in this ecosystem we're [36:41] always looking around like ah is that [36:43] going to become a thing is that going to [36:44] become a thing that was from 1994 onward [36:47] never a question about the internet [36:49] never yeah in the IPO press [36:53] cycle Mark andreon is quoted as saying [36:56] that quote Netscape will soon reduce [36:58] Windows to a poorly debugged set of [37:00] device [37:03] drivers it's such a good quote and [37:06] there's so much behind it too if you [37:08] really dwell in that quote what does it [37:10] mean if one of the things he's saying is [37:14] Windows is a platform upon which [37:17] independent software vendors write [37:19] applications so Windows is the way that [37:23] currently people write software for [37:26] businesses and consumers to use use and [37:28] if we are going to reduce Windows to a [37:30] poorly debugged set of device drivers [37:32] what I'm implying is these crappy static [37:34] web pages that get served right now that [37:37] is merely a step on our journey to [37:40] enabling Rich web applications think [37:43] JavaScript CSS eventually you know Java [37:46] and Flash the web will be a way that [37:50] developers write their applications [37:52] that's right there implicit in the quote [37:55] and so when they're saying we're going [37:56] to reduce windows blah blah blah it's [37:58] saying okay Windows has all this stuff [38:00] right now for developers but essentially [38:03] you're going to use Windows or any [38:05] operating system just to boot it up [38:06] connect to all your peripherals your [38:08] screen and your mouse your keyboard and [38:10] everything and you'll open your browser [38:11] and you'll do everything through the [38:12] browser and that scared the hell out of [38:16] Microsoft not specifically this quote [38:19] but Microsoft had come to the same [38:21] conclusion too of oh my God if the web [38:24] becomes the platform of the future all [38:27] the reasons why we have all this [38:29] incredible business you know people [38:32] feeling the need to use our operating [38:34] system to be able to get access to their [38:36] favorite software and for developers to [38:38] build applications on a platform to get [38:39] access to the users that could go away [38:42] and in the same memo that you were [38:43] quoting earlier the internet Title Wave [38:46] Bill Gates famously says and when I say [38:48] famously it's because the Department of [38:49] Justice later grabbed this quote and [38:52] used it as an exhibit Bill writes a new [38:55] competitor born on the internet is [38:57] Netscape their browser is dominant with [39:00] a 70% usage share allowing them to [39:03] determine which network extensions will [39:06] catch on they are pursuing a [39:07] multi-platform strategy where they move [39:10] the key API the application programming [39:12] interface into the client to commoditize [39:15] the underlying operating system I mean [39:18] they got it immediately the web is an [39:21] application platform that completely [39:23] reduces our value you can see why it was [39:26] so important to Microsoft to beat [39:30] Netscape to bring the internet in the [39:33] form of Internet Explorer into windows [39:36] and have Windows maintain its role as [39:38] the dominant platform so all the stuff [39:40] we'll cut off their Air Supply you know [39:42] it was existential and how amazing is [39:46] this it's an application platform of the [39:48] future that is distributed as a Windows [39:51] app I mean Windows had huge market share [39:55] at this point I don't know 80 90% [39:57] eventually over 90% market share the way [39:59] that Netscape could get to Consumers was [40:03] because Microsoft had all these [40:04] computers out there running Windows it [40:05] was like this ultimate trojan horse that [40:07] they could build the platform of the [40:09] future through Microsoft yep so Windows [40:14] 95 launches couple weeks after the [40:16] Netscape IPO Internet Explorer is not [40:20] baked in at least not in the retail box [40:22] version you can buy it for $50 as part [40:26] of the Plus pack that I was referencing [40:29] before and install that and add it into [40:32] windows and Microsoft will make money on [40:34] the sale of that software but that of [40:37] course does nothing to make a dent in [40:39] the free version of Netscape Navigator [40:41] that is out there right if Microsoft's [40:43] goal is to cut off the air supply David [40:45] as you already quoted of Netscape the [40:48] goal is ubiquity instantly and we don't [40:50] care about making money we just need to [40:52] get this thing out so the internet [40:53] doesn't kneecap our business and we can [40:55] sort of embrace and control it or [40:57] perhaps Embrace and extend it netscape's [40:59] run continues the Netscape stock triples [41:04] over September October November net [41:07] skape is now A10 billion public company [41:11] insane and I don't think making very [41:13] much money on their server software yet [41:15] all the market cap creation is [41:17] attributable to people believing they [41:19] have the dominant platform of the future [41:21] and not based on their current [41:22] financials yeah basically all of the [41:25] hype train that had been behind the [41:26] information Super Highway has now [41:28] completely ported over to Netscape [41:31] that's true what's our tracker for the [41:33] internet Netscape everybody Pile in yeah [41:35] I could make an analogy to today but I'm [41:37] gonna spare us all make this episode [41:39] Timeless I'm gonna make the episode [41:40] Timeless okay and then on December 7th [41:45] 1995 Bill Gates announces that Internet [41:48] Explorer is now free and it will be [41:51] bundled in with every single copy of [41:54] Windows 95 going forward and on that day [41:57] Netscape stock drops by about a third [42:00] and never recovers that was the high [42:03] watermark for Netscape it's over after [42:07] that yep and for good reason I mean [42:10] there's a very difficult to learn lesson [42:13] but you learn it once you never forget [42:14] it if your [42:16] distribution decides to compete with you [42:19] and decides to make that a priority your [42:22] business is over in a minute yep and [42:24] that's exactly what happened I mean this [42:25] is now the of Internet Explorer it [42:28] doesn't happen overnight but it's [42:30] inevitable by the end of the next year [42:33] in [42:34] 1996 Microsoft has now done deals with [42:37] AOL compy serve and Prodigy all the old [42:40] online services to ditch whatever [42:42] browsers they were using and bundle in [42:44] Internet Explorer and by the end of that [42:47] year in 96 Internet Explorer passes 20% [42:50] market share 97 it passes 40% market [42:53] share 98 it passes 60% market share and [42:57] then by the year 2000 Internet Explorer [43:00] basically has for all intents and [43:02] purposes 100% worldwide browser market [43:05] share if you look at the Internet [43:08] Explorer market share chart over time it [43:11] is the most perfectly rounded Hill that [43:13] you will ever see it goes from zero in [43:15] '95 to like 100 in 2000 and then all the [43:19] way back down to zero in [43:22] 2010 which is the next chapter of this [43:24] story is how on Earth did they lose that [43:27] Monopoly that they had in the browser [43:30] but before that there's this interesting [43:32] moment of reflection here why did [43:35] netscape's business dry up because their [43:38] business was made from selling server [43:39] software well the way to have the best [43:42] server software is to also control the [43:44] client people are very interested in [43:47] making sure that their websites run [43:49] perfectly using the experience that [43:51] everyone has and when you can no longer [43:53] claim hey a whole bunch of Internet [43:55] users actually use our browser [43:57] do I really want to buy my server [43:59] software from you or should I just be [44:01] open to buying it from anyone that I can [44:03] sort of it's the lowest cost and the [44:05] best value with the most features all [44:07] that so they sort of lose the [44:08] Competitive Edge in the revenue side of [44:10] the company on top of that it's just [44:12] really hard to recover for companies [44:14] that have a 80% draw down or whatever in [44:16] their stock price there was a lot of [44:18] excitement around the company that then [44:21] goes away suddenly all these employees [44:23] are [44:23] undercompensated it's a company killing [44:26] event and [44:27] all the market cap and excitement was [44:29] all on the C it wasn't because of the [44:31] revenue right so to this point Microsoft [44:33] has not changed their business model [44:35] they simply vanquished a potential [44:38] future that was dangerous for them [44:40] they're still doing the same thing as [44:42] ever selling Windows licenses through [44:44] oems and to Consumers at retail yep [44:47] there a couple more fun little tidbits [44:50] from this era in August [44:52] 1997 is when the famous Mac World [44:55] happens where Steve Job JS returns to [44:58] the company yes and Bill Gates shows up [45:00] on the satellite feed and you know of [45:03] course this moment is legendary but [45:06] studying it from this lens I realized [45:09] there's this whole other aspect to it [45:11] that I didn't know before so what bill [45:14] and Steve announced on stage it's also [45:16] so telling that bill couldn't even be [45:17] there in person he joins by satellite [45:20] yeah so there are four points to the [45:23] partnership one is the $150 million [45:26] investment from Microsoft and Apple two [45:30] is the 5year commitment on the part of [45:32] Microsoft to ship Office for Mac yep and [45:35] those are the big ones that everybody [45:36] talks about which by the way saved Apple [45:40] the company would have been completely [45:41] out of business because it was so [45:45] existentially important to anyone using [45:48] an Apple computer to use office that if [45:50] office you know Microsoft decides oh [45:52] we're going to stop developing office [45:53] people stop buying Max and you know the [45:55] company's already in such a financial [45:57] position it's just over so the third [45:59] deal point was they agree to end all [46:02] patent disputes so this is the end of [46:05] all the back and forth that we talked [46:06] about in part one but then the fourth [46:07] point which I didn't even remember at [46:09] all was that Internet Explorer would [46:12] become the default browser on the Mac [46:16] displacing Netscape and that continued [46:20] from 1997 until 2003 when Safari became [46:24] the default browser on and knowing now [46:27] and knowing the head that bill was in I [46:29] got to imagine that's the reason he did [46:31] the deal well it's funny I actually do [46:34] have some color on why he did the deal [46:36] Steve Jobs wanted to message this as [46:40] Microsoft believes in the Mac as a great [46:44] way to use the office suite they believe [46:46] in us and as a company and so they're [46:48] investing $150 million in making this [46:51] commitment to help us get through this [46:53] difficult time and this money by the way [46:55] just to help people understand Apple was [46:58] worth about $2 billion at the time so [47:01] this is Microsoft buying 8% of Apple [47:05] yeah wow Steve cleverly identified this [47:10] moment as a time to call Microsoft and [47:13] say hey I know we're through all the [47:15] patent issues that big lawsuit I have [47:19] more oh I think you guys are using some [47:22] of our stuff I don't want to sue you I [47:26] know the OJ is sort of looking at you [47:29] guys for antitrust right now Apple was [47:32] aware that Microsoft would be interested [47:35] in appearing collaborative with another [47:39] major player in the ecosystem and so we [47:41] sort of have the leverage to say Hey [47:42] what if you guys invested in us and did [47:44] this big commitment to Office for Mac [47:46] it's super important it help us get [47:48] through this difficult time and [47:50] Microsoft said back well it's really [47:52] important to us to have IE everywhere [47:54] and so they rolled it all into one big [47:56] deal no one's going to sue anyone all [47:58] the IP is cross- licensed and Microsoft [48:01] gets the win with ie Apple gets the win [48:03] with the investment in office and we can [48:05] all walk away interesting interesting [48:08] apple is saved truly Apple would have [48:10] gone out of business had Steve Jobs not [48:12] seized this opportunity it was a [48:14] critical business deal for both of them [48:17] yeah so to close the book on Netscape in [48:22] November 1998 AOL acquires Netscape in [48:26] an all stock deal for just over $4 [48:29] billion but again all stock and this is [48:33] just a little over a year before the [48:34] Time Warner merger and this moment here [48:38] is just the absolute peak of Microsoft [48:42] as a consumer technology company I mean [48:45] I think maybe the absolute peak of any [48:47] consumer technology company ever I mean [48:50] think about the market power that [48:52] Microsoft has at this point in time [48:55] Apple has an existential Reliance upon [48:59] them they have completely crushed [49:02] Netscape you know quote unquote cut off [49:04] the air supply there's nobody else there [49:06] is nothing else except Microsoft Google [49:10] is three years from being founded [49:12] Facebook is nine years from being [49:14] founded yeah there's Yahoo sure there's [49:17] real competition in the Enterprise at [49:19] this point Sun Oracle but in terms of [49:22] what your point is the consumer [49:24] technology landscape yeah [49:27] they had ultimate power yep yeah but [49:29] David I don't know the whole thing of [49:31] you can just decide to and then you [49:33] completely Vanquish your biggest [49:35] existential Threat by cutting off their [49:36] Air Supply like shouldn't that be [49:40] illegal [49:42] well so I mean to lead you a little bit [49:44] into our next section [49:46] yeah well listeners I think you know [49:49] what is coming next based on David and I [49:51] coily alluding to it but before we get [49:53] there we've really been talking about [49:55] this idea of [49:57] development platforms we were talking [49:59] about the web as a potential development [50:01] platform of the future even as far back [50:03] as 1994 people building web applications [50:06] or you know Windows 95 in its aing but [50:08] what makes for a great development [50:11] platform well this is a great time to [50:14] thank our friends at JP Morgan payments [50:17] to talk about their developer platform [50:20] to build a great one it requires a [50:22] culture focused on empowering others and [50:24] investing with a longtime Horizon [50:27] and JP Morgan payments knows this well [50:29] we've talked about how they're much more [50:31] than a global bank they're now investing [50:32] $7 billion a year in technology and R&D [50:36] but you may not know that they've also [50:37] been investing heavily in their [50:38] developer ecosystem and over the last [50:40] year they've really embraced this Dev [50:42] first mentality with their API powered [50:44] Payment Solutions yep and if you zoom [50:47] out it's clear that traditional Finance [50:48] folks aren't the only decisionmakers [50:50] anymore it's also your developers your [50:52] engineers your product managers and so [50:54] on who now have a seat at the table when [50:56] it comes to finding the best payments [50:58] platform for your company many of these [51:00] modern digital first categories like [51:02] ride sharing e-commerce or travel [51:04] couldn't exist without smooth and [51:06] seamless payments built into the product [51:08] experience natively yes we know that [51:10] many of you listening are developers who [51:12] will be excited to learn about JP [51:14] Morgan's new payments developer portal [51:16] that's essentially a One-Stop shop to [51:18] build solutions for your business on top [51:20] of their trusted and scaled platforms [51:22] while this is just a first step in a [51:24] long journey for their developer portal [51:26] JP Morgan is really taking a Long View [51:28] in investing for the future working hand [51:30] inand with their clients Dev teams beta [51:32] testers and payment industry experts to [51:35] launch learn and iterate and they've got [51:37] a robust end to-end road map of payment [51:39] apis coming in the future across [51:41] treasury Commerce embedded finance and [51:43] even value added services like account [51:45] validation that are going to be truly [51:46] unique in the industry anyone can make [51:49] an account to create projects [51:50] collaborate with team members generate [51:52] access tokens and try out payment apis [51:54] that help accept manage or send payment [51:56] in a sandbox environment which is great [51:58] I can speak from my past experience as a [52:00] developer that it is very nice when your [52:02] head's down you can just use self-serve [52:04] well-documented apis and code samples I [52:07] actually read through the quick start [52:08] guide and thought it was very easy to [52:09] follow with JP Morgan you get to rely on [52:11] their experience and security so you can [52:13] focus on your core functionality this [52:16] season we've talked a lot about how [52:17] they've been powering secure Innovation [52:19] with an ecosystem of Payment Solutions [52:21] trusted by some of the world's largest [52:22] and most Innovative companies not to [52:24] mention 200 years of banking experience [52:26] your payments just work the first and [52:29] every time you get security in huge [52:31] scale we've said it before JP Morgan [52:33] moves 10 trillion a day over 50% of all [52:36] e-commerce transactions in the United [52:38] States pass through their platform we [52:41] encourage you to learn more about their [52:42] API powered Solutions built for [52:44] developers at jpmorgan.com Acquired and [52:47] let us know if you've tested out their [52:48] new payments developer portal in the [52:50] slack if any listeners are heading to [52:52] fintech Devcon in August you'll be able [52:54] to learn more about all of this directly [52:57] from their team okay David so we've [53:00] arrived the famous 1998 Microsoft versus [53:05] the US Department of Justice Anti-Trust [53:07] trial yes and I was thinking about it in [53:10] the transition at the end of the browser [53:12] Wars there well you didn't like my [53:14] snarky comment yeah well we were being [53:16] gled about like oh this should be [53:17] illegal that's really the question here [53:20] all that power that Microsoft had it had [53:22] probably never been concentrated in the [53:26] hands of one company like that and [53:27] probably never will be again and the [53:29] question is was that illegal and did [53:32] Microsoft do anything wrong right we're [53:36] getting into a whole bunch of very [53:38] interesting questions here and I asked [53:39] it exactly to sort of pop open the can [53:41] of worms but there's a question of what [53:44] actually is legal in the US what [53:46] actually is legal in the EU then there's [53:48] this interesting question sort of [53:50] emotionally for everyone who is working [53:51] on software at Microsoft the vast vast [53:54] majority of people are not really [53:56] focused on what is the business and [53:59] competitive strategy most people who [54:00] worked on any of this stuff their whole [54:02] goal was I want to ship great software [54:04] and make things that people love to use [54:07] and I want to work with people that I [54:08] love making it with and so if you ask [54:10] most people who worked on any of this [54:12] their opinion is I don't know we were [54:14] trying to just make the best software [54:15] out there which is very interesting to [54:17] square with this growing public [54:21] perception that Microsoft is being a [54:23] bully especially public generated by [54:25] their competitors [54:27] right and then the literal legal [54:29] question of did they do something [54:31] illegal Because the actual antitrust [54:34] laws are a super different thing than oo [54:37] does this feel anti-competitive in some [54:39] way to me and then there's the other [54:41] dimension too of as a consumer am I [54:45] unhappy that I get a worldclass web [54:47] browser included in my operating system [54:50] well David now you're cracking open the [54:52] issue of consumer harm the consumer [54:54] welfare standard that the whole thing is [54:56] based so take us into the story yeah so [55:00] the Microsoft antitrust Saga actually [55:04] started not with the Department of [55:06] Justice and not in [55:08] 1998 but with the Federal Trade [55:10] Commission the FTC all the way back in [55:14] 1990 when they opened an investigation [55:17] into the company about whether it was [55:20] violating antitrust laws this centered [55:23] on the notion of per process licensing [55:27] which we discussed in our last Microsoft [55:29] episode yep so in July 1993 the FTC [55:33] Commissioners vote on whether Microsoft [55:35] is a monopoly that deserves further [55:37] action and penalties and they deadlock [55:39] at 2 to2 which means essentially a win [55:42] for Microsoft no action would be taken [55:44] against the company this is a huge [55:47] Victory the antitrust case of the US [55:50] federal government against Microsoft [55:53] should be closed at this point in time [55:55] yep theoretically they could have [55:58] examined any monopolistic practice at [56:00] this point and they said just the one [56:04] narrow thing that we were worried about [56:06] they agreed to stop doing and we in [56:09] voting 2 two we see no other issues that [56:11] we need to investigate yep Microsoft you [56:14] are good as far as the US federal [56:16] government is [56:17] concerned however the very next month in [56:21] August [56:23] 1993 the Department of Justice picks up [56:26] the case which is pretty unprecedented [56:30] one Department in the US federal [56:31] government essentially investigates a [56:34] company about whether it is abusing its [56:36] Monopoly power declines to prosecute [56:40] them for it and then another department [56:43] within the federal government the very [56:45] next month says essentially well we [56:48] don't think you did it right we're going [56:49] to do it Microsoft is now all of a [56:52] sudden basically standing trial for the [56:54] same accused crimes twice yeah [56:56] theoretically Double Jeopardy is not a [56:58] thing and in fact several members of the [57:00] FTC commission opposed this whole [57:03] process and tried to refuse to turn over [57:05] their notes to the justice [57:07] department but nonetheless the doj [57:10] proceeds and the next year in July 1994 [57:15] Microsoft just settles with them rather [57:17] than going to trial they're like all [57:18] right look we just want to be done we [57:21] are going to settle with you doj we're [57:23] going to be done with the US federal [57:24] government here and in that settlement [57:27] they agree to enter into what folks may [57:31] know the famous words a consent decree [57:34] and that means they consent in this case [57:37] that they are not going to tie the sale [57:40] of Microsoft application products to the [57:43] sale of Windows meaning they can't say [57:46] like hey oems or businesses or consumers [57:49] or whoever if you're buying Windows you [57:51] have to also buy office or X or whatever [57:54] else that we're selling in our app appli [57:56] group but importantly as part of the [57:59] consent degree they remain free and [58:01] clear Microsoft does to integrate [58:04] additional features into the Windows [58:07] operating system which brings us right [58:10] back to internet explore is it a product [58:13] or is it a feature exactly and this is [58:15] so messy because I think David you just [58:17] use the exact language which is they [58:19] cannot tie these application products in [58:23] a bundled sale however they absolutely [58:27] can integrate new features yes so what [58:30] is Internet Explorer and this also looks [58:33] the other way at the whole idea of [58:36] software development and platforms which [58:38] is it is a continuously changing [58:41] landscape where over time in the [58:44] interest of users platforms do more and [58:47] more and more things that applications [58:49] used to do and so the whole notion that [58:52] they're going to write that sentence and [58:54] then call it good what is an application [58:57] today might be a feature years down the [59:00] line but the law is written and we have [59:02] to pay attention to that sentence [59:04] constantly re-evaluated in the context [59:07] of the current time yeah I mean today [59:10] could you imagine purchasing a device [59:14] that has an operating system and that [59:17] device not having an internet browser as [59:20] part of the core system no you can't [59:23] even imagine that of course it's a [59:25] feature well is it a feature it's [59:26] actually it's literally an application [59:28] it is a bundled application as it exists [59:30] today so this is the gray area this is [59:33] the gray area and you know look if you [59:35] ask Bill and Microsoft and Jay Allard [59:37] all the way back to the original memo it [59:40] was absolutely intended to be a core [59:42] feature of the Windows operating system [59:45] having an internet browser as part of it [59:47] clearly motivated by the idea that we [59:51] want our Windows platform to maintain [59:53] the power it enjoys from its Monopoly [59:55] market share share so there's a [59:56] sympathetic view for sure of hey this is [59:59] core functionality to an operating [60:01] system whether it's a feature or an [60:02] application that we bundle and also [60:05] clearly the reason you are incentivized [60:06] to ship your own browser is to cut off [60:08] the air supply of potential competitors [60:10] that develop the platform of the future [60:12] yes so in October 1997 the justice [60:16] department files a motion in Federal [60:18] District Court stating that by bundling [60:20] Internet Explorer with Windows Microsoft [60:23] has now violated the 1994 consent decree [60:27] against product tying and it's important [60:29] to know what they're basically asking is [60:31] this is not about future versions you [60:33] know we know you're doing some kind of [60:34] Windows 98 thing we're saying right now [60:38] stop shipping IE bundled into Windows [60:41] Microsoft insists this is an integrated [60:44] product you cannot do that and it's not [60:47] even necessarily a legal argument yet of [60:50] we're allowed to do this cuz it's an [60:52] integrated feature they're saying we [60:55] ship a p of code and you actually cannot [60:59] just rip out Explorer and if you [61:01] remember at this time you could do all [61:03] sorts of crazy stuff like you could [61:04] paste a web address in Windows Explorer [61:08] and it would render even though it [61:10] wasn't Internet Explorer so there [61:11] actually was like if you think back to [61:13] that sort of vision of the browser is [61:16] integrated into the windows shell and it [61:18] sort of happened a browser was not [61:20] really at least Internet Explorer was [61:22] not really its own Standalone thing it [61:25] was deeply integrated now could they [61:27] have pulled it apart as a different [61:29] question if they really wanted to and [61:31] also remember the fact pattern here [61:33] isn't exactly great for Microsoft of [61:35] well they did ship Windows 95 without [61:37] Internet Explorer in the beginning so [61:40] right so the federal judge Thomas [61:43] Penfield Jackson orders them to do it [61:45] anyway or more specifically he ordered [61:47] Microsoft to ship a version of Windows [61:50] to the PC makers the oems or original [61:53] equipment manufacturers that didn't [61:55] include ie so that those oems could load [61:57] those onto the PCS that they were going [61:59] to ship to customers if they wanted to [62:01] and Microsoft said we told you we can't [62:03] do that but you're a judge and you're [62:04] ordering us to so they do and surprise [62:07] surprise when you just disable a bunch [62:09] of code that other code depends on it [62:11] doesn't work so then of course two [62:13] things happen judge Jackson is not [62:16] pleased since it appears Microsoft is [62:18] complying with the letter of the law but [62:20] violating the spirit and sort of [62:21] thumbing its nose and being arrogant so [62:24] that's Thing One Thing two is obviously [62:26] the PC makers don't actually ship this [62:28] version of Windows so it never sees the [62:30] light of day and so things get real [62:33] Petty real fast the doj asks the court [62:35] to hold Microsoft in contempt yeah whole [62:38] bunch of back and forth Microsoft [62:41] appeals judge Jackson's order and in [62:44] early May 1998 the appet court rules [62:48] that Microsoft can continue shipping [62:51] windows with ie bundled into it and also [62:54] continue to bundle any other features [62:56] that they want as part of Windows as [62:58] long as it benefits consumers and this [63:00] is interesting because this is when it [63:01] really hammers home the idea of what we [63:05] the US courts care about is consumer [63:08] welfare we haven't explored the idea of [63:11] if Microsoft is a monopoly or not yet [63:15] but for now what we are saying is as [63:18] long as what they are doing is in the [63:19] consumer best interest they're not [63:21] causing harm they're not raising prices [63:24] then it's okay so then one week later on [63:29] May 18th 1998 the doj announces a brand [63:34] new enormous wide ranging antitrust [63:39] lawsuit against Microsoft for violating [63:41] the Sherman Antitrust Act and abusing [63:44] its Monopoly power to suppress [63:46] competition and this investigation is [63:49] way bigger than just is it okay if they [63:52] tie Internet Explorer with the shipment [63:55] of windows [63:56] this is a is the company of Monopoly and [63:59] B are they doing anything across their [64:03] entire business to abuse that Monopoly [64:06] power in the disinterest of consumers [64:09] right it is not necessarily illegal to [64:11] be a monopoly it is illegal to abuse [64:14] your Monopoly power correct so this [64:16] court is examining both of those [64:17] questions one is Microsoft a monopoly [64:20] two are they abusing their power this is [64:24] really bad for Microsoft really bad and [64:27] it's worth decoupling did they do [64:29] anything wrong from just legal strategy [64:33] by holding a very firm line early of [64:37] we're appealing this decision we [64:38] couldn't possibly be doing anything [64:41] wrong here Microsoft is starting to take [64:42] this super aggressive stance and the [64:46] Department of Justice is then like wait [64:48] you didn't give an inch you're not open [64:50] to just this one thing the tying of [64:52] Internet Explorer and windows okay we're [64:56] gonna look at everything yep you can see [64:58] how it sort of gos them into like okay [65:01] we're gonna bring the big lawsuit right [65:03] but this whole suit you can also see [65:07] from Microsoft's perspective a feeling [65:09] of like betrayal by their government you [65:13] know like hey this is the third time we [65:16] are being tried for what feels like the [65:18] same crime Ben you said Double Jeopardy [65:21] isn't a thing earlier you know what is [65:23] this triple jeopardy come on I thought [65:25] this was supposed to be a free country [65:27] where we can build [65:28] businesses what the hell yeah and you [65:31] know Microsoft Folks at the time too are [65:33] starting to get this inkling of why are [65:35] they doing this are consumers really mad [65:38] at us you know who's being helped here [65:40] and they're starting to realize hm there [65:43] is a lot of lobbying going on behind the [65:45] scenes of Netscape and everyone else we [65:49] competing against trying to find a way [65:52] to call us anti-competitive yep which [65:55] which we should say is always true in [65:58] these big antitrust lawsuits but that [65:59] was certainly happening in this one yep [66:02] and so for all these reasons including [66:04] Ben as you say the legal strategy they [66:06] started with in the first place of we're [66:08] going to fight everything they say like [66:10] all right we're fighting this and we're [66:13] gonna fight it hard yep so Ben you [66:17] talked to a lot of people here take us [66:19] through what happens in this big trial [66:22] through the fall of 98 into 99 so the [66:25] first question question that everyone is [66:26] sort of wondering is did Microsoft have [66:28] a monopoly here well the fact is that [66:31] they had over 90% of PC operating system [66:34] sales so you know I'm not a judge but at [66:37] first glance there you think like okay [66:39] they have Market power so in August of [66:42] 1998 judge Jackson issues a pre-trial [66:46] order that all [66:48] depositions shall be submitted during [66:50] the trial only in transcript form and so [66:54] for folks who aren't in this world or [66:56] looked at lawsuits before a deposition [67:00] is when the council goes and does a [67:02] bunch of interviews beforehand you're [67:04] not being called as a witness in the [67:06] trial but it's basically information [67:08] gathering y interview process yes so on [67:11] August 27th Bill Gates is deposed by the [67:15] doj's appointed prosecutor David boy for [67:18] 20 hours and I think this happens over [67:22] multiple days actually on YouTube which [67:26] is interesting to note you can watch 12 [67:28] of the 20 hours I think I've watched [67:30] eight or nine of it but it's just hours [67:32] and hours and hours of just some you [67:34] know light bedtime viewing of uh Bill [67:37] Gates being asked questions so the [67:39] strategy that Gates and the Microsoft [67:41] legal team used was one that was [67:43] tailored for this pre-trial ruling if [67:46] you watch the video you can see that the [67:49] strategy is essentially never give an [67:50] inch avoid saying anything that can be [67:53] used against you and Microsoft walked [67:56] out of it feeling like they were pretty [67:57] successful in this and but when you say [67:59] tailored for the pre-trial ruling you [68:01] mean tailored with the assumption that [68:04] this is only going to be delivered as a [68:06] written transcript right there will be [68:09] no video no recording of right these [68:14] depositions and yet I just watched the [68:15] video on YouTube so what's going on here [68:17] how did that happen yeah so if you're [68:19] watching the video though it's very easy [68:22] to think this guy is rude pedantic and [68:26] disrespectful I'm not out on a limb [68:28] saying that opinion if anybody watches [68:30] this video that is just the obvious [68:33] takeaway yeah at a certain point they [68:35] argue over the definition of definition [68:38] is that right yes so a couple examples [68:41] I'm not exaggerating here the deposition [68:42] really does come across as just showing [68:44] pure disdain for the prosecutor and the [68:46] questions he's asking Bill Gates rat [68:48] holes on things like refusing to answer [68:51] questions about memorandums since they [68:54] were not memos but emails so I couldn't [68:57] possibly answer you on the question [68:58] about the memorandum at one point he [69:02] does look at David boy and ask him how [69:04] he would Define the word definition of [69:07] course while smirking the whole time and [69:09] so the whole thing is like very [69:11] obviously tailored with this idea that [69:14] I'm going to give you pages and pages [69:16] and pages of which you will have nothing [69:19] that can be used against me and that is [69:21] the whole strategy I don't care how I [69:22] come across I don't care how ticky tacky [69:25] the languages you know he sits and [69:27] pauses forever he'll say well you asked [69:30] me what the person who sent this was [69:31] referring to how should I know what [69:33] they're referring to I didn't write the [69:35] email you'd have to ask them I I don't [69:37] know and so it's 20 hours of this well [69:41] somehow and I actually don't really know [69:43] how this happened after the deposition [69:46] is recorded on October 9th the judge [69:50] then issued a reversal saying that [69:52] videotaped depositions are indeed [69:54] allowed to be used in court o yeah how [69:58] did this hold up and if you give a great [70:01] prosecutor like David Bo this [70:03] opportunity he uses it masterfully and [70:06] so throughout the trial he'd show little [70:08] Clips at strategic moments in the trial [70:11] where he either wanted to give the press [70:12] something juicy to write about that day [70:14] because there's a whole press section in [70:15] the back going and listening to all the [70:17] witnesses every single day or he would [70:19] play something he knows is going to get [70:20] to rise out of the judge and if the [70:22] judge makes an expression then the Press [70:24] writes about oh the judge Jud is leaning [70:25] this way or that way also he would use [70:28] it anytime there was an opportunity to [70:31] feel sympathetic for Gates or anyone at [70:33] Microsoft and then he would show a clip [70:36] that sort of clearly causes you to lose [70:38] any sympathy or leaning and so it was [70:40] just dripped out in this really clever [70:44] way Y and certainly went a long way [70:46] towards shaping the decision but also [70:49] shaping more importantly public opinion [70:51] about Gates and about Microsoft Y how [70:54] did this whole up though didn't [70:56] Microsoft appeal the change from [70:59] recordings not being allowed for [71:00] depositions to recordings being allowed [71:03] and is a great question David and one of [71:04] the things that I read to prepare for [71:06] this episode is a book called World War [71:09] 3.0 which is exclusively about this [71:11] trial and the author has this comment on [71:14] it Microsoft feared that judge Jackson [71:17] was a foe he had made a number of [71:19] pre-trial rulings deemed hostile to the [71:21] company they were especially unhappy [71:22] that he modified the pre-trial order [71:24] that deposition shall only be submitted [71:26] in transcript form issuing a new order [71:29] allowing videotape depositions Microsoft [71:31] suspected that Justice had somehow [71:33] prevailed on Jackson to amend his [71:35] earlier Court ruling Jackson [71:37] categorically denied this but does not [71:40] recall exactly why he issued the October [71:42] 9th ruling they groused but only in the [71:45] most unguarded private moments because [71:47] they were terrified of offending him [71:49] that Jackson was biased and would rule [71:51] in favor of the government so your [71:53] question of how does it hold up [71:55] I guess there was no formal challenge of [71:58] that change in Rule and part of it [72:00] probably was just because they realized [72:03] they had a long way to go with the judge [72:04] and didn't want to agitate too much WoW [72:09] interesting it also sounds like maybe [72:11] they didn't realize yet how disastrous [72:15] these tapes getting out was going to be [72:16] for Bill and for the company yeah I [72:19] think that's right interesting well okay [72:24] all of this [72:25] starts to culminate in November 1999 [72:29] these trials take forever when judge [72:32] Jackson issues a finding a fact that [72:35] Microsoft is indeed a monopoly in the [72:38] operating systems business now remember [72:41] it's okay to be a monopoly it's not okay [72:43] to abuse the power but simply the fact [72:46] that the judge has now issued his [72:49] opinion that it is a monopoly everybody [72:52] knows this probably means the other shoe [72:55] is about to drop and more specifically [72:58] the finding was that the network effects [73:01] from the large installed base so that's [73:04] users and large body of applications so [73:08] apps makes it prohibitively expensive [73:10] for a competitor to develop its PC [73:12] operating system into an acceptable [73:14] substitute for Windows which yeah of [73:18] course obviously that's what our whole [73:20] episode one was about correct right so [73:22] the finding of fact is hey it's Monopoly [73:25] but again not necessarily illegal to be [73:28] a monopoly only illegal to abuse [73:30] Monopoly power right so couple months go [73:34] by after the finding of fact and then on [73:37] June 7th 2000 judge Jackson issues the [73:41] final judgment in the case and he rules [73:45] that Microsoft did indeed abuse its [73:50] Monopoly power and as a remedy for [73:53] having done so he orders that Microsoft [73:57] be broken up into at least two separate [74:02] companies separate operating system [74:05] company and a separate applications [74:07] company just like the Standard Oil [74:10] breakup order however many years it was [74:13] before 90 I think also what this is [74:17] completely lost to history unless you [74:19] are a tech [74:20] Oldtimer Microsoft was ordered that was [74:24] the ruling by the court to split up yes [74:27] it wasn't just that oh Microsoft lost [74:30] the doj case no the ruling was Microsoft [74:34] will be split up by order of the United [74:38] States government and there's a whole [74:40] bunch of additional Provisions in this [74:42] Steve Balmer had to work at one company [74:44] and Bill Gates had to work at the other [74:46] they could not work at the same company [74:48] each of those two after they picked [74:50] their companies had to divest all of [74:53] their shares from the one that was was [74:55] not their employer so they couldn't have [74:56] this conflict of interest it is crazy [74:59] imagining this world that could have [75:01] been I mean clearly this didn't happen [75:04] but for a moment in time this was the [75:06] position of the United States government [75:08] it's totally w i mean can you imagine if [75:11] there was like the Gates company and the [75:13] bomber company I mean sort of as we're [75:15] going to talk about in the rest of this [75:17] episode that is what happened but in a [75:19] very different way it's also worth [75:22] pointing out from late 1999 [75:25] when the findings of fact came out over [75:27] the next 12 months Microsoft's market [75:29] cap dropped from 600 billion to 270 [75:33] billion which was a 55% drop now this [75:36] coincided with the dot bubble and the [75:39] CEO change that we're going to talk [75:40] about shortly but the perception of [75:43] Microsoft this super high flyer [75:46] completely fell off a cliff from this [75:48] ruling imagine if a ruling comes out [75:51] tomorrow that Apple needs to be broken [75:52] up and iOS needs to be separated from [75:56] the devices and you need to be able to [75:58] buy a phone without iOS what do you [76:00] think that's going to do to the [76:01] company's market cap yeah not exactly [76:03] the same thing of course because this is [76:05] not about devices well right but I'm [76:07] just I'm making a similar type of scale [76:09] analogy right like what the impact would [76:12] be yes so do you know the technicality [76:16] that was discovered no I don't I know [76:18] that Microsoft immediately appeals of [76:20] course it was discovered later in June [76:23] 2000 that judge Jackson had secretly [76:25] been meeting with reporters in his [76:27] chambers before the rulings were [76:29] delivered it's not allowed and so judge [76:31] Jackson was removed from the case yes [76:34] the reporters all had these embargoed [76:36] stories they could drop immediately [76:37] afterwards and everyone was like how did [76:39] you what wow that's wild yeah what a [76:43] freaking crazy Escapade here like [76:47] there's no other way to put it so this [76:50] is June of 2000 by the way the appeal [76:52] then takes a long time so there a [76:55] meaningful moment in history I think [76:58] about 15 16 months where the official [77:02] ruling is Microsoft should be going [77:04] through the preparations to do their [77:05] breakup right that is what the world [77:09] believes as far as anyone knows yep so [77:12] the appeals court removes judge Jackson [77:14] from the case they install a new judge [77:16] to readjudication [77:24] the parties toward a settlement [77:27] especially 9911 happens and then I think [77:30] that's a galvanizing factor to pull the [77:32] parties into the room and say hey this [77:34] has gone on too long and we need to put [77:36] this behind us yep also there was a [77:39] political Administration change from the [77:41] Clinton Administration to the Bush [77:42] Administration yep so then in November [77:46] 2001 just a couple weeks after the [77:48] Windows XP launch the doj and Microsoft [77:52] finally completely settled the case also [77:55] can we just say this case is brought [77:58] against Windows 95 Windows 98 comes out [78:01] and then before we have a resolution [78:03] Windows XP comes out Windows 98 you mean [78:06] the uh marketing update to sell back to [78:08] school PCS [78:10] yes but like insane right and the whole [78:14] time Internet Explorer shipping with [78:16] Windows right the whole time so November [78:19] 2nd 2001 the settlement is proposed at [78:22] this moment in time internx expor has [78:25] right around 90% market share right if [78:28] you are in the camp of Microsoft was a [78:30] monopoly was abusing its Monopoly power [78:33] you're like well this was a complete [78:35] failure of process the damage is done [78:37] right yeah meanwhile also if you're in [78:40] the Microsoft camp of what the hell is [78:42] our government doing you're also like [78:44] what the hell nobody is happy here right [78:47] an Innovative company that built the [78:49] most important product for that [78:51] technology phase meanwhile there's this [78:53] whole new thing going on with the [78:55] internet and like we need to figure out [78:58] how to legally navigate that transition [79:01] right yeah we have enough existential [79:03] threat to our business from technology [79:06] Trends happening yeah that to try and [79:08] navigate that with our hands tied behind [79:10] our back because of these legal [79:12] proceedings like come on right so 2002 [79:16] the settlement is finally approved it [79:19] reverses the order to be split up [79:22] obviously Microsoft is still one company [79:25] officially the ruling that Microsoft did [79:27] indeed have a monopoly is upheld they [79:30] put in place a 5-year consent decree and [79:34] the terms are that Microsoft is not [79:36] allowed to enter into contracts with PC [79:39] makers that excluded competitors I mean [79:42] fine two windows had to be interoperable [79:46] with non-microsoft software which of [79:49] course it does it always was yeah it's [79:52] the developer platform they have to [79:54] write API documentation and make their [79:57] apis such that developers can build [79:59] applications on top of them that is the [80:01] purpose of the company so okay three an [80:06] independent technical committee was [80:08] created to field complaints from [80:11] competitors okay they created a call [80:14] line that is it wow that's it am I [80:17] missing something David that's my [80:19] understanding of what it is yeah I no I [80:21] don't have anything [80:22] else but okay that is the letter of the [80:27] resolution here the actual cost of [80:32] this was immense nothing could have been [80:35] bigger I mean we spent the whole first [80:38] section of this episode talking about [80:40] how Microsoft was so powerful had never [80:44] been more powerful and there probably [80:47] never will be a more powerful company [80:49] than Microsoft in the late [80:51] 1990s this is what destroyed it oh that [80:54] is a take right there I think that we [80:55] will debate at the end of the episode oh [80:58] well the back half of the episode is [81:00] about the incredible story about how [81:02] Microsoft rebuilt itself in a completely [81:05] new market into now again the most [81:07] valuable company in the world but let's [81:10] just talk about what the actual cost was [81:12] not in terms of money it certainly [81:15] didn't actually impact Internet Explorer [81:17] or Windows XP was a huge success sells [81:20] over half a billion copies gets used [81:23] over its lifetime on probably a billion [81:24] PCS it unifies Windows under the NT [81:27] architecture has the Bliss wallpaper [81:29] amazing but the true cost is what it did [81:33] culturally and emotionally to Microsoft [81:36] I mean we talked to all these people and [81:39] God it was like death being there I mean [81:41] to believe for 16 months that the [81:43] company was going to be broken up for [81:45] Bill to have this really embarrassing [81:48] video of him all over the press and to [81:50] have the narrative change about Bill [81:53] change about the company you change [81:55] about for every employee working at the [81:57] company to like oh you're the best and [81:58] brightest in America to you guys are [82:01] evil and why are you working at this [82:03] company yeah it exposes the difference [82:05] too in the legal strategy of both sides [82:07] where Microsoft's strategy was to refute [82:11] Point by point every allegation brought [82:13] against them to the point where they [82:15] were trying to refute Netscape we don't [82:19] view Netscape as an existential threat [82:21] to us and they should have just probably [82:23] acknowledged you know Bill literally [82:25] wrote a letter that got published a memo [82:28] saying that Netscape is a competitive [82:30] threat born on the internet but they [82:31] wanted to refute every single point and [82:33] knock of an inch meanwhile all David boy [82:35] and the doj wanted to do was Destroy [82:38] Microsoft's credibility so that every [82:39] time they brought a witness there were [82:41] emails or there was a deposition that [82:44] basically called into question are they [82:47] really telling the truth on the stand [82:48] can they really not remember that and it [82:50] just blowby blow made Microsoft look [82:53] like they were licus and that has to [82:56] leak into the company culture that has [82:57] to make you on the one hand feel like [83:00] your government is attacking you but on [83:02] the other hand start to question and say [83:04] why did we do this again I thought we [83:06] were just trying to make the best s [83:07] where were we trying to do something [83:08] illegal and I just didn't know about it [83:11] right it's worth talking about some of [83:12] the other pieces of Fallout it did slow [83:15] Microsoft down there were huge amounts [83:18] of protocol documentation that needed to [83:20] happen so if anyone's running a software [83:22] company you know that if your iteration [83:24] times are slower and you just have [83:26] permanent new drag on your development [83:28] process you are going to fall behind and [83:31] I think that was one that was felt by a [83:33] lot of employees and managers who [83:35] suddenly can do less with the same [83:37] amount of resources that they have there [83:39] was also a bunch of private lawsuits Sun [83:43] AOL real networks Microsoft was paying [83:46] out billions of dollars in these private [83:48] settlements that followed the doj their [83:50] civil suits not to mention state [83:53] attorney generals were also suing [83:54] Microsoft left and right and [83:56] international many of the state AGS for [83:59] years who brought the suit together with [84:01] the doj did not accept this reversal and [84:05] so they tried to continue independently [84:07] suing Microsoft which was painful for [84:09] another fiveish years we made it all the [84:12] way to 2009 before they settled their EU [84:16] version of this antitrust case I mean [84:19] that's another what seven years after [84:22] the reversal and and in May 2011 that is [84:27] when the final consent decree finally [84:30] expired so basically from 1990 until [84:34] 2011 21 years of the company's life the [84:37] majority of the company's life had been [84:40] spent under some sort of antitrust [84:43] scrutiny or active litigation wow and [84:46] obviously the company thrived through [84:49] much you know if not all of that but [84:52] were consumers ever harmed I continue to [84:54] wonder this it was horrible for [84:57] Microsoft even though there weren't any [84:59] real material changes they had to make [85:01] but effectively they won right which I [85:05] guess they should have because it's not [85:06] clear that there was negative impact to [85:08] Consumers there was all kinds of [85:09] negative impact to existing competitors [85:12] or future potential competitors but that [85:14] is not the US Standard for antitrust law [85:17] especially at this point in history and [85:19] so I guess the right answer is the right [85:23] thing happened eventually but it was [85:25] awful to get there and it had all sorts [85:29] of indirect negative impact on the [85:32] company yep so I said a minute ago I [85:35] think it killed [85:38] Microsoft's immense dominant consumer [85:43] Technology power and the biggest reason [85:46] I say that we didn't talk to Bill Gates [85:49] as we were preparing for this but is [85:51] what this whole thing clearly did to [85:53] Bill Gates [85:54] yeah yeah Microsoft had one competitive [85:57] advantage that no other company had and [85:58] that was Bill Gates and for whatever [86:01] sets of reasons I mean I can imagine so [86:03] many thinking about like if I were in [86:06] that seat going through that bill at [86:09] Microsoft was never the same person [86:11] after this in fact Bill stepped down [86:15] before the final ruling from Judge [86:17] Jackson yeah so in July 98 right as this [86:21] big huge doj antitrust suit is heating [86:23] up [86:25] Steve bomber gets promoted to president [86:27] of the company bill is still CEO but [86:29] Steve is now promoted to president and [86:31] is the clear number two and then they go [86:34] through the trial the deposition the [86:37] November 99 finding of fact that [86:39] Microsoft is a monopoly and then Ben as [86:41] you're referring to on January 13th [86:45] 2000 Bill Gates announces that he is [86:47] handing the CEO role of Microsoft over [86:50] to Steve and that he is moving to a [86:53] newly created position as Chief software [86:55] architect and he will remain chairman of [86:57] the company but he is no longer going to [87:00] be CEO and then of course it's just a [87:02] few months later that the breakup [87:05] verdict comes down yeah going through [87:08] something like this has to feel personal [87:10] and has to change you forever I can't [87:11] imagine how it wouldn't totally [87:14] especially when again it's not clear to [87:18] me how consumers were harmed so this [87:21] constant battle this war that was w on [87:24] forever and ever and ever and ever it [87:27] totally distracted Microsoft and as [87:29] anybody can attest especially in the [87:31] tech industry if you are distracted you [87:33] just fail because you need to have all [87:35] of your best resources making stuff [87:38] building stuff focused on a firing on [87:40] all cylinders clear North Star strategy [87:43] and so if you tie up a company for five [87:45] years and you lose your leader through [87:47] it I mean somebody we talked to [87:49] characterized this period as like a [87:51] mental breakdown for the whole company [87:53] and I think that's kind of the best way [87:56] to characterize it yeah it's not fair to [87:59] blame everything we're about to talk [88:00] about all the future consumer failings [88:02] on this but it is helpful to keep this [88:04] in mind and say okay why perhaps did [88:08] they not fully have their wits about [88:10] them yes and so the transition to Steve [88:14] bomber happens this is the context under [88:18] which Steve bomber became the CEO of [88:21] Microsoft so I talked to a whole whole [88:24] bunch of people who were at Microsoft in [88:26] this era and one thing that every single [88:31] person brought up that never gets talked [88:34] about is how much Steve was the [88:38] emotional Rock for the company when this [88:41] was happening all the stuff everybody [88:43] thinks about Steve you know the running [88:44] around on stage the yelling the [88:46] screaming developers developers [88:48] developers when do you think all this [88:51] happened the crazy dancing on stage I [88:54] love this company that was in September [88:58] of 2000 when they thought they were [89:00] going to get broken up and Steve was [89:03] there trying to keep everybody moving [89:06] forward everybody we talked to was like [89:09] I don't know how he did it and it meant [89:11] so much it's actually shocking they held [89:14] on to as much talent as they did in a [89:16] 15-month period of people assuming the [89:18] company was about to be split right [89:21] knowing that context for me at least it [89:24] completely changed my perception of [89:28] Steve and of the company during this [89:29] time fascinating so when Steve takes [89:32] over his agenda is three things and I [89:34] think in basically priority order number [89:37] one hold the company together [89:39] emotionally I love this company that was [89:43] job number one just like keep everybody [89:45] coming to work job number two clean up [89:49] this antitrust mess and then job number [89:51] three I think was hey let's keep this [89:54] company like growing and winning and I [89:56] think it's kind of fair to say he did [89:58] all three so we just talked about one [90:01] emotionally holding the company together [90:03] two one of the very first things Steve [90:06] does when he becomes CEO is he promotes [90:08] Brad Smith to general counsel who Brad [90:12] Smith is still of course leading all [90:14] this in Microsoft to this day he's now [90:16] president and Steve tells Brad go make [90:19] peace so actually this is amazing Brad's [90:24] fin interview with the Microsoft board [90:26] of directors oh I was wondering if you [90:29] found this yep for his job you know to [90:31] be promoted to General counsil his [90:34] PowerPoint presentation to the board is [90:36] just one slide that has one sentence on [90:39] it it's time to make peace and that is [90:43] totally what he goes and does and he [90:46] says okay I'm going to figure out what [90:49] settlements we can live with and I going [90:51] to go settle everything [90:54] and this company just needs to move [90:56] forward and it doesn't matter that we [90:57] all feel it wasn't fair it doesn't [90:59] matter that we all feel this was a sham [91:02] of a process we just have to move on and [91:04] we have to live in a new reality and you [91:06] kind of need a new set of people to do [91:09] that it's kind of amazing that Steve was [91:10] part of the Old Guard and the new guard [91:12] to do this because how can you say I'm [91:14] going to put how unfair I feel this was [91:17] aside and just focus on moving forward [91:21] that is an extremely difficult [91:22] compartmentalization exercise and so for [91:25] Brad to come in and say like I'm going [91:26] to be the guy who is able to disregard [91:28] the past and figure out how we and I use [91:31] this phrase in the first episode become [91:33] a trusted partner to governments around [91:35] the Free World I mean how crazy is it [91:38] that this Microsoft that we just talked [91:41] about for the last hour became the [91:44] Microsoft that can do no wrong from a [91:47] regulatory perspective the only one [91:48] that's not under active antitrust [91:51] investigation today by the federal [91:53] government the one that is a massive [91:56] provider of software and services to the [91:58] US and its allies at the government [92:00] level right the reversal here is it [92:03] doesn't get talked about enough what an [92:05] amazing job Brad and the company did to [92:09] reverse this perception so then that [92:11] leaves job number three on Steve's [92:13] agenda of be successful continue to have [92:16] Microsoft be a leading technology [92:17] company and hopefully still grow revenue [92:20] and profits and Bill Gates is still [92:22] chairman of the board like not only is [92:24] he a full-time employee being the chief [92:25] software architect it's not that it's [92:27] like a sham that he's not the CEO but he [92:30] is a very present voice at the table in [92:35] these big decision-making moments and so [92:38] for how do we become a company that [92:41] continues to innovate and make great [92:43] products despite all this he still has [92:45] Bill as the technical leader of the [92:47] future products yes and absolutely bill [92:50] was still there and Steve had Bill and [92:52] they were running the company together [92:54] absolutely but what's so interesting is [92:57] Microsoft right at this time basically [92:59] starts a transformational journey from a [93:03] technology company RIT large a consumer [93:05] and sort of Enterprise technology [93:07] company to the Enterprise technology [93:11] company and that is a muscle that as we [93:13] talked about last episode Steve had been [93:14] building for a while but boy does he [93:17] really come into his own here and [93:19] Microsoft the entire Enterprise [93:22] Juggernaut that it built [93:25] the bulk of it really is post doj it is [93:27] like new business and new markets that [93:29] they are getting into yep so then the [93:32] question becomes how did Microsoft build [93:35] this phenomenal Enterprise business and [93:39] along with that release XP the most [93:41] successful Windows operating system ever [93:44] and then we're going to talk about Vista [93:45] and then we're going to talk about zoom [93:47] and search and Bing and Windows mobile [93:49] and Windows 8 and yeah all that but [93:52] before we do we would like to thank [93:55] hugee partners of ours here in this [93:57] season of acquired service now yep [94:00] service now is the AI platform for [94:03] business transformation helping automate [94:04] processes improve Service delivery and [94:07] increase efficiency over 85% of the [94:09] forun 500 runs on them and over the past [94:11] few years they've joined companies like [94:13] Microsoft as one of the most important [94:15] Enterprise technology vendors in the [94:17] world and speaking of Microsoft and [94:19] service now they just announced a huge [94:21] expansion of their partnership [94:23] specifically integrating the two [94:24] companies Enterprise AI assistant [94:27] starting in the fall customers will be [94:28] able to interact with service now's now [94:30] assist AI assistant directly within [94:33] Microsoft co-pilot yeah it's telling for [94:35] the magnitude of this partnership to see [94:37] SAA nadela appearing in the keynote at [94:40] service now's big annual event knowledge [94:42] last month it had Echoes of that Bill [94:45] Gates 1997 macw World video that put [94:48] apple back on the map not that service [94:50] now needed putting back on the map yes [94:53] and like that historic announcement from [94:55] Bill committing to Microsoft Office for [94:57] the Mac this partnership is also huge [94:59] service now's now assist will be [95:01] integrated with Microsoft co-pilot and [95:03] will be available directly from office [95:05] apps starting with Microsoft teams in [95:07] August the AIS are integrated into one [95:10] seamless user experience without [95:12] actually sharing data so if for example [95:15] a user asks co-pilot in teams about how [95:19] the company's laptop policy Works behind [95:21] the scenes co-pilot shares that request [95:24] and context with now assist and now [95:26] assist accesses internal company policy [95:29] with the right permissions for that user [95:31] and Returns the answer to co-pilot in a [95:33] rich card with options for the user to [95:36] kick off a workflow via now assist in [95:38] the future Microsoft co-pilot will also [95:40] be integrated the other way into nsis so [95:43] it can automatically generate office [95:45] files like PowerPoint presentations and [95:47] Excel spreadsheets directly from assets [95:49] and knowledge in the service now [95:51] platform it's pretty awesome for both [95:53] companies and especially awesome for [95:55] Enterprise users so if you want to learn [95:57] more about the service now platform and [95:59] how it can work with your company's [96:00] Microsoft Services go over to [96:02] servicenow.com Acquired and when you get [96:05] in touch just tell them that Ben and [96:06] David sent you all right so to [96:09] contextualize how this Enterprise [96:11] business was built it is worth [96:13] understanding the shape of Microsoft's [96:15] business like the divisions what [96:17] products generated what Revenue even [96:19] before all this doj stuff so if we go [96:21] back to 1996 Bill Gates gave a great [96:24] interview where he was talking about the [96:26] kind of four businesses that they're in [96:28] today oh this is the wired interview [96:30] with Kevin Kelly right yes it's on [96:32] YouTube it's great it is great so [96:34] there's Windows which he calls one [96:36] business there's NT slack office there's [96:40] office which he calls a $4 billion do a [96:43] year business and those three businesses [96:46] together are over 90% so you can think [96:48] about it as Windows and he said NTB back [96:51] office but this is the Enterprise an [96:53] office yeah which is so funny that bill [96:55] thought of it as NTB back office it [96:57] really exposes that Steve was the one [96:59] who had the passion for the Enterprise [97:01] yes bill was like it's like this stuff [97:03] that businesses buy but I'm going to [97:05] refer to it by its Microsoft product [97:08] name of one of the products we sell [97:10] which is NT and then the last 10% is [97:13] everything else so there's MSN [97:15] e-commerce games encyclopedia Maps joint [97:18] ventures DreamWorks and NBC so he's [97:21] talking about the interesting thing the [97:23] server business which is a different way [97:26] he refers to NTB back office all the way [97:29] back in '96 said it's the fastest [97:31] growing business even faster than [97:32] Windows or office so they sort of know [97:34] they're on to something yeah but they [97:37] haven't quite cracked the go to market [97:40] motion the pricing the service [97:44] organizationally how do they sort of fit [97:46] it in that all comes later or the [97:48] products either really yes that's a [97:50] great point the fact that he calls it [97:52] back office this is so telling okay so [97:55] we did talk last time about NT and Dave [97:58] Cutler and the heroics that he performed [98:00] to right NT Windows NT though was still [98:06] a client operating system architecture [98:09] for a user to use a personal computer [98:12] with NT basically was Enterprise ready [98:15] it was like very networked for work [98:17] groups it ran on only the most high [98:20] power PCS but you're right David it was [98:24] designed for the thing that you know the [98:26] first 25 years of Microsoft was all [98:28] about which is PCS it's not like oh [98:32] we're a systems company that makes stuff [98:35] for all use cases all over your [98:37] Enterprise it's no we make stuff that [98:40] runs on a box sitting in front of you [98:43] yes [98:45] and discovering this [98:47] distinction is what Microsoft in this [98:50] next era really really need nailed and [98:54] they discovered that the [98:56] Enterprise is not about users it's about [99:01] it and it's about systems for Better or [99:04] For Worse yes and discovering that and [99:07] the products and the sales motions that [99:10] Microsoft could then go use to sell to [99:14] Enterprise it and sell [99:16] systems was a new you know multi hundred [99:19] billion dollar market that Microsoft [99:22] could now go attack and offense in in [99:24] this post doj landscape whereas they're [99:27] playing defense everywhere else hey here [99:29] our market share is zero we can do [99:31] whatever we want here well it wasn't [99:33] zero but they were fighting Sun IBM [99:37] Oracle yes really IBM but Sun yeah to [99:41] Oracle Etc and it was perfectly suited [99:44] to Steve's strengths so Ben if you've [99:47] ever heard of these now sort of strange [99:50] sounding Microsoft products SQL Server [99:53] active directory exchange Dynamics [99:57] SharePoint SharePoint was technically [99:59] within office but it is one of these [100:01] systems types products these are all [100:04] every single one of those names I just [100:05] mentioned become multi-billion doll [100:08] Revenue Enterprise it server products [100:13] that are built and sold during the Steve [100:15] bomber era of Microsoft and what's so [100:18] like honestly beautiful about this is [100:21] they work in concept [100:24] with Windows and office on the PC client [100:28] side so like this is the client server [100:30] era that Microsoft really dominates here [100:33] and Microsoft within Enterprises all [100:36] these new server products work best with [100:41] Windows operating system devices running [100:45] Microsoft Office applications on them [100:48] and those Windows operating system [100:51] devices and those office applications [100:53] work best with the Microsoft server [100:55] products you now have a full system [100:59] solution from one technology vendor as a [101:02] major Enterprise it's like the most [101:05] incredible three-sided technology [101:08] flywheel ever built and one benefit from [101:11] this which of course if you're Microsoft [101:13] you don't want to lean on this benefit [101:14] but they end up doing it is if you make [101:17] everything integrated together work well [101:20] and come from one vendor nothing [101:22] actually has to be best of breed and so [101:25] you're no longer competing with any [101:27] point Solutions you offer the whole [101:29] thing sure yeah you can consider going [101:32] and buying that other vendor's directory [101:34] service or that other vendor's email [101:37] server but are you really because you [101:39] buy everything from us and it all works [101:40] pretty well together yep the very very [101:44] very best example of this that most [101:46] listeners can probably tangibly relate [101:49] to as well is exchange EMA and [101:53] calendaring service and Microsoft [101:55] Outlook and windows it all has active [101:59] directory that syncs across everything [102:02] in doing all this research it seemed to [102:03] me that once a Enterprise adopted active [102:09] directory they were going to tip and [102:11] they were going to buy the rest of the [102:12] software too because whoever manages the [102:15] source of Truth for who are all the [102:16] people and what are all the resources [102:18] you know devices and everything that my [102:20] company owns everything else needs to [102:23] reference that canonical set of proper [102:25] nouns whether it's email whether it's [102:27] calendar so that was this incredible [102:30] sticky product that then you could just [102:32] keep attaching more and more stuff to [102:34] any Enterprise need oh we got you [102:36] covered and hey it works with active [102:37] directory yep and the whole product [102:40] effort here started with database in [102:44] 1998 Microsoft takes SQL server and it [102:48] was the first real Enterprise ready [102:51] database that can Ral IBM and Mainframe [102:55] databases Oracle [102:57] databases and of course unlike IBM it [103:02] runs on x86 Intel architecture so the [103:06] pitch now to Enterprise it is everything [103:09] we just said about why working with [103:11] Microsoft server products is better for [103:13] the whole ecosystem reasons also total [103:16] cost of ownership don't pay IBM tons of [103:18] money for their main frames just go buy [103:21] cheap x86 windows boxes from Dell or [103:25] whomever and use that as your it server [103:28] architecture fascinating I don't think I [103:30] quite understood that and so basically [103:32] you then have NT as the operating system [103:35] SQL is the database and then you've got [103:37] all these other applications that [103:38] basically run on that stack and here's [103:41] where exchange and Outlook and [103:42] everything comes in this is right as [103:45] email is taking off as like the killer [103:49] application in Enterprises and so now [103:53] Microsoft shows up and says we've got [103:55] this great new product for you it's [103:57] called exchange and like maybe you were [103:59] using Lotus Notes before which of course [104:02] developed by the legendary Ray Azie he's [104:04] going to come back up here in a minute H [104:06] Lotus gets acquired by IBM for three and [104:09] a half billion in 1995 you're buying [104:11] Lotus Notes from IBM come take a look at [104:15] exchange exchange has Email exchange has [104:20] calendaring exchange has address book [104:23] exchange has Outlook it is a first class [104:27] included in the bundle of Microsoft [104:30] Office Office application that you you [104:33] know Mr and Mrs Enterprise are now going [104:35] to get for all your users and it works [104:38] just beautifully and perfectly with our [104:41] exchange email calendaring an address [104:43] book service it sells itself [104:46] [Laughter] [104:48] basically and then you were talking [104:49] about active directory that led to [104:51] active directory of oh okay well now [104:55] you've got your whole database [104:57] architecture running on Microsoft you've [105:00] got your email and your calendaring [105:02] architecture running on Microsoft you've [105:04] got your windows machines out there well [105:07] you've got all these employees within [105:09] your company all these users with all [105:12] these devices you kind of need to manage [105:14] them and you need to know who has what [105:16] security access and how to find each [105:18] other and where should the mail get [105:19] routed and all that well we've got this [105:22] great new product for you it's called [105:23] active directory yeah it's pretty [105:25] incredible that's all on the sort of why [105:29] it's good for customers on the why it's [105:31] good for Microsoft Steve also pioneered [105:35] this bundling idea which is once you [105:37] sign the Enterprise agreement you get [105:40] access to all of this and if you're a [105:43] customer that's only using 30% of the [105:45] things in the bundle if you have [105:47] business needs that involve some [105:50] Microsoft product that comes for free in [105:52] your bundle you're going to adopt that [105:53] and guess what you just became a [105:55] stickier Microsoft customer I feel like [105:58] this often goes overlooked in the like [106:00] oh Microsoft's a big boring Enterprise [106:02] company right now there was a tremendous [106:04] amount of business model innovation in [106:06] figuring out that bundling like that [106:09] with additional products can create [106:12] stickiness which eventually creates more [106:14] Enterprise value for your company [106:16] because you've got these long durable [106:17] compounding revenue streams oh and all [106:19] your customers are growing so you have [106:20] the whole land and expand thing there [106:22] and [106:23] the thing underpinning it all is the [106:25] software itself has zero marginal costs [106:28] so you can bundle in all this stuff for [106:30] free because it actually doesn't cost [106:32] you anything I know it's Enterprise [106:34] software so it's not as sexy or exciting [106:36] or thought about as much as consumer [106:38] software but truly The Innovation that [106:40] was happening here was among the most [106:44] that has ever happened in a technology [106:46] company because Microsoft was figuring [106:48] all this out again these were not [106:50] lessons that people knew in the IB M era [106:53] that came before this in the Enterprise [106:55] there were no users Microsoft is now [106:58] figuring out how to build and sell [107:00] Enterprise technology systems in this [107:03] new era to businesses where there are [107:05] users of the technology and on the [107:08] business side yeah what you just said [107:10] like this is crazy Microsoft said okay [107:13] we're not going to just sell you the [107:16] software we're going to introduce this [107:18] thing called an Enterprise agreement [107:20] where you based on the size of your [107:23] company will pay us a certain dollar [107:25] amount per year per employee actually I [107:28] think it was per device but in these [107:30] days it was like you know most employees [107:32] just had one device and we've got you [107:35] covered everything that you would want [107:38] access to in our whole Suite of software [107:41] products inclusive of Windows and office [107:45] I mean it's not just you know a Salesman [107:47] comes to you and sells you Windows this [107:48] is Microsoft amortizing their go to [107:52] market C across all of their products [107:54] because when you show up at an [107:55] Enterprise you got lots of stuff to sell [107:57] them yep so now Microsoft has turned a [108:02] one-time sale of software into an annual [108:04] annuity that is going to keep growing [108:06] every year and is going to grow with [108:08] headcount and a key feature of the EA is [108:11] that it is a threeyear [108:13] agreement which means that you really [108:16] need everything to be aligned to pull [108:18] this off there's something pretty [108:20] convenient that you may have noticed [108:21] about Windows and office they both tend [108:23] to release an operating system or a new [108:26] package of office once every three years [108:29] or so and so every customer no matter [108:32] when they sign the agreement is [108:33] essentially guaranteed one upgrade [108:36] during their lifetime totally here's [108:39] something else that you get now as a [108:43] Enterprise it buyer in the Enterprise [108:45] agreement world with [108:47] Microsoft your [108:50] needs as it [108:53] are actually pretty different than your [108:56] users they're actually very different so [108:59] like if you are an employee of a large [109:02] company at this time you are using a [109:06] Windows PC at your office what are the [109:10] set of things that you want from that [109:12] device well you probably want to be able [109:14] to procrastinate you might be able to [109:17] want to play some games you probably [109:18] want to poke around the internet and you [109:20] definitely want it to be easy to use and [109:23] you definitely don't want restrictions [109:25] on there yeah you're willing to make [109:27] tradeoffs like uh if you can get a [109:29] little bit more efficiency but trade off [109:31] some security that's fine if you can you [109:35] know maybe use some pirated software but [109:37] it makes you better at your job that's [109:40] also fine you're acting with your own [109:43] agency not necessarily the company's [109:45] best interests in mind right you want to [109:47] run some VBA macros you know etc etc [109:51] okay now you are a corporate it [109:54] administrator and all of a sudden you [109:56] have to manage all these Rogue agents [109:59] all over your systems Rogue agents [110:01] called your employees you want the [110:04] ability to restrict your users from [110:05] doing what they can do you want to say [110:08] like no you cannot upgrade this software [110:10] without us doing it you cannot install [110:12] anything you cannot run these macros you [110:14] cannot visit these websites etc etc etc [110:18] one part of that is productivity but a [110:20] large part of that Ben as you said is [110:21] security yeah security privacy legal [110:25] compliance am I going to get hacked am I [110:27] going to get sued are we going to lose [110:29] data well Microsoft's got a beautiful [110:32] solution that they can sell you and with [110:33] the Enterprise agreement you can [110:35] customize all of this and we will give [110:37] you exactly what you want yes but now [110:40] David you're starting to expose a couple [110:42] features of Enterprise adoption which [110:47] have trade-offs if you're Microsoft oh [110:49] yes oh yes they do first of all if you [110:53] are a user you want the latest and [110:55] greatest software with all the most [110:58] Innovative features your it [111:00] administrator has a lot of incentive to [111:03] say I don't really want to go train [111:05] everyone on anything new so if the [111:07] software never confused anyone that's a [111:11] win even if it means we never get any [111:13] new features and so suddenly and I have [111:16] a direct quote from someone who is an [111:18] executive in office told me when I was [111:21] in office I always thought we could stop [111:22] bundling new features for 10 years and [111:24] it would be fine no one would notice I [111:27] think people would probably pay more for [111:29] it office got to this point where and I [111:32] think step soski even writes about this [111:34] in hardcore software in his book and on [111:36] his substack that at some point they [111:38] were trying to ship features that the [111:41] PMS thought were great and users would [111:42] love they would do this user research [111:43] they would hear that people want them [111:45] and then the Salesforce would run back [111:46] to them and say no no no no no no no do [111:48] not include that are you kidding me I'm [111:50] going to have all these objections in my [111:52] sale if you make me take this new [111:55] feature or take this ribbon or take this [111:58] you know any big UI change everything [112:01] has to be small and iterative and not [112:03] add any training or confusion I joined [112:06] the corporate Workforce in 2007 when I [112:09] graduated from college and I was an [112:10] investment banking analyst on Wall [112:12] Street at UBS I started Midsummer 2007 [112:17] and our corporate it systems my windows [112:22] laptop was so locked down we were using [112:25] XP of course in 2007 you were using XP [112:29] yes we were using Office 2003 of course [112:33] and like over everyone's dead bodies [112:36] would any of that change everything was [112:39] firewalled we could not access we [112:41] couldn't install anything we couldn't [112:42] access tons of websites I remember when [112:45] I first started we could still access [112:47] miniclip.com and so the analysts were [112:51] playing tons of games and like pretty [112:53] quickly it caught on and you know that [112:54] got the kibos [112:56] so yeah and I'm sure UBS as a customer [112:59] loved every single bit of [113:02] that the other big thing that you are [113:04] talking about which you were hinting at [113:05] with h VBA macros the key to Enterprise [113:09] is backward compatibility saying look we [113:12] don't necessarily need to promise you [113:15] anything to groundbreaking we need to [113:17] meet your needs today and be the most [113:20] cost efficient you know total cost of [113:22] ership driven system that meets your [113:26] needs and your employees are fine with [113:28] and from here on out everything's going [113:30] to stay compatible any modifications you [113:32] make Enterprise or software you use and [113:35] rely on we won't break no matter what [113:39] and we will continue to support those [113:40] versions you are using yes and [113:43] Enterprises love that and we're going to [113:45] put a pin in this right now and we're [113:47] going to bring it back toward the end of [113:48] this episode in an really illustrative [113:51] way that it can deep deeply deeply hold [113:53] you back if you are Microsoft and you [113:55] have built an entire brand and [113:57] reputation around your backwards [113:59] compatibility yep one stat and then one [114:04] point I want to make to highlight all [114:05] this by 2007 analysts estimated that 40% [114:11] for zero of all of Microsoft's Revenue [114:15] which I think was about $51 billion that [114:18] year so 40% of $51 billion came from [114:23] multi-year Enterprise agreements so [114:25] these threeyear agreements that you're [114:27] talking about Ben that covered windows [114:29] that covered office that covered all the [114:31] products that Microsoft offered except [114:32] like Xbox 40% of all dollars were [114:35] flowing from multi-year EAS and then [114:39] another 15% of all dollars that [114:43] Microsoft was earning as Revenue were [114:45] flowing from single year EAS wow so more [114:49] than half the company's Revenue 55% % [114:53] yes more than half the company's revenue [114:55] is all from this by 2007 so I mean it [114:59] was really the first seven years of [115:01] Steve's tenure as CEO yes already tipped [115:05] the balance into majority and the vast [115:09] vast majority of the rest of Microsoft [115:12] revenues you know the other 45% of the [115:14] company was the OEM Windows business [115:17] that was 30% so if you look at Microsoft [115:19] Revenue in fiscal 2007 50 55% is this [115:23] new Enterprise motion 30% is the old [115:26] windows business you know Dell and [115:29] Lenovo and whoever like you know selling [115:32] laptops to Consumers and paying [115:34] Microsoft for the operating system Y and [115:36] only 15% of the company's revenue is [115:40] anything else yeah it's funny I wasn't [115:42] going to bring this up here but since [115:43] you brought up oems the OEM business [115:46] model is completely transformational for [115:49] Microsoft when they figured out actually [115:51] we should be just selling software [115:54] directly to Consumers instead we should [115:55] be selling them to the PC maker and the [115:58] PC maker should do our distribution so [116:00] here's a couple stats in the 90s the Box [116:02] software that Microsoft would use to [116:05] sell Windows their gross margin on a [116:08] copy of Windows was [116:10] 29% oof that's not good they had to [116:13] print the disc which had actual real [116:16] cost especially on floppies you had to [116:18] put it in the Box you had to ship it to [116:19] the retailer you had to split profits [116:21] with the retailer you had to pay the [116:22] sales and marketing costs I mean it's [116:25] like real material C this is not a zero [116:27] distribution cost zero marginal cost [116:29] business in the Box software retail [116:32] world but when they're selling through [116:34] an oem Channel their gross margin was [116:37] 75% because you just ship the bits to [116:39] the OS once and then the PC manufacturer [116:42] and takes it from there not only is it [116:45] amazing because you get that 75% versus [116:47] 29% of gross margin it's also an amazing [116:51] way to scale because you do a deal with [116:54] every OEM you know as you're going down [116:56] the line it's the Visa networks of [116:58] networks thing that I think we alluded [116:59] to this last episode too you just get [117:02] each of them scaling on their own can [117:04] ACR to you without you doing additional [117:07] work to do the scaling yourself and so [117:09] David it's interesting you're talking [117:10] about how 85% of the business by 2007 [117:13] was either Enterprise sales of the EA or [117:17] OEM I mean they'd basically Kick the Can [117:19] to the curb on that crappy retail box [117:21] softare model and they're just doing the [117:25] whale hunting with their Salesforce and [117:27] doing these Enterprise agreements which [117:29] of course have great margin structures [117:32] and the oems and our annuities annual [117:35] annuities exactly way better business [117:37] model in every way they pivoted the [117:39] whole business to the two best ways to [117:42] sell software and completely eliminated [117:44] the bad way to sell software yes and one [117:47] of which they figured out post doj and [117:49] it became Yeah by 2007 over half the [117:52] revenue of the company which is crazy [117:54] now Ben when you said put a pin in a [117:56] minute ago and I know we're going to [117:57] come back to this after all the consumer [117:59] failures we're about to talk about there [118:02] is a downside to this when it becomes [118:04] your customer when you become an [118:06] Enterprise business the quality of the [118:10] software especially the user facing [118:12] software is no longer priority number [118:14] one and this wasn't a problem for the [118:19] company until in 2007 with the iPhone [118:25] but let's rewind and talk about [118:27] everything that happened in consumer [118:28] software at Microsoft until then yeah [118:31] what was going on with Windows releases [118:33] during that time and I think through [118:34] storytelling the windows releases we can [118:36] then understand the state of the company [118:39] so Windows XP why was Windows XP such a [118:42] big deal well it was a big deal [118:45] technologically it was a big deal for [118:47] users and it was a big deal because it's [118:50] pretty wild that Microsoft amidst all [118:53] the antitrust stuff we were just talking [118:55] about during the 1998 to 1999 the [118:59] rulings in 2000 the settlement proposal [119:02] in 2001 they developed and released an [119:06] operating system amidst all of that and [119:09] an awesome one yeah so what was Windows [119:12] XP technically well for the previous [119:15] better part of a decade they had two [119:18] parallel development efforts going on [119:21] there was Windows NT for the Enterprise [119:25] and there was Windows 9x you know [119:29] Windows 95 Windows 98 for consumers and [119:33] both of these had the same API that [119:36] developers could write their [119:37] applications for But ultimately the way [119:39] they were implemented the way [119:41] interoperability worked compatibility [119:43] worked user experience everything about [119:44] it was actually completely different [119:46] because it was a completely different [119:48] implementation of those apis and so the [119:52] knock against NT was always well you [119:54] need really beefy Enterprise grade PCS [119:57] to run it and it's not as nice and [119:59] intuitive and the knock against the [120:01] windows 9x call it 95 was that yeah it [120:05] looks pretty but it's not powerful I [120:06] can't actually do anything it was like a [120:09] friendly interface but not a powerful [120:12] set of functionality that came with the [120:14] operating system and so [120:17] XP did The Impossible where they figured [120:21] out how to to take the ease of use of [120:23] the 9x interface and make it run on top [120:28] of NT the whole thing is built on the NT [120:31] kernel and it has the friendly [120:34] approachable ease of use that you are [120:35] used to in Windows 95 and 98 amazing [120:40] amazing so the lineage of that 9x code [120:42] base that came all the way from Windows [120:44] 3.0 or maybe even one or two I don't [120:46] know how long code lived but interface [120:48] manager right exactly is now dead and so [120:52] you had the NT lineage of I guess maybe [120:55] even you could say it started with os2 [120:57] but Windows NT Windows 2000 and then [121:01] Windows XP so everybody's running XP now [121:05] there's two additions there's home and [121:07] there's professional oh got to get the [121:09] professional I always got the [121:10] professional did you every time I built [121:12] a new pc oh gotta go pro I didn't even [121:14] know what Pro meant I definitely didn't [121:16] need Pro because I was not a corporate [121:17] office worker but got to go pro it came [121:20] with all kinds of great stuff they've [121:22] got this great slide it's a fun [121:24] announcement to watch the emphasis on [121:27] digital photography digital music [121:29] digital video home networking it ushered [121:32] us into this age of you probably have [121:33] media that you're using on your computer [121:36] Apple famously owned this as a corporate [121:39] identity with their digital Hub strategy [121:41] but you know Windows XP plenty of people [121:44] were importing digital photos off their [121:46] camera to Windows XP that was a a sort [121:48] of big exciting use case for it lot of [121:51] Napster clients running on Windows XP [121:54] machines lot of Napster clients yes so [121:59] like I did for every Microsoft Windows [122:03] release I went and watched the [122:05] keynote the keynote is extremely strange [122:09] think about what a Steve Jobs keynote [122:12] was back in the day or what a WWDC [122:14] keynote is like today or a Google [122:17] IO this keynote opens with a gospel [122:21] choir singing America the [122:23] Beautiful and is followed by Bill Gates [122:27] and Rudy Giuliani walking out on stage [122:30] together and talking about how bad [122:34] terrorism is and of course the thing you [122:36] need to know about this keynote is the [122:39] date yes so this happens one month after [122:42] 911 in New York City and it really [122:46] underscores what a strange time it was [122:48] in the US if you had this once in three [122:51] years product release and it was going [122:54] to be in New York October of 2001 you [122:58] probably have this question should we [122:59] even do it should we make it all about [123:02] the First Responders it grounds the [123:05] whole thing in a very specific moment in [123:08] history when you're watching it in a way [123:10] that no other Tech event really ever has [123:13] been grounded in history before so a few [123:16] other things that jump out during the [123:18] keynote Bill Gates is not the CEO Steve [123:22] is and yet Bill Gates is the one walking [123:24] out with Rudy Giuliani to kick things [123:26] off and that's a strange and somewhat [123:29] telling element of what Bill's role at [123:31] the company was now you could argue he [123:33] was the public facing figure he was the [123:36] founder of the company it seems very [123:38] natural but also at some point why isn't [123:41] the CEO the one doing the keynote [123:44] another thing about Windows XP there was [123:46] a new release of office right at the [123:49] same time as XP this is a classic [123:52] Microsoft move they are able to create [123:54] great applications available on day one [123:57] which makes the OS more valuable and so [124:00] from the applications perspective [124:01] they're able to ensure that they get [124:03] great market share since they're always [124:05] adopting the latest and greatest Windows [124:07] platform right away so Windows success [124:09] begets office success and it's important [124:12] to remember that that worked for many [124:14] many years and if you remember back to [124:16] the last episode Lotus 123 and Word [124:19] Perfect smoked Microsoft in Microsoft's [124:23] own backyard during the Dos era [124:25] Microsoft's productivity apps did not [124:27] get real adoption in DOS which is crazy [124:31] so when they were making windows they [124:33] basically swore never again they ensured [124:34] that they were going to be very early [124:36] with applications on those platforms and [124:38] so as Windows took off office also got [124:41] huge market share and it's smart to [124:43] remember this lesson and carry it [124:45] forward for years maybe a decade but [124:47] again they may have been on this [124:49] strategy a few years too long [124:52] forever it kind of became gospel at [124:54] Microsoft so with Windows goes the [124:56] company and so you need to do things to [125:00] make sure that Windows is going to [125:01] continue to succeed because that is our [125:04] company's platform and livelihood it's [125:06] almost like the old Disney adage so with [125:09] animation goes the company and until [125:12] 2014 Microsoft felt the same way yep and [125:16] yes that is true for all the traditional [125:19] reasons in the XP time frame the reason [125:21] was also true in part one of our [125:24] Microsoft series it's even more true as [125:28] Microsoft becomes an Enterprise company [125:30] because Windows is at the heart of the [125:32] Enterprise agreement the whole value [125:35] prop of all of our server Technologies [125:38] is they work great with your windows [125:39] devices on your network right and so [125:42] there's strong incentives everywhere for [125:44] Microsoft to Ure that Windows is the [125:48] standardized platform that everyone [125:50] wants to have on their [125:52] PCS because it kind of makes everything [125:55] else work and so of course they're going [125:57] to release a new version of office that [125:58] shows off the latest and greatest of [126:00] Windows and I think this XP time frame [126:02] is the Showcase moment of when that was [126:06] a great strategy and we'll contrast that [126:08] later yep the other thing to know about [126:11] this XP time frame is last episode we [126:15] talked about the incredible secular [126:17] growth trend of the PC that was this [126:19] crazy Tailwind for Microsoft one of the [126:22] greatest Tailwinds you could ride in [126:24] business history PC shipments I believe [126:26] the stat David was that they grew 98% [126:30] per year over the 11 years between the [126:32] founding in 1975 and the IPO in [126:35] 1986 the crazier thing is even as late [126:39] as 2001 with Windows XP they were still [126:43] riding this Tailwind the US household [126:47] penetration of personal computers again [126:49] flashing back pre-ipo was only 8% so [126:53] that whole doubling year over year over [126:55] year by Microsoft's IPO they still only [126:58] got to around 10% of penetrating the US [127:02] by [127:03] 1997 13 years later it grew to [127:07] 37% and after a couple years of XP being [127:10] in Market 2003 it had grown to [127:14] 62% so I think the craziest stat is [127:16] actually that last one 2003 feels like a [127:19] modern moment in history but PC's were [127:21] still only in 62% of us homes wow that's [127:26] crazy the PC wave is just one of the [127:29] greatest secular Trends in history [127:32] particularly if you have a monopoly [127:34] share of that market yeah and they as [127:38] defined by the US government did Ben [127:41] Define define for me I there's just no [127:44] question of as this Market grows are you [127:46] going to be able to continue to [127:47] participate in it's like yeah we [127:49] basically are a tracker for that market [127:51] like if grows we grow with it yep and [127:54] now might be a good time certain [127:56] Microsoft fans have probably been [127:57] listening to this episode and gripping [127:59] their phones with all their strength [128:01] like when are you gonna talk about [128:04] Xbox we're going to talk about Xbox [128:06] briefly right now we will do a whole [128:08] another episode on Xbox someday maybe [128:10] David well make your case and then let's [128:13] talk about it my case is I love my Xbox [128:16] so well it's important to know Microsoft [128:19] didn't start in gaming with the Xbox [128:22] Windows 95 they shipped DirectX X that [128:25] changed the world they became a real [128:27] gaming platform because of that is this [128:30] unbelievably clever set of apis that [128:32] went entirely around windows amazing [128:34] piece of technology you put Microsoft on [128:36] the map and you have the whole rise in [128:37] PC gaming for the next six years even [128:39] before the Xbox yes totally that's funny [128:42] Microsoft is so huge that this is one of [128:45] the things that kind of gets lost to [128:46] history but you are absolutely right [128:48] DirectX was so important in that late [128:52] '90s era for PC gaming you know Quake [128:55] Counter Strike everything that happened [128:58] halflife later that was enabled you know [129:02] Doom came before and was really just [129:05] like the genius of carac as a programmer [129:08] to enable a first-person shooter to [129:10] happen on a PC Hardware without [129:12] something like direct X and Hardware [129:14] acceleration but yes everything that [129:16] came after that the birth of the first [129:17] person shooter genre huge story to tell [129:20] another day but you're right that leads [129:22] into Xbox and Microsoft's entry into the [129:26] Home console crazy that happened in [129:29] November 2001 so just like a couple [129:31] weeks after the XP launch it was a big [129:33] time for Microsoft and how crazy is this [129:36] they thought they were getting broken up [129:38] right as they're launching a video game [129:40] console and this operating system that [129:43] they've been working toward for like [129:44] eight years yeah this is also part of my [129:46] argument of Microsoft was such a [129:49] dominant consumer technology company [129:52] before doj because even though all this [129:55] stuff comes out right after it's the [129:57] momentum still from before that's [129:59] carrying Microsoft through to it yeah [130:02] okay while we're in Xbox land should we [130:03] finish our Xbox right now for this [130:06] episode sure Xbox has become an [130:08] important part of our world but not an [130:12] important part of Microsoft's business [130:15] agree David and I sort of heard people [130:17] utter things in our research like Xbox [130:19] has kind of been a lifetime Breaky [130:21] business or it's never meaningfully [130:23] contributed to Microsoft so I tried to [130:25] figure out as much as I could from [130:26] financial statements and I got to thank [130:28] Alex at the science of hitting it's a [130:30] great substack for helping me with this [130:32] if you look there was a division called [130:34] entertainment and devices that was part [130:37] of their old reporting structure and if [130:40] you look at the end reporting over time [130:44] let's start back in [130:46] 2006 they generated $4 billion doll in [130:50] Revenue lost $1.4 billion operating loss [130:55] so this is five years after the Xbox has [130:57] come out loss making 2008 they do 8 [131:02] billion in Revenue $400 million in [131:05] profit so like even as it's becoming a [131:08] real business at steady state yeah as [131:10] 360 is yeah coming into teeny margins [131:14] yeah totally 2009 8 billion in Revenue [131:19] 100 million in operating income 2010 [131:22] another 8 billion in Revenue 700 million [131:26] this is $700 million to Microsoft in [131:28] this time frame they do call it 20 [131:30] billion do of profit yeah what's 600 you [131:34] know 700 [131:35] million there's a great quote we'll [131:37] bring it up again later but we got to [131:40] talk to Steve again to bomber as we were [131:42] preparing for this episode and he had [131:44] this amazing quote to us about some of [131:46] his Acquisitions that didn't go well he [131:48] said we only lost money it's funny but [131:51] it's such an important point in the [131:53] context of Microsoft money is not the [131:56] scarce resource the scarce resource is [131:58] time and talent and focus yes David that [132:01] is exactly the right Point Microsoft [132:03] since year two or three has never been [132:05] Capital constrained and Bill Gates says [132:07] this in an interview anytime we've [132:09] thought about making an investment it's [132:12] just do we have enough talented people [132:15] to pull that off on any given year I [132:18] can't deploy all of the dollars [132:19] available as a CEO as a capital [132:22] allocator because I'm constrained by the [132:24] amount of smart people we have to pull [132:26] it off that is a much different position [132:29] that most businesses are in yes but is [132:33] absolutely the case not just for [132:35] Microsoft but for all the atale tech [132:37] companies these days you know the top [132:39] five market cap companies in the world [132:41] yep money is not the issue right in fact [132:44] you're making my point for me if I had [132:45] to make the case of why Xbox has been [132:48] somewhat of a Folly and perhaps not [132:51] worthy of a full acquired episode it [132:53] would be there was a lot of Microsoft's [132:56] best people worked on Xbox this is a [133:00] group of people that went and created [133:02] Xbox Live that by 2012 had 40 million [133:05] subscribers so people who built a core [133:07] competency of running a big online [133:10] service I mean these are some of the [133:12] best product people the Aesthetics of [133:15] Xbox from a physical perspective but [133:17] also the software I just think it was a [133:20] sinkhole of some of Microsoft's best [133:23] product people and just hardest working [133:26] people the culture in Xbox was so hard [133:28] driving to produce at least in this [133:31] point in history up to the 2010 time [133:33] frame very little in the way of [133:36] contributing to Microsoft's business but [133:39] soaking up a huge amount of the talent [133:41] imagine if that sort of product design [133:44] sensibility was deployed across the rest [133:47] of Microsoft yeah totally I think X Box [133:51] live is debatable we'll come back to [133:53] this with Azure yes Xbox Live was one of [133:58] the original pioneering internet [134:02] services subscription Services across [134:05] any category of software and technology [134:08] and the DNA and experience that [134:10] Microsoft built from that served it [134:12] extremely well I mean yes I think there [134:15] are two gigantic benefits look the [134:18] gaming Market is massive and important [134:21] and if you could try to own one market [134:23] in the world today in the world of [134:24] entertainment it's gaming I'm just [134:25] saying Microsoft didn't up until that at [134:28] least this point in history but it's the [134:31] right Market to go after they were not [134:33] successful in capturing value from it at [134:36] this moment in history but you're right [134:39] the two big things that they were able [134:40] to do is build out a core competency of [134:42] running a big online service which [134:45] totally led to Azure which we'll talk [134:48] about later and two it really did make [134:51] Microsoft relevant with a whole new set [134:53] of consumers when Microsoft was [134:55] completely irrelevant in their lives [134:57] should we talk about Vista [134:59] yes oh boy so there's a little taale off [135:03] of XP that'll lead to Vista we got to [135:05] talk about the code names too yes so the [135:08] Windows XP code name David was what [135:11] Whistler of course like the beautiful [135:13] ski mountain real close to Vancouver lot [135:17] of seattleites go there it's a favorite [135:18] of many a Microsoft employee and [135:21] blackhome the ski mountain right next to [135:23] Whistler is yes blackhome which became [135:26] the name for the theoretical release [135:28] that they wanted to do like just a year [135:30] or two after Vista we're going to follow [135:32] hot on the heels of that oh boy but [135:36] black home started becoming pretty [135:38] technically hairy so they decided to [135:40] push the date out another reason they [135:42] had to push the date out was Windows XP [135:44] for all of its usability and reliability [135:47] was very insecure and so Microsoft had a [135:50] whole thing where they thought they were [135:51] going to spend like 3 months putting out [135:53] a service pack they spent the better [135:55] part of two years iterating on Windows [135:58] XP to come out with a release that [136:00] really people at Enterprises could trust [136:02] as no viruses you know this is safe to [136:05] deploy in your Enterprise this was [136:06] Service Pack 2 I think was what [136:08] ultimately yeah Windows XP SP2 is the [136:11] stuff of Legend like that's the good one [136:14] so that pushes black holm's date out and [136:17] it also ties up a lot of the talent that [136:20] Microsoft needs to start working on the [136:22] Next Generation operating system which [136:24] again they thought was going to be a [136:25] fast follow for anyone who skied up [136:27] there there's this great Ski Lodge [136:30] Restaurant right between the two [136:31] mountains called the Longhorn Saloon yep [136:36] Longhorn baby that sounds like a great [136:38] name for a modest release to follow XP [136:41] before we get to the big hard changes [136:43] that are going to come in Black [136:45] home boy oh boy David the look on your [136:48] face so I remember being like a teenager [136:51] in high school at this point in time and [136:54] reading all about Longhorn blackhome all [136:59] this stuff on the internet you know on [137:00] these new tech sites these blogs like [137:03] this is g to be amazing I remember [137:06] downloading like new shells for Windows [137:08] XP to mimic the longhorn UI with the [137:11] sidebar and the clock on the side oh man [137:15] what a disaster well this was part of [137:18] the belief behind Longhorn they wanted [137:21] to Market all the cool stuff they were [137:24] doing for it through these sort of like [137:27] developer blogs and fan blogs even [137:30] though the product didn't have a ship [137:31] date yet and so everyone got really [137:34] well-versed in what was coming in [137:37] Longhorn and then everyone was kind of [137:39] sitting on their hands like where's [137:40] Longhorn they've been really telling us [137:42] about Longhorn in a way that you would [137:45] never see today no one's dripping out [137:47] the features of something that is [137:49] potentially still years away from a [137:51] release and ultimately then years go by [137:56] five years go by yeah could you imagine [137:59] if apple on their developer site Apple [138:02] were just like hey here's you know iOS [138:05] 23 here's all the great new features [138:07] we're building I mean the funny thing is [138:09] they actually kind of did that this year [138:10] with all the AI features all of those [138:12] are coming soon over the next year dot [138:15] dot dot which I'm not saying that's a [138:17] bad strategy in the current environment [138:18] but it is a different strategy for Apple [138:21] anyway Longhorn is teased for 5 years [138:24] all the David rosenthals out there are [138:25] kind of like what the heck Microsoft [138:27] I've been excited for all this crazy [138:28] stuff you're showing me what's going on [138:31] well what happened behind the scenes [138:34] David what was the initial [138:39] technical spark that was supposed to be [138:43] the [138:44] Cornerstone of Longhorn well there were [138:49] three pillars of I think it was all [138:51] originally supposed to be black home and [138:53] then they were like no no we're going to [138:54] pair it down to Longhorn but it all [138:55] ended up getting added back into [138:58] Longhorn the first of which was called [139:00] Avalon and it was a new graphics engine [139:03] that used direct Hardware acceleration [139:06] so I think the vision for this was kind [139:07] of like Hey we're going to take direct X [139:09] and bake it into the operating system [139:11] and allow the operating system to use [139:14] GPU Hardware acceleration yeah that's [139:17] more or less it all these code names [139:19] ended up referring to multiple things [139:21] because it was emblematic of the [139:23] organizational disarray inside the [139:25] windows development team but anyway it [139:28] sounds great right we can render all [139:30] these really great graphics as a part of [139:32] the operating system because it's GPU [139:34] accelerated who doesn't want better [139:36] graphics of course right the thing that [139:38] ultimately happened is the oems were all [139:41] trying to make netbooks and so they're [139:42] furious at Microsoft about saying the [139:45] next new release of Windows which is 5 [139:47] years since Windows XP they really [139:50] really counting on a new version of [139:52] Windows to drive PC sales and the one [139:54] that they're getting requires pretty [139:57] good gpus like a gaming PC yeah so it [140:01] was like a kind of a total Miss with [140:03] what their OEM Partners were looking for [140:06] but if you did buy a nice PC and you did [140:08] eventually end up with a copy of Windows [140:09] Vista this is why you got to see the new [140:13] what uh what did they call the oh the [140:16] arrow interface Arrow that's right the [140:18] blue shiny sort of thing that was like [140:20] kind of ripping off Mac os's [140:24] Aqua I mean call Spade is Spade over [140:27] here we can tell where your true loyalty [140:29] lies well I'm just saying like if you [140:30] run a company where you make all your [140:32] own hardware and your own software then [140:34] it's much easier for you to Hardware [140:37] accelerate all the graphics in the [140:38] operating system but when you're [140:39] counting on OEM Partners you need really [140:41] good communication there yes yeah that [140:44] was one the other one was a new web [140:47] services framework called Indigo which I [140:50] don't know I did a lot of research and I [140:51] couldn't figure out what it was supposed [140:52] to be I think it was kind of a fever [140:55] dream of like let's stuff the internet [140:57] fully into Windows ultimately there was [141:01] a very fundamental architecture shift [141:04] that just did not pan out if you [141:05] remember from the last episode we talked [141:07] about Chicago Windows 95 and Cairo this [141:10] theoretical thing that never shipped [141:12] that was going to be the next Generation [141:13] operating system well they basically [141:16] they tried it again did the same thing [141:18] again yeah I think Bill Gates was a big [141:20] fan of this Vision because it was really [141:21] technically ambitious where they had an [141:24] objectoriented file system where the [141:26] file system could specify data types and [141:30] then every application would plug [141:34] directly into the data types that the [141:38] file system knew about so there was [141:39] these sort of Standards like a calendar [141:42] invitees a calendar invite and the [141:44] operating system has its own fields for [141:48] date and time and notes and that means [141:51] you're not always traversing directory [141:52] trees whenever you're trying to search [141:54] through stuff and also it meant that the [141:56] operating system could actually reach [141:58] into the data within files that were [142:01] being stored by applications so it was [142:03] sort of a standard way of storing files [142:05] in an easy to search way and what you're [142:07] talking about here is the third pillar [142:09] of Longhorn blackhome which is winfs [142:12] right yes ultimately [142:15] wifs they tried to build it many times [142:19] there was a lot of offsites and [142:22] architecture reviews and talking about [142:26] how great it was when in practice there [142:28] was never any pull from application [142:30] developers that they wanted this in the [142:32] first place and this was a huge part of [142:35] the wheel spinning of well we can't do [142:37] all this other stuff in the operating [142:39] system until we figure out the spec and [142:42] the implementation for [142:44] wfs and then once we have that then we [142:46] can start to do all this other stuff [142:48] part of the other stuff was the [142:50] development team wanted to bake net [142:53] directly into the bits of the operating [142:55] system that shipped in the box and on [142:57] your PCS so net was everywhere so [143:00] ultimately what happened here is and I [143:03] heard this from a developer there was [143:06] many different groups who were all [143:07] compiling their own sub projects and [143:09] they could sort of run them but when it [143:10] came time to try to actually do a build [143:13] of this operating system and say hey [143:16] we've had too many offsites and [143:17] architecture reviews and restarts and [143:19] this is in this is out let's just try to [143:21] do like a build of the OS that we could [143:23] deliver they never built Longhorn like [143:26] it did not compile they could not [143:28] integrate all the different projects [143:31] into one and they ended up [143:33] reking from an old Windows Server [143:36] version or something and adding things [143:39] in one by one piece meal to try to [143:42] figure out you know in year four how can [143:45] we get something shippable out to [143:48] Consumers so we can say this is our next [143:50] generation operating system and what is [143:52] the minimum acceptable set of stuff that [143:54] we can put in such that it looks and [143:56] feels new yep so okay Longhorn vist this [144:02] is truly a disaster for the company well [144:05] so 100% it is but they were trying to [144:09] talk about it like it wasn't so I [144:11] watched the launch announcement for this [144:13] too they kind of have to they can't [144:15] really say like nobody should upgraded [144:17] this so they come out first of all it's [144:19] Bill Gates Again in 2006 six years after [144:22] Steve Balmer has become [144:24] CEO my opinion on this is they clearly [144:27] had no idea what to talk about in the [144:28] keynote because the one feature that I [144:32] can kind of really remember as a [144:33] flagship feature is that alt tab [144:35] switcher that was 3D that kept bringing [144:37] the windows closer and closer and closer [144:39] to you you know they've got the widgets [144:40] they've got the sidebar it's Arrow they [144:43] had one feature that people hated there [144:45] was a Revolt called user access control [144:48] which the theory makes sense protect [144:50] users from running malicious blah blah [144:52] blah but in practice it would just [144:54] overwhelm you with dialog boxes all the [144:56] time and everyone's just trying to [144:57] figure out how do I turn off the dialog [144:58] boxes so they're standing up there at [145:01] the keynote the whole thing the [145:05] marketing message is the wow starts [145:08] now oh boy oh boy it's a completely [145:11] incohesive incoherent set of things [145:14] they're launching consumers didn't like [145:17] it businesses tried not to upgrade even [145:19] as late as 20 [145:20] 9 so three and a half years after launch [145:24] something like that three qus of [145:25] corporate PCS are still running XP and [145:29] had never upgraded to Vista oh it's even [145:31] worse than that you may have this in [145:32] your notes but Microsoft oems were so [145:36] unhappy because consumers didn't want to [145:37] buy Vista machines Microsoft had to [145:40] extend the ability for their OEM [145:43] Partners to keep selling XP machines to [145:46] Consumers for another two years after [145:48] this just brutal this was kind of the [145:51] windows culture at its worst I worked in [145:53] office so I have a bias here when I was [145:55] at Microsoft but they weren't super ship [145:57] date driven whereas office would set a [145:59] ship date three years in advance and [146:01] then they would hit it exactly office [146:03] had all these really robust procedures [146:06] for shipping you know a triage process [146:09] an escalation process a zero bug bounce [146:12] everything was run in this Dev test PM [146:15] Triads the excuse was this General guise [146:18] that this is too hard to use your [146:21] processes like we're doing Alchemy over [146:22] here and because we're doing systems [146:25] level programming none of your software [146:28] development principles work on us and so [146:32] ultimately this was the failure mode of [146:36] a process that really did work for a [146:37] while really did enable technical genius [146:39] really did enable solving hard computer [146:41] science problems and this is effectively [146:44] the company smoking their own Supply and [146:46] just believing they were smarter than [146:47] everyone else and what consumers wanted [146:49] didn't matter and if they could come up [146:51] with some hallucinated cool technical [146:54] thing then that is what they should [146:55] spend years doing and fighting about and [146:58] then Force into the market and the [146:59] market just didn't take it one bit yep [147:03] and a couple other things on this one so [147:07] when Vista actually shipped just sort of [147:09] you know [147:10] process-wise it was Ben as you're saying [147:12] a complete reset so Brian Valentine [147:15] comes over from exchange you know in [147:17] Windows server in the Enterprise world [147:20] to take over managing getting something [147:23] out the door and just cut all the [147:25] features cut all the pillars of the [147:27] windows Longhorn Vision it still takes [147:31] two years in that process to get it out [147:34] and then immediately afterwards Brian [147:36] leaves the company lots of other great [147:38] Engineers leave the company too they go [147:40] down the street to Amazon and then Brian [147:42] ends up leading the entire engineering [147:44] platform team for [147:46] amazon.com wow oh I didn't know that's [147:48] where Brian went interesting [147:50] y yep he went to Amazon he was like a [147:53] named top senior level executive at [147:55] Amazon for a long time wow you know and [147:58] then the other thing about this whole [148:01] process that this is purely my own [148:03] speculation like nobody said this but [148:05] just as I've been thinking and [148:07] reflecting on how seminal a moment the [148:11] antitrust stuff was to Microsoft after [148:16] the height of their consumer power right [148:19] beforehand [148:21] I think this might be a case where Bill [148:23] no longer being CEO and just being Chief [148:26] software architect really impacted this [148:29] process when you're a CEO you have to [148:31] engage with your OEM Partners you have [148:33] to engage with Enterprises you have to [148:35] engage with customers and not that it's [148:37] all Bill's Fault by any means but this [148:41] Longhorn black home [148:43] disaster then as you say was a case of [148:45] getting high on your own Supply within [148:47] the company and if he's only spending [148:48] his time on technical decisions you need [148:50] some introduction into that feedback [148:52] loop some governor on how deep to go in [148:55] rearching Windows for rearching [148:57] windows's sake right and you know [149:01] remembering back to part one too it's [149:03] not just that bill was a great engineer [149:05] he was a great business person one of [149:08] the greatest of all time he trained from [149:10] birth it's like what company am I going [149:12] to be CEO of right the issue with [149:15] Microsoft is that there is only one Bill [149:17] Gates Bill was the best engineer bill [149:19] was the best lawyer bill was the best [149:22] deal negotiator to figure out what the [149:23] right BD situations were bill was not [149:26] the best Enterprise relationship Builder [149:29] I don't think Bill had a passion for [149:31] empowering the Enterprise and you know [149:33] making sure that businesses succeeded [149:35] the way that Steve was but nobody should [149:38] ever sell Bill Gates short and say he [149:40] was just the technical genius that would [149:42] be wrong totally yeah anyway this is [149:46] really bad at the same time as Vista [149:50] actually is coming out in late [149:52] 2006 this is when Apple starts running [149:55] the Mac versus PC ads which is just [149:58] brutal and oh boy if you're Microsoft [150:02] does that hurt and like mac sales are [150:04] irrelevant even today in 2024 Max sales [150:08] are 8% of the market and what were they [150:10] at this point in time like two or 3% [150:13] zero I don't even know like you know it [150:15] doesn't matter but the point is not that [150:18] apple is taking massive am of market [150:20] share from Windows it's that they are [150:23] hitting a nerve with consumers with [150:26] Enterprises and within Microsoft itself [150:29] most importantly of shoot we are way [150:31] behind here and the part that really [150:34] hurt about all those Mac versus PC ads [150:36] there so many were just straight up true [150:39] oh you know I'm a PC and I crashed again [150:42] to be in the Halls at Apple when they're [150:44] firing at all cylinders Steve Jobs is [150:47] back the iPod was a Smash Hit you're [150:50] developing the digital Hub strategy Macs [150:53] are starting to sell because of that [150:55] your iPod attach rate with Max is [150:57] actually working people are buying Mac [150:59] Mac are becoming the option that [151:01] students are starting to pick as they're [151:03] picking their college computer Market [151:05] Shar is rising and Microsoft comes out [151:08] with Vista you just have to be besides [151:10] yourself with this gift you've been [151:12] giving like oh my God look at this [151:13] opening here's where it mattered and the [151:16] timing mattered so much too this sets [151:19] the stage for the iPhone because Apple [151:22] now in STK contrast to Vista and with [151:25] these ads is training consumers with the [151:28] benefits and the joy of it just works [151:31] yeah and what did the Mac do it just [151:33] worked and what did the iPod do it just [151:35] works and what did the iPhone do it just [151:37] worked right yeah because there was this [151:40] pent up demand I remember people in 2005 [151:42] and six when the rumor started there was [151:45] this almost like glint in people's eyes [151:46] what if Apple made a phone wouldn't that [151:49] be awesome it is remarkable like the [151:50] iPhone delivered on all that promise but [151:52] there actually was wow what if we had [151:55] technology as good as the stuff that [151:57] Apple makes in the form of a phone [151:59] wouldn't that be great because phones [152:01] are so crappy so I think yeah you're [152:03] right there's something there there was [152:04] a training of associating the Apple [152:06] brand with well it was really a setting [152:08] of the Apple brand promise at this [152:10] moment in time yes that's a great point [152:14] and I think we got to call it here the [152:17] death of Microsoft as a relevant [152:19] consumer techn company they never [152:21] recovered from this as a leader well [152:24] yeah I think that's correct there's a [152:26] lot to talk about in their consumer [152:28] technology offerings I also think this [152:31] is the death of Microsoft as an [152:33] interesting platform for [152:35] developers who is writing Vista apps the [152:38] win32 API as a potential Target for my [152:42] new interesting Innovative application [152:45] it's just not a thing anymore you have [152:47] to write a Windows desktop app at this [152:49] point in history [152:50] because it's where a bunch of the users [152:52] are if you need a desktop app for real [152:55] for real but probably you're just [152:57] writing a web app you've lost developer [153:00] hearts and Minds which is the path to [153:02] losing relevance I think with one [153:04] exception I think you were totally right [153:06] nobody is writing Vista apps but the [153:08] only people left who are writing Windows [153:10] apps period are Enterprise developers [153:12] writing custom software for Enterprises [153:14] y that's a great point that is the [153:16] exception and of course anyone that had [153:19] big applications for mac and windows so [153:21] Adobe is a great example of they're [153:23] going to keep that up forever but where [153:26] these new disruptive software players [153:28] are coming from they're just not gonna [153:30] have Windows apps Facebook is not [153:32] writing a uh Windows vist app [153:36] correct so the biggest things to hurt [153:39] Microsoft coming out of Vista are what [153:42] we just talked about losing developers [153:44] what we just talked about losing users I [153:47] mean consumers who are excited to buy a [153:49] computer they're just not excited to buy [153:51] a Windows Vista PC but the biggest thing [153:54] is they lost years of their very best [153:58] talent I mean Vista was a black hole as [154:01] it just kept growing and growing growing [154:02] it would suck in more teams and as it [154:04] sucked in more teams you would get the [154:06] talent that it would suck in but then it [154:08] also would suck in executive and [154:11] distinguished engineer Talent from [154:12] elsewhere to come fix it and so [154:16] Microsoft is about to be in a place [154:17] where they need to compete and [154:19] understand a change landscape in Social [154:21] in Mobile in search they still have to [154:25] fight the browser War I mean IE is peing [154:27] and about to start falling off a cliff [154:29] and are completely consumed by Vista so [154:32] I think a lot of the consumer stuff can [154:35] be answered by Steve Balmer wasn't [154:38] really a consumer oriented technologist [154:41] that seems fair I think that is true [154:43] sure true but that's not the whole [154:45] answer Vista consumed a bunch of the [154:47] smartest people even if they had the [154:49] right Vision to chasing and the doj had [154:52] just crippled the culture among many [154:54] other things and they were still [154:56] recovering from that 100% in that [154:59] conversation that we had with Steve [155:00] where he you know made the comment about [155:03] my Acquisitions my mistakes we just lost [155:05] money on the bad ones the Genesis of [155:08] that conversation was about Vista you [155:09] know he was reflecting he said that [155:11] probably was the worst moment actually [155:15] right in my tenure as CEO because all of [155:18] that best talent everything you just [155:19] said Ben it was off the field it wasn't [155:21] playing it was out of commission right [155:24] money is not a scarce resource so bad [155:26] Acquisitions whatever who cares it's [155:28] just money but consuming a huge [155:29] percentage of Microsoft's most talented [155:31] Engineers that's company killing y I [155:34] mean hell even taking Brian Valentine [155:36] off of exchange exchange was freaking [155:38] killing it in the Enterprise and he goes [155:41] and spends two years getting Vista out [155:44] the door and then goes to Amazon o that [155:47] sucks brutal one other mic Micosoft [155:49] exact put it to me it hurt so bad that a [155:54] bunch of our best systems people were [155:57] leaving the company driving across the [156:00] lake going to work for an online book [156:02] seller and then building that online [156:05] book seller into the market leading [156:09] Enterprise compute company that is a [156:12] black eye right there yeah oh I can't [156:16] wait to talk about that okay Ben I'm too [156:18] excited for azer let's do search let's [156:20] do Mobile Windows 8 Zoom let's get all [156:23] that oh yeah and then let's talk Cloud [156:25] baby and David unexpectedly there is a [156:27] through line through all of them there's [156:29] a cohesive story that leads to Azure [156:31] here oh yeah but before we do that this [156:33] is the perfect time for another one of [156:35] our favorite companies and longtime [156:37] acquired partners pilot.com for startups [156:40] and growth companies of all kinds pilot [156:42] handles all of your companies accounting [156:44] tax and bookkeeping needs and is by far [156:46] the largest startup focused accounting [156:48] firm in the entire us also as on our [156:51] first Microsoft episode we have to give [156:54] pilot CEO Wasim daher a special shout [156:57] out here because he is the only acquired [156:59] sponsor CEO who is also a research [157:02] source for us on the same episode back [157:05] when Wasim was a student at MIT he [157:07] interviewed Bill Gates for the school [157:09] paper and he dug it up and sent it to us [157:11] Bill talked about Microsoft's fourth [157:13] coming 40 gbyte portable media center [157:17] and it was a pretty fun time cap we will [157:20] link to the PDF in the episode sources [157:22] yep so back to Pilot and speaking of [157:24] Bill we talk all the time on acquired [157:26] about one of his Seattle neighbors Jeff [157:28] Bezos and the AWS inspired Axiom that [157:31] startups should focus on what makes your [157:33] beer taste better in other words only [157:35] spend your limited time and resources on [157:38] what's actually going to move the needle [157:39] for your company for your product for [157:40] your customers and Outsource everything [157:42] else that you need to do that doesn't [157:43] fit that bill and accounting is just [157:46] example number one of this every company [157:48] needs it it needs to be done by a [157:49] professional you don't want to take any [157:51] risk of something going wrong but at the [157:52] same time it has zero impact on your [157:54] product or customers yep so enter pilot [157:58] pilot both sets up and operates your [158:00] company's entire Financial stack so [158:03] Finance accounting tax even higher level [158:05] CFO surfaces like investor reporting [158:08] everything from your general ledger all [158:10] the way up to budgeting and the [158:11] financial sections of your board decks [158:13] and they've been doing this for years [158:14] across thousands of startups in Silicon [158:17] Valley and elsewhere there's nobody [158:18] better who you can trust to both get [158:20] your Finance right and make it easy and [158:23] painless for your company and when you [158:24] say thousands of startups pilot has done [158:27] this for open AI air table scale as well [158:30] as large e-commerce and other companies [158:31] so it's not just that they have [158:33] experience across startups they can also [158:35] keep working with you as you scale to [158:37] the growth phase and Beyond so if your [158:39] company wants to start focusing on what [158:41] makes your beer taste better go to [158:42] pilot.com Acquired and tell them that [158:45] Ben and David sent you thank you to [158:48] Pilot okay so search and the alternate [158:53] title of this chapter could be an [158:55] acquisition that wasn't I think that's a [158:58] uh lost to history moment is the [159:01] acquisition that almost happened here [159:02] for $47 billion oh okay so I'm just [159:07] actually curious do you know the company [159:09] I'm referring to do you know the the [159:10] deal well it's funny the way you phrase [159:13] that I'm thinking like oh did Microsoft [159:14] try and buy Google and I don't know [159:16] about it but the number of course you're [159:18] talking about Yahoo yes yes okay so [159:21] let's set some context before we get to [159:24] this 2008 Yahoo attempted acquisition so [159:28] there were two companies that had [159:29] developed programmatic advertising [159:32] technology to serve and Target online [159:35] ads especially in search there was [159:37] double click the market leader and there [159:40] was a quantive Microsoft had lost the [159:43] doubleclick acquisition to Google they [159:45] bought a quantive and that didn't go [159:47] well it was $7 billion and they ended up [159:50] declaring basically the whole thing of [159:51] write off so Microsoft is desperate for [159:55] search market share and between their [159:57] internal efforts with MSN search and I [160:01] believe it was called Windows Live [160:02] search yep they were not making much [160:05] progress there and at the same time [160:09] Internet Explorer had totally languished [160:11] Microsoft had completely taken their eye [160:14] off the ball of the browser Wars from 10 [160:16] years earlier and iie was just widely [160:19] regarded as a garbage browser and web [160:22] developers hated it because it made you [160:24] write a bunch of weird custom stuff so [160:26] randomly things wouldn't work in IE [160:28] users hated it because basically nothing [160:31] new was coming every time a new version [160:33] of the operating system would ship it [160:34] just felt like it's the same old [160:36] Internet Explorer over and over again [160:38] and you have Firefox coming on the scene [160:40] starting around 2007 where it was really [160:42] making a dent and Google was the default [160:45] search from Firefox y Firefox was [160:49] awesome it had tabs iie didn't have tabs [160:52] at the time that's right oh my God you [160:55] know Safari I don't think Safari had [160:56] tabs either Chrome wasn't a thing yet [160:59] and so I know I'm on the one hand [161:01] talking about search on the other hand [161:02] talking about the browser but it's the [161:04] same pot of gold but it's the same thing [161:06] right yeah it turned out search was the [161:08] business for the browser yes so the [161:11] thing that you kind of have to realize [161:14] is the [161:16] browser is the front door to search [161:20] search is heavily heavily monetizable [161:24] and if you're Google and you can [161:25] monetize it directly that's great but [161:28] let's say you're not Google let's say [161:29] you're Firefox or Microsoft or apple and [161:33] you don't have this [161:35] incredible business model of people [161:37] bidding on the keywords for search and [161:39] all the R&D to go into making search [161:41] good but you actually do have the user [161:43] attention the front door well you get to [161:45] monetize it too the rumors are that [161:47] Apple makes something on the order of [161:49] $20 billion a year today in 2024 from [161:53] Google as being the front door to Google [161:55] sending all of the iPhone search traffic [161:58] to Google this is the traffic [162:01] acquisition cost in Google's financial [162:03] statements absolutely and so if you can [162:05] be in the business of operating a scale [162:08] search engine or you can be in the [162:10] business of directing traffic to a [162:12] scaled search engine who is willing to [162:14] pay you for that traffic it's going to [162:16] be a great business so David as you just [162:18] said the way to monetize the browser is [162:22] owning and operating or directing to a [162:24] search engine so search isn't going well [162:27] at Microsoft at first it was sort of [162:30] because they just didn't take it [162:31] seriously enough when Google first [162:33] started in 1998 I think there was a lot [162:36] of skepticism that the auction-based [162:40] advertising business would really work [162:43] and then there was skepticism that it [162:44] would really scale and then when it went [162:47] public people are sort of looking at it [162:49] freaked out at how profitable it was and [162:52] then even after that people didn't [162:53] really realize that being the market [162:56] leader at search was way better than [162:58] being number two there's these massive [163:00] massive returns to scale and the reason [163:03] for that is just pure Marketplace [163:04] liquidity if you have the most searches [163:08] you can create the best data from the [163:10] searches and you can return all the best [163:12] results because you have the most data [163:14] and on the advertiser side you have the [163:16] most advertisers who are willing to come [163:18] in and bid to the highest possible price [163:21] you just get to make the most money by a [163:23] country mile versus other search engines [163:27] and then it locks in even further [163:28] because you can spend more cacks on the [163:30] data centers and more on R&D and make [163:31] the search better and more performant [163:33] and faster and all that Google is search [163:35] as monetized by a ad-based auction is [163:39] one of the world's true Marvels it's one [163:42] of capitalism's greatest discoveries we [163:44] may or may not do an episode on Xbox [163:47] someday Ben that's for Ben and I to [163:48] discuss you know privately for parents [163:51] to discuss after the kids go to bed but [163:53] we're definitely going to do an episode [163:54] on Google yeah it's criminal that we [163:56] haven't so Microsoft is really nowhere [164:00] to be seen in Search and part of it was [164:03] just thinking oh well search is just a [164:05] feature of MSN but there's all these [164:07] other reasons to come to MSN or hey this [164:10] is a product in the portfolio of Windows [164:12] Live and you know we can kind of do it [164:14] with the talent that we have here [164:16] ultimately someone needed to grab [164:19] leadership at Microsoft early 20022 2003 [164:22] shake them and say nothing else matters [164:25] in the next 5 years except you figuring [164:29] out how to meaningfully participate in [164:32] search Revenue because that is just the [164:34] next big wave in technology and it's a [164:37] fantastic business yeah you needed the [164:41] equivalent of the J Allard Windows the [164:43] next killer application on the Internet [164:45] or the sinowski cornellis wired memo so [164:49] in 2008 Microsoft puts a deal on the [164:51] table that gets bid all the way up to [164:53] $47 billion to buy Yahoo this was [164:58] effectively their last Hail Mary to [165:01] become relevant in search they actually [165:04] didn't launch Bing until 2009 Google was [165:07] started in [165:08] 1998 and went public in 2004 and [165:13] Microsoft got serious about a branded [165:15] search engine in [165:17] 2009 but clearly before that they're [165:20] starting to realize this is a big deal [165:23] we need to participate in it you know [165:24] what do we do so they'd been negotiating [165:26] to buy Yahoo 2008 after a bunch of [165:29] negotiating and flying back and forth [165:31] finally both David Philo and Jerry Yang [165:34] fly up to Seattle and Steve Balmer goes [165:37] to Boeing field and they have a meeting [165:39] at the airport this is one of the great [165:41] what if scenarios this feels like a [165:44] episode of Entourage right it's totally [165:47] right there's conflicting report of what [165:49] happened from what I can tell Bill and [165:53] Steve kind of looked the Yahoo guys in [165:56] the eye and [165:58] decided these guys are kind of jerking [166:00] us around they really don't actually [166:02] want to be a part of Microsoft at all [166:04] and this has gotten so expensive that if [166:07] we execute the transaction or God forbid [166:10] they even try to negotiate up even [166:13] higher it's just not going to go well [166:15] because it's going to be an organ [166:16] rejection here so the deal complet [166:18] completely falls apart it's interesting [166:20] to try to look at the deal and figure [166:22] out even at that high price of $47 [166:24] billion was it a good deal for [166:27] Microsoft so here's how to pull it apart [166:30] I'm laughing here I was hoping I could [166:32] surprise you and be at the end of it [166:33] like have something to say here but I [166:36] think you're going to take my thunder [166:37] let me know when I do and tell me if it [166:39] rhymes with schib [166:42] Baba yes yes yes it's because I'm just [166:46] loving I'm smiling the whole time I'm [166:47] imagining you know all of you list being [166:49] like $47 billion for Yahoo what are they [166:53] smoking keep going so here's how to pull [166:56] it apart Yahoo had about 15% market [166:59] share of search which I think was number [167:01] two Google was way way way ahead and so [167:05] on the face of it you're thinking wait [167:06] $47 billion to buy 15% market share in [167:10] search but there's actually two other [167:13] Assets in here there's Yahoo Japan yep [167:17] and there's a state [167:19] in Alibaba not just a steak which [167:22] famously is one of the greatest [167:24] Investments of all time [167:26] 40% stake [167:28] 40% of Alibaba so collectively those two [167:32] assets together are worth over 30 [167:35] billion so if you back it out it's [167:37] really only like $15 billion to buy 15% [167:41] of the search Market yeah David what is [167:43] Google's revenue today alphabet's annual [167:46] revenue in 2023 was over $300 [167:51] billion so would you want to pay a [167:55] billion dollar per percent of market [167:58] share of that market yeah sure it's just [168:01] money why not I mean it's the craziest [168:03] thing this would have been ludicrously [168:05] profitable to spend only $15 to buy 15% [168:10] of the search Market which is way bigger [168:13] than100 billion and still growing this [168:16] is the thing people just always continue [168:19] underestimate and underappreciate about [168:21] the search Market is it's just so large [168:23] and so profitable yes now because this [168:26] is counterfactual and we actually don't [168:28] know what would have happened Yahoo [168:30] completely went away they sold for5 [168:33] billion and their most recent [168:34] transaction to be co-owned by Verizon [168:37] and Apollo so there's this real question [168:39] of like okay if Microsoft bought all [168:41] that traffic would they actually have [168:43] been able to harness it and build a [168:44] Google like business or would it have [168:47] just gone the way that Yahoo is going to [168:49] go anyway but to make the bull case on [168:52] that Bing is a good business it just has [168:55] a small market share Microsoft succeeded [168:58] finally in 2009 at attracting all the [169:00] right talent and taking it really [169:02] seriously and building a super viable [169:06] search engine that does I don't know [169:08] something like a billion a year in [169:10] profit well you know then they do go on [169:13] in launch bang and they actually sign a [169:15] commercial deal with Yahoo to provide [169:17] the search on the back so they're not [169:18] getting the Yahoo traffic yaho is still [169:20] monetizing the traffic but Bing is [169:22] getting all the data from doing [169:25] performing the searches for Yahoo and I [169:28] know you know this but most listeners [169:29] will not remember you know who the [169:32] leader of Bing was for a brief period in [169:34] its early days here Satia Nadella Satia [169:37] Nadella that's right other piece of SAA [169:40] trivia that I very much suspect you do [169:42] not know SAA joined Microsoft first in [169:46] 1992 from Sun micro and he joined as an [169:49] evangelist for Windows NT he joined as [169:52] an evangelist an evangelist H and then [169:55] he got his first product job do you know [169:59] what product it would never ship but a [170:02] product in development that job was o no [170:05] I don't tiger server the cable soft [170:10] information Super Highway fever dream [170:14] where did you find that it's in his book [170:17] oh my gosh it's in hit refresh was he [170:19] actually I always thought he was in the [170:20] marketing side well it's a little [170:22] Microsoft product managers are marketers [170:24] they don't live in the engineering org [170:27] yeah fascinating there's so many things [170:29] about his history that are not a part of [170:30] the common narrative like he I think [170:34] worked in Dynamics their Salesforce [170:37] competitor their CRM for a while which [170:39] they had Acquired and then he I think [170:42] ran bis talk server which was another [170:44] one of these Enterprise server products [170:46] and then Bing wow and then they plucked [170:48] them out a bing to go run Azure right [170:50] and go run server and tools which oh we [170:52] we will get to that all right pause [170:55] pause okay some of the other fun tidbits [170:58] of this Yahoo deal so Bing powers Yahoo [171:02] search Microsoft does the ad sales for [171:06] both sides so while Microsoft doesn't [171:08] get the user you know they don't have [171:09] the direct relationship with the users [171:11] they do get to build up their [171:13] Marketplace liquidity on The Advertiser [171:16] side and as you mentioned there's a huge [171:20] data advantage of actually powering the [171:23] search while Yahoo gets 88% of the [171:26] revenue in the deal for the first five [171:29] years arguably the value from this [171:32] basically all acred to Microsoft because [171:35] they ended up building out not only a [171:37] proper advertising business which now is [171:41] used on a number of different sites I [171:42] think even in partnership with Netflix [171:44] for their ad supported tier but also [171:47] being [171:49] once it had all the Yahoo traffic needed [171:51] to be a scaled web service yes like a [171:56] distributed computing system that [171:59] operated at 247 uptime with super low [172:03] latency fast response time and huge [172:05] scale God that sure sounds like the [172:07] cloud it sure sounds like the cloud so [172:10] you've got Xbox Live where they have 40 [172:13] million users you have a scale search [172:16] engine which is like the number one most [172:19] difficult distributed computing problem [172:22] that if you get good at that you can get [172:23] good at lots of other stuff the [172:24] ingredients are really starting to come [172:25] together for the right talent and DNA at [172:28] the company to do well in building out [172:30] the cloud so funny the other big piece [172:33] of it honestly maybe even the biggest [172:35] piece of it from talking to folks was [172:37] hot mail yes Microsoft had acquired hot [172:40] mail back in 1997 and ran it the whole [172:42] time it eventually became outlook.com I [172:44] think but you know it's running a [172:46] consumer web service for decades at [172:49] scale yep but oh my God it's so [172:54] unfortunate that they didn't buy Yahoo [172:57] just because of the Alibaba stake [172:59] nothing else would have mattered 40% of [173:01] Alibaba at IPO in 2014 when Alibaba IPO [173:05] that stake was worth $92 [173:07] billion well obviously Microsoft is not [173:09] a hedge fund but like right these things [173:11] are hard because how long would [173:12] Microsoft have held that yeah totally [173:15] but man if they did that'd be crazy it's [173:18] so funny yeah one of the greatest [173:21] Venture Investments of all time yep okay [173:24] so at the end of all this you might be [173:25] wondering why was search so important [173:28] how did Microsoft get so obsessed with [173:31] the search engine why are they still [173:32] running Bing today why has it been this [173:35] sort of white whale for them where they [173:37] continue to try over and over and over [173:38] again to do search deals or Acquisitions [173:40] or things like that well search [173:44] monetizes incredibly well so Microsoft [173:47] is sitting there realizing okay we're a [173:50] technology company historically what [173:53] we've done is sell licenses of our [173:55] software and people have paid us [173:58] directly but there's this new business [174:00] model emerging that appears to just [174:04] scale infinitely where you can make [174:07] three times or more off of each user [174:11] again using software but not selling the [174:13] software to them that's a much better [174:16] business if I could take 7 billion [174:18] people in the world and sell them [174:20] windows or I can take 7 billion people [174:22] in the world and have them use my search [174:24] engine for free and then make the money [174:26] from the advertisers I'm going to make 3 [174:28] to 5x more money from the advertisers [174:31] that I actually would selling them [174:33] software so suddenly this kind of [174:35] becomes existential where the windows [174:38] Revenue isn't going away but actually [174:40] the next generation of Economics [174:43] generated from software is not selling [174:46] the licenses it is monetizing via [174:49] advertising it's funny I never thought [174:51] about it this way but really what search [174:55] did and what Google does is you go from [174:57] selling software as a technology company [175:00] to selling [175:02] everything yep and the offline economy [175:04] is much bigger than the software economy [175:07] and so everyone has to acquire customers [175:10] whether you make software or tents or [175:13] airline tickets and there's only a small [175:15] set of dollars that goes to software so [175:17] if I just pull up my credit card [175:18] statement each month how much software [175:20] am I paying for versus how much [175:22] everything else am I paying for and even [175:24] if you say well that's not really fair [175:26] because it's advertising for everything [175:27] else it's not everything else directly [175:29] even a small percentage of my everything [175:31] else turns out to be way bigger than my [175:33] software budget so at the end of the day [175:36] Microsoft made a browser they didn't [175:38] monetize that browser they monetized [175:41] using it to defend an operating system [175:45] that they sold licenses to eventually [175:48] gole Google comes along and creates a [175:49] browser they also don't sell that [175:51] browser but they monetize all the [175:54] traffic coming through that browser and [175:56] they do it way better than Microsoft [175:59] does at monetizing selling [176:02] licenses maybe put it more simply [176:05] Microsoft built a browser had a bunch of [176:07] share and then kind of looked around and [176:08] said we don't really know what to do [176:11] with it I guess we'll use it for defense [176:15] and Google built a browser and said we [176:17] know exact what to do with this and they [176:19] used it for offense yeah such a good [176:22] point well while we're talking about [176:25] using software to sell everything and [176:27] not software that sure makes me think a [176:30] lot about social and [176:32] Facebook Microsoft has some history [176:35] there during this period too doesn't it [176:37] they do David should we take a brief [176:39] aside to talk about Microsoft's [176:41] intertwined history with Facebook [176:44] absolutely so it's October of 2007 [176:48] Microsoft is missing search and they're [176:53] realizing social seems to be a wave [176:56] that's coming five six years after [176:59] search also a great online advertising [177:02] business we now deeply understand and [177:05] regret not being a bigger player in that [177:08] business we can't let it happen again so [177:11] what do we do we're not going to build [177:13] one of these internally we know better [177:15] than to do Google+ and so we are not [177:19] Capital constrained and so we are very [177:21] willing to try to do large Acquisitions [177:24] because we've got money lying around but [177:26] we don't have talent lying around and we [177:28] don't have DNA and brand to be able to [177:31] do this right exactly so what do you do [177:33] you try to buy Facebook Microsoft puts [177:36] an offer on the table it's a very [177:38] complex deal structure but effectively [177:41] what it does is it lets Facebook [177:44] shareholders cash out over a long period [177:46] of time as the company's value grew so [177:49] you're not taking all your money off the [177:50] table today and so the important thing [177:54] to take away though is a very big dollar [177:56] valuation news outlets reported it to be [177:58] worth $24 billion and again this is way [178:01] back in 2007 3 years after the founding [178:05] of Facebook right we not that long after [178:08] the yahoo1 billion offer exactly [178:12] Facebook's not interested I'm pretty [178:15] sure Zuck doesn't even respond to the [178:18] offer some of Zuck lieutenants have been [178:20] meeting with Microsoft people saying if [178:22] you get the number in this range they [178:24] sent in an offer we'll have to ask Mark [178:26] about it in [178:28] Jas we will no dice so instead they work [178:31] out an investment and a commercial deal [178:33] so the terms of the deal are in October [178:35] 2007 Microsoft invests $240 million for [178:41] 1.6% of Facebook so for those trying to [178:44] do the math at home that is a $15 [178:46] billion value on the deal Microsoft will [178:51] get the exclusive right to sell banner [178:53] ads on Facebook internationally until [178:57] 2011 so again Microsoft cleverly is [179:00] using this as a way to bootstrap The [179:02] Advertiser side of their marketplace now [179:03] they have all this inventory to sell now [179:06] it's interesting to think about much [179:08] David like your comment about Alibaba if [179:11] Microsoft sold all of it at IPO which I [179:14] don't think it did that would be a 7x [179:17] okay that's a pretty good growth [179:18] investment in you know a few years not [179:20] bad right from 2007 to Facebook's IPO in [179:24] 2012 yeah 5e 7x that's good for growth [179:27] investment that's great I'll take that [179:29] if they held for another two years and [179:32] sold in 2014 that would Adit a [179:34] 14x I actually don't know when they sold [179:36] but I feel like these are helpful guard [179:38] rails to understand what this [179:40] appreciation could have been either way [179:43] it's not really relevant to Microsoft as [179:46] I said earlier they're not a hedge fund [179:47] in money is the least important thing to [179:49] them right they're constrained by Talent [179:53] execution ability Focus DNA to pull it [179:56] off focus but they're not constrained by [179:58] cash so who cares if you 10x your 240 [180:01] million over five to seven years so it [180:04] seems like the actual interesting part [180:06] of this deal is the fact that they had [180:08] the right to sell Facebook's [180:09] International ads for four years and the [180:12] companies kind of became friendly so [180:14] Facebook on the pages for businesses [180:18] would use Bing Maps and there was all [180:19] this sort of reciprocal things that the [180:22] company did together and a lot of [180:24] Microsoft people went to Facebook like [180:27] friend of the show VJ Raji coo at stat [180:29] Sig now a lot of great early Facebook [180:32] folks came from Microsoft yep so right [180:35] around this time June of 2008 Bill Gates [180:38] leaves the company full-time it is an [180:41] actual retirement I'm no longer Chief [180:43] software architect I am still chairman [180:45] of the board but I'm going to go be the [180:48] full-time at the foundation exactly at [180:49] the bill Melinda Gates Foundation which [180:51] of course is right at the time the [180:54] iPhone you know of course came out in [180:56] 2007 but 2008 was when I think it was [181:00] iOS 2.0 right with the uh SDK app store [181:04] opening up and SDK comes out and the [181:08] world completely transforms there's a [181:11] pretty rough quote in Time Magazine from [181:14] Bill's retirement that you know was [181:16] obviously written about bill here but I [181:18] think it's just kind of more applicable [181:20] to all of Microsoft DNA at this point in [181:24] time Gates is probably getting out of [181:27] Technology at the right time funnily [181:29] enough it's not really a business for [181:31] nerds anymore Gates was at the center of [181:34] the personal computer Revolution and the [181:36] internet Revolution but now the big [181:38] Innovations are about exactly the things [181:40] he's bad at the iPod was an aesthetic [181:43] Revolution Myspace was a social [181:45] Revolution YouTube was an entertainment [181:47] Revolution [181:48] this is not what Gates does techology [181:51] doesn't him anymore that's a stupid [181:54] quote that's just too [181:55] reductionist yeah it is totally too redu [181:58] and it's too personally about Bill and [182:00] that's just completely not right but [182:02] that was sort of the view at the time I [182:04] mean this really shows you how [182:06] irrelevant people thought Microsoft was [182:09] yes that is why we included it no one [182:12] would have been saying this about [182:13] Microsoft in the Windows 95 time frame [182:15] but after the obsession with Enterprise [182:18] the complete failure in consumer markets [182:21] but importantly the complete ignoring of [182:24] what the exciting developer platforms [182:26] were at the time open source the web I [182:31] mean if you think about where all the [182:32] development efforts were going it was [182:33] the lamp stack the uh Linux Apache MySQL [182:37] PHP the stuff Facebook was written on [182:39] that's in a different Universe from [182:41] Microsoft's Enterprise developer [182:43] customers so I just think you need [182:45] developer excitement if you're going to [182:47] have consumer excitement I mean or you [182:48] need to develop every interesting app on [182:50] your platform yourself but that's just [182:51] not how it goes so that consumer and [182:53] developer excitement goes hand in hand [182:55] right I said to put a pin in the iPhone [182:58] and the downside of Microsoft becoming [183:01] the Enterprise Juggernaut when we were [183:02] telling that [183:03] story and all of the huge advantages in [183:07] lock in that Microsoft had built up the [183:11] iPhone changed that calculus because the [183:13] iPhone kicked off Shadow it and bring [183:16] your own device and kicked off the user [183:18] revolt against it and this is what this [183:23] quote [183:24] encapsulates Yeah I think that's right I [183:26] think a set of Technologies were [183:29] breaking through people were just going [183:31] to use those devices and that software [183:33] no matter what the era today is one [183:36] where users have way more choice in what [183:39] they use at work than they did in that [183:41] early 2000s era and the iPhone sort of [183:43] forced that door open yeah choice and [183:46] expectations of what that soft Ware and [183:48] Hardware is going to be like and [183:51] Microsoft for all of its great victory [183:54] in the Enterprise over this period just [183:57] fundamentally did not have any of that [183:59] DNA in the company anymore and that's [184:02] what this quote from time is pointing [184:03] out they were all over an [184:05] Xbox okay fair enough or they were [184:07] developing cool new stuff that would [184:09] then get killed because it's not a part [184:10] of the Windows machine I mean you look [184:12] at Courier you look at kin in Mobile you [184:15] look at all these things that they would [184:16] let him get so far and then they'd be [184:18] like ah you guys don't get it [184:22] Windows is the center of everything and [184:25] if it doesn't make Windows look great or [184:28] it competes with Windows that's not what [184:30] we're doing and I think that DNA was too [184:34] strong to [184:35] overcome disruptive Innovations exactly [184:39] so let's talk about those things and [184:41] what's happening in Mobile okay so let's [184:45] rewind what was Microsoft doing in Mob [184:47] mobile so far a lot actually a lot [184:51] actually is Right Microsoft was obsessed [184:55] with all sorts of things in particularly [184:57] Bill Gates for decades before they [185:00] became true one of which was Bill Gates [185:02] was always talking about mobile [185:04] Computing so much so that in the key [185:07] slide in the Windows XP presentation one [185:10] of the big bullet points is mobile [185:14] Computing all the way back in 2001 and [185:18] Gates thought natural user interfaces [185:20] was going to be a thing multi-touch [185:23] tablet Computing pen Computing yeah I [185:26] had a tablet I had a Microsoft XP tablet [185:29] edition I think is what it was called PC [185:31] in college yeah from the early 2000s and [185:34] sometimes even before that these were [185:36] Bill Gates's visions of the future that [185:38] he thought were pretty close and so in [185:41] the world of I suppose they were early [185:45] smartphones Microsoft had developed [185:47] Windows mobile so what was this is like [185:52] iOS not really no what Windows mobile [185:55] was was an operating system for handset [185:58] makers to adopt and put on their [186:02] handsets and you know these things kind [186:04] of looked like blackberries or mostly [186:06] keys with a little screen and when you [186:09] looked at it it looked like Windows it [186:11] had a little start menu and it was much [186:15] like the rest of the Enterprise strategy [186:17] David [186:18] designed around all working seamlessly [186:21] together with your Windows PC and [186:23] exchange and your corporate Network [186:25] because surely people at home consumers [186:28] were not using smartphones these were [186:30] for business people who you know these [186:32] were issued by their Enterprise so it [186:34] fits pretty squarely into the Enterprise [186:37] category now how did Microsoft think [186:41] about this product they thought about it [186:44] as an ingredient into [186:48] the handset makers [186:50] product [186:51] Microsoft was somewhat at the whim of an [186:54] oem in the computer ecosystem you know [186:57] that Dell could install some more stuff [186:59] on top of Windows and customize the [187:01] installation but it was still Windows XP [187:04] no matter who the PC was from it was a [187:06] pretty standard thing that really wasn't [187:09] the case with Windows mobile phones the [187:12] handset makers could modify the code of [187:16] Windows mobile [187:18] so when you bought a handset first and [187:20] foremost you were trusting the product [187:23] quality of the people at the handset [187:25] maker and they had several ossos that [187:28] they could buy and effectively start [187:30] from one of which was Microsoft we [187:32] handset makers we make a [187:34] phone and we know how to interconnect [187:37] and do all the carrier stuff with the [187:39] carriers because they're our partners [187:41] and we kind of needed to do a bunch of [187:43] computer stuff too like email and stuff [187:45] so can you guys do all that and then [187:48] we'll make sure when we get that from [187:49] you that we'll start changing your code [187:51] to make it work with our phone and we'll [187:54] do all the phony stuff not an iPhone [187:57] what was Microsoft's position in Mobile [187:59] yes they had Windows mobile but no it [188:02] was nothing like what smartphones would [188:05] become because of the way that the [188:07] iPhone reset everything yeah and at some [188:11] point in this journey here post iPhone [188:14] but still in this sort of weird Windows [188:16] mobile era [188:18] Microsoft buys danger the company that [188:21] made the T-Mobile Sidekick you remember [188:23] that which was awesome you like pushed [188:25] button and then it would like flip [188:26] around and suddenly you were on a [188:27] sideways keyboard oh yeah it was in [188:29] entourage I remember watch an Entourage [188:32] Turtle had one I think that's right and [188:34] that was Andy Rubin before he started [188:35] Android oh you got me I was hoping I [188:37] could stump you I was going to say do [188:39] you know who of course co-founder of [188:41] danger was it was Andy Ruben what do you [188:42] think I do for a living what do you [188:44] think I do for a [188:46] living amazing Andy had already left [188:49] danger and started Android which would [188:51] be the very thing that would sort of [188:54] destroy their mobile business yes but [188:57] let's get there so 2007 in January the [189:01] iPhone is announced it won't come out [189:03] until July the iPhone comes out it's the [189:06] most spectacular technology demo since [189:09] the mother of all demos the old Doug [189:11] angle B one way back in the day [189:13] consumers are all in awe the existing [189:16] mobile industry people can't really [189:18] believe it's real the founder of [189:21] blackberry basically said I think his [189:23] exact quote is how did they do that then [189:25] later says we'll be fine you have Palm [189:29] who was already saying things like I [189:31] believe the CEO even before the [189:33] announcement said the PC guys are not [189:34] just going to figure this out they're [189:35] not just going to walk in famously David [189:38] I know you have it Steve Balmer has a [189:40] quote after the announcement yes he says [189:44] it's never going to work at $500 which [189:47] is the full quote and you could totally [189:50] see that phones at this point in time [189:53] Flagship phones were costing like $100 [189:56] with carrier subsidies and Steve's like [189:59] $500 that price like who's GNA buy that [190:02] right there is actually two quite [190:03] interesting things about this quote one [190:05] Steve is being the company salesperson [190:08] if a competitor drops this amazing bomb [190:10] and you're interviewed and you have a [190:11] whole bunch of Enterprise customers who [190:12] are sort of looking to you what do you [190:15] say you say are things still great [190:17] there's things really expensive of [190:19] course you say that you are literally [190:21] always selling all the time and so I [190:23] always take some issue with that two [190:25] apple legitimately had a business model [190:27] Innovation there with the carrier [190:29] subsidy yeah well it wasn't the original [190:31] the original iPhone didn't have it it [190:33] was then later I think the 3G I mean the [190:36] mobile industry to this point had been [190:38] how do I make the cheapest possible [190:40] phone it's certainly not a scaled down [190:42] version of a Mac which is what the [190:45] iPhone was so that was a completely [190:47] different Paradigm this is a tiny [190:48] computer not a kind of crappy embedded [190:52] system that is optimizing for pennies [190:54] and apple basically said we don't care [190:56] if it's really expensive we just think [190:58] this is the user experience bar and we [191:00] will figure the business model out and [191:02] eventually my god did they figure the [191:04] business model out and the carrier [191:05] subsidies were that Innovation but [191:09] Windows mobile was that old Paradigm [191:11] embedded systems cheap as possible [191:14] Hardware were a couple cents either way [191:17] determines whether your phone's going to [191:18] sell or not and so it was pretty [191:21] shocking so iPhones start selling [191:24] they're selling well it's 2008 apps [191:28] start coming out it's 2009 sales start [191:32] really picking up and finally Microsoft [191:36] decides hey what we're going to do is we [191:38] have this old asset Windows mobile we [191:41] can repurpose some of that to make this [191:44] new thing called Windows phone but [191:48] unfortunately everything we're [191:49] optimizing for is different the new [191:52] ecosystem expectation is a super high [191:55] quality user experience yeah and so [191:59] there's this way that we used to work [192:01] with all of our Hardware Partners which [192:02] basically said we will make the software [192:05] work on whatever you can come up with [192:07] the crappiest Hardware you can think of [192:09] and we'll make it work it's kind of like [192:10] the Roku strategy the way that they work [192:12] with all the embedded TV makers and the [192:16] new strategy to be we will dictate [192:19] really intense Hardware requirements [192:21] because now with Windows phone we are [192:26] making a promise to users to compete [192:28] with the iPhone where Microsoft is [192:30] backing that up the Microsoft brand is [192:33] first and you know we're defining a [192:34] really breakthrough new user interface [192:37] called Metro that actually came from the [192:40] zoom which is funny that that's its [192:42] lineage now how did it actually play out [192:45] Microsoft tried to use their existing [192:47] business model we will sell you an [192:49] operating system we will charge you a [192:51] royalty we will sell you OEM [192:54] manufacturer an operating system correct [192:57] and you sell that phone people want good [193:00] phones now so you can probably generate [193:02] some nice margins on that good phone [193:04] because the iPhone really set the bar [193:06] there's just one problem with trying to [193:08] maintain your old business model it's [193:11] that you don't have the same competitive [193:13] set that you used to you now have Google [193:16] Google has acquired Android Google has [193:19] transformed Android from a Blackberry [193:21] clone into an iPhone clone the software [193:26] is open source and so Google's value [193:28] proposition is they go to all those same [193:31] manufacturers that Microsoft used to [193:33] work with HTC Motorola and say hey how [193:38] about a deal for 0o yep deal Point [193:41] number one here you go it's free deal [193:45] Point number two you can even have the [193:47] source code deal Point number three we [193:50] aren't Microsoft look at what they did [193:53] to the PC makers do not let them do that [193:55] to you you know those PC makers they [193:57] make no money zero to profit dollars in [194:00] the value chain ACR to these PC makers [194:02] they all acre to the software vendor [194:03] it's literally the same people who did [194:05] that to the PC makers why would you let [194:07] them do that to you and remember in this [194:10] mobile world every scent matters and so [194:13] Microsoft is trying to ask for you know [194:15] whatever it is some single number of [194:17] dollars for a licensing fee to the EOS I [194:20] mean I think I'm unders shooting but [194:22] let's even say it's five bucks that is a [194:25] mountain of difference between $0 and $5 [194:28] in a low margin business in the total [194:30] bill of materials of these things [194:32] exactly and Google also only really [194:36] cares about their services that they [194:39] monetize through advertising so one of [194:41] the deal points in there I think this [194:43] may be varied by geography but is oh [194:46] yeah you got to use Google services on [194:48] there too but by the way they're Best in [194:50] Class and they're free you don't have to [194:52] pay anything for that either yeah I [194:53] think at first it was you can have it [194:56] open source but you don't get any of our [194:58] services or you can take the whole thing [195:01] and you take all of our services but our [195:03] services are great and guess what the [195:04] Play Store is one of our services so if [195:07] you want all the apps then you have to [195:08] take all the other Google services too [195:10] right right right now keep in mind how [195:13] does Google make money they make money [195:16] on search [195:18] so Google from the moment they figured [195:21] out hey we can run a call it 2002 when [195:25] Google's search business model was [195:28] really hardened and it was evident this [195:32] will scale it's ludicrously profitable [195:35] it's very high value per user Google's [195:37] going to be the number one at it it's [195:39] almost like if you really thought about [195:42] it you could have figured out that [195:45] Microsoft wouldn't win in Mobile H yeah [195:48] it's a really circuitous path but if [195:52] Step One is Google makes a ton of money [195:54] on search then step two is Google should [195:58] try to get all the searches so then step [196:01] three is Google needs to have the front [196:03] door to search and so you have to count [196:06] on Google being the actor that figures [196:09] this all out step four is Google figures [196:12] out what the next platform is and makes [196:14] sure that they are guaranteeing all the [196:17] search volume comes to Google from them [196:20] so what do they do they invent or buy a [196:22] mobile operating system what do they do [196:25] after that the next step they give it [196:26] away for free because again all they [196:28] care about is all the search volume and [196:30] so therefore unless Microsoft adopts [196:32] Google's business model they're [196:34] immediately screwed this is such a good [196:38] point Microsoft's [196:41] competitor was not apple and the iPhone [196:44] it was Android yeah [196:47] it's a little bit butterfly flaps its [196:49] wingsy but there is a direct line over a [196:52] 10year period from Google Finds Its [196:55] web-based search business model and [196:58] Microsoft cannot employ its traditional [197:01] business model and win in Mobile like [197:03] Microsoft will lose in Mobile and [197:06] there's some pivots in there I think the [197:08] biggest moment when the door really shut [197:10] is when Verizon freaked out after the [197:12] apple and AT&T deal and said we need an [197:16] answer and they decided that answer was [197:19] Droid and they put like a gajillion [197:21] dollars behind the Droid advertising [197:24] campaign the Moto Droid yep yeah and so [197:27] I think at that point it was sort of a [197:29] two-horse race Microsoft probably could [197:31] have figured out a way to get in before [197:33] that but it is all related to Google [197:35] finding that orthogonal business model [197:38] yep and it's F you know Microsoft did [197:40] have B at this point in time so they did [197:42] have a business model that they could [197:44] have used if they've been willing to go [197:46] free on Windows phone and it would have [197:48] taken a big culture shift at Microsoft [197:50] to say we're an advertising company [197:51] right Microsoft is not a company that is [197:54] at least at this point in time [197:56] comfortable with their bread being [197:58] buttered from advertising I mean they're [198:00] the PC company they want to sell [198:01] software to people using PCS they're the [198:03] software company yeah they sell software [198:05] might be via Enterprise agreements but [198:07] they sell software yep so then if you [198:10] really believe this step by step by step [198:13] thing then actually what Google should [198:15] keep doing is is finding things that [198:19] Microsoft [198:20] sells and figure out which ones are the [198:23] cheapest per user to run and then give [198:26] those away for free and so Outlook [198:30] Exchange gez Gmail hm word excel [198:34] PowerPoint oh G Suite [198:37] workspace and all they're doing is [198:38] they're just looking at Microsoft's core [198:40] value propositions they charge money for [198:42] and Google says would it really be that [198:44] expensive if we just gave that away for [198:46] free [198:47] and the more of those that they do a [198:50] it's good for Google's business model [198:52] because they just get more data a closer [198:54] relationship with you you're doing [198:56] either more queries or you're [198:57] interacting on their platforms in ways [199:00] where they have other ways to show you [199:01] ads maybe while you're off platform now [199:03] they know a lot about you from data [199:05] they've collected blah blah blah but [199:07] even if it doesn't actually make more [199:09] money for Google they make so much money [199:12] in their Core Business that if it [199:13] hampers Microsoft then it's a good thing [199:16] to do yeah interesting and that is [199:19] totally true I totally buy it in the [199:21] consumer world and in the Enterprise [199:23] World Microsoft lockin is still as [199:26] strong as it ever has been yes and [199:28] Google has really not figured out how to [199:32] be an Enterprise company Excel is still [199:35] the main way that spreadsheets are done [199:37] around the world I bet a lot of [199:38] listeners use Google Sheets we do too we [199:41] love it also use Excel but if you do use [199:43] Google Sheets you are in the minority [199:45] globally you think yeah I guess because [199:47] most consumers don't actually use [199:49] spreadsheets yeah yeah Enterprise [199:51] spreadsheet work is done in Excel full [199:53] stop yep that brings us to Nokia but I [199:59] think let's save Nokia for the end here [200:01] yeah Nokia is our Koda what it does [200:03] bring us to is a realization from the [200:06] very top of Microsoft that the profit [200:09] pools in Mobile are changing and this is [200:11] a thing that I think Steve Balmer also [200:13] doesn't get credit for Bill Gates was [200:15] obsessed correctly [200:17] with being the software company it was a [200:21] brilliant business strategy to be the [200:23] software platform and then everything [200:26] around you had to interoperate with you [200:28] and again the profit pools in the PC [200:31] World just acred to software vendors it [200:34] was remarkable how the PC manufacturers [200:37] over time had no profits and Microsoft [200:38] had tremendous profits Steve Balmer [200:41] realized pretty early I think because of [200:44] the Google Android thing mobile was [200:46] going to shake out differently future [200:48] Hardware platforms were not guaranteed [200:51] to have the same profits in the value [200:53] chain the way that the PC did and so he [200:56] was pretty aggressive about actually we [200:58] need to be in the hardware business and [201:00] I know that seems really unattractive to [201:03] us as a company because we've been in [201:05] the software business there's this great [201:07] general rule that it's really hard for [201:08] any business to enter a lower margin [201:11] business than the one they are currently [201:13] in yeah Amazon can go from e-commerce to [201:15] AWS right Amazon can go anywhere but for [201:19] Microsoft you know you're selling [201:20] software licenses it's hard to even get [201:22] into Cloud because cloud is a lower [201:24] margin business you know you have to [201:25] operate those data centers than just [201:27] selling the licenses so if you're [201:30] Microsoft and you've been making [201:31] software all these years and you've been [201:33] enjoying those margins and suddenly [201:36] you're realizing uhoh we have to be in [201:38] the hardware business or at least if [201:41] we're not in the hardware business or [201:42] the search business we're not going to [201:44] enjoy any of the profits in the mobile [201:47] era that's a difficult conundrum yeah [201:51] but to Steve's credit he acted they [201:53] released surface they tried to buy Nokia [201:56] yep let's talk about let's start with [201:58] surface and Windows 8 which we got to [202:00] talk briefly about Windows 7 before that [202:03] because Windows 7 was awesome so Steven [202:05] sinowski he ends up running office [202:08] product management after the Vista [202:10] disaster he gets drafted to come over [202:13] and run Windows and Ben like you're [202:16] saying the office culture was known as [202:17] we ship we ship product that's what you [202:19] were part of that was the culture that [202:21] Steven said mindbending that 3 years in [202:24] advance a date is set and then 6,000 [202:26] people ship on that date no matter what [202:29] y so he comes in for Windows 7 and he [202:33] does that for Windows we were talking to [202:35] him and he had this great analogy of [202:38] windows at this point in time didn't [202:40] need technical Vision it was trying to [202:42] be the Dodge Viper that's what Longhorn [202:45] was it needed to be the Toyota Camry and [202:48] he comes in and he makes Windows 7 the [202:51] Toyota Camry of PC operating system I [202:53] love that analogy that's exactly right [202:55] it's so good it's exactly what it was [202:57] and it's exactly what everyone wanted [203:00] it's what the consumers who were still [203:02] using Windows wanted they wanted it to [203:03] just work and most importantly it's what [203:05] the Enterprises wanted everyone's like [203:07] hey it's like XP but modern or it's like [203:09] Vista without all that random stuff and [203:11] all the regressions that Vista had yes [203:15] Do Not Crash my device or my network [203:18] thank you nice easy start button in the [203:20] lower corner normal predictable menu [203:23] fast search Fast file system I honestly [203:27] can't tell you a feature that launched [203:29] in Windows 7 I have no idea but I [203:32] remember I had a Windows 7 laptop when I [203:34] first joined madona and it was great yep [203:38] it ran everything the way you expected [203:39] it to and so the product that sinowski [203:42] shipped there was just as much the new [203:45] organization as it was the actual [203:48] product that customers experienced it [203:50] was a much more slim down team it was [203:53] Dev test PM it was ability to hit a ship [203:55] date it was a proper planning and vision [203:58] process a lot of what the team was doing [204:00] with Windows 7 was yeah yeah we feel [204:03] like we can do seven with our arms TI [204:04] behind our back but let's start thinking [204:07] about the future about what we're really [204:09] going to do now that we have all the [204:10] infrastructure in place to really ship a [204:12] interesting product and [204:14] then Apple out with the iPad in 2010 [204:19] yeah so Windows 8 Vision had kicked off [204:22] the planning process for what it's going [204:24] to be had kicked off and they're [204:26] starting to play actually Stephen puts [204:28] these videos on YouTube they're awesome [204:30] to watch the original vision of what [204:31] Windows 8 should be and they're kind of [204:34] out on a limb they're saying the future [204:36] is touch the future is tablets we think [204:40] that's going to be a dominant Computing [204:42] Paradigm and on the one hand Bill Gates [204:44] has been saying this for years [204:46] so there's sort of like a cultural [204:49] acceptance with the idea on the other [204:52] hand it really hasn't manifested in the [204:54] market so it's a little bit dangerous to [204:56] say but the iPhone has now come out and [204:58] multi-touch has shown this is the way to [205:01] do it so but you're right the iPad [205:03] coming out really validates whoa big [205:06] tablets with [205:08] multi-touch actually might be the [205:11] Computing Paradigm for the future and oh [205:13] my God we've been in planning we've been [205:15] in development for you know a year or [205:17] two already this thing hits the market [205:19] we're right we are so right we've been [205:22] validated it validates the product [205:24] vision and it [205:26] terrifies bomber and Microsoft [205:30] leadership because they just watched [205:32] what happened what Apple did to the [205:33] phone market the iPad sure as hell looks [205:37] like it's going to try and come do that [205:39] too Microsoft's core PC market I mean [205:43] right the original jobs keynote [205:45] introducing the iPad lays out his vision [205:48] right of like the PC is going to become [205:50] the pickup truck and the iPad is going [205:52] to become the car and that would be a [205:55] truly terrible thing for Microsoft [205:57] especially if that goes into the [205:59] Enterprise as the iPhone is clearly [206:01] going into the Enterprise on the phone [206:03] side now there was a little bit of a [206:05] Folly in believing that the iPad was the [206:08] PC of the future standing here today we [206:10] all can look at unit sales and realize [206:13] oh the iPad was not the PC of the future [206:15] it had its place but it did not in any [206:18] way replace PCS and it turns out that [206:22] Apple scaling up the iPhone [206:24] metaphor was good for tablets but that [206:27] never was going to take over most PC use [206:30] cases today in fact the phone has far [206:32] more replaced the PC than the tablet has [206:35] 100% yes but back in 2010 that sure [206:38] looked pretty terrifying as a prospect [206:41] to Microsoft the other thing that's [206:43] happening around this time is I've said [206:46] this a number of times but Windows [206:48] despite having great Revenue great [206:50] profits massive penetration in the [206:52] Enterprise and momentum almost just like [206:56] staying power in consumers because [206:57] people were just used to it it was not [207:01] relevant for the next Frontier totally [207:03] not it did not have hearts and Minds it [207:05] was not where the excitement was it was [207:07] not what people were building for so [207:10] there's sort of a two birds with one [207:12] stone attempt with Windows 8 One Touch [207:16] tablets we are going to get out ahead of [207:20] Apple I mean we're going to try to out [207:22] apple apple here and we're not going to [207:23] let what happened in phone happened to [207:26] us in our core Market of [207:27] PCS two we need a new developer platform [207:32] yeah we need to bring developers back [207:35] everyone's building for the web web is [207:38] agnostic to what operating system it [207:40] runs on can we create a platform that is [207:43] so exciting for Developers [207:46] that they're going to use it and we [207:48] should lean into the Technologies people [207:50] are already using so the Windows 8 Touch [207:54] mode Metro UI development environment [207:57] was HTML 5 because all these web [208:00] developers are already writing their web [208:03] apps we want to support that too and [208:05] we're going to build a whole new tool [208:07] chain so that their HTML 5 Windows 8 [208:11] apps run really well on arm processors [208:15] because these tablets are going going to [208:16] run on arm processors yep and we're [208:18] going to make our own the Surface RT yes [208:21] so that's the two-headed dragon of [208:23] Windows 8 is new developer platform and [208:26] touch first and the way the touch first [208:29] manifests in the operating system itself [208:32] is the desktop is now just an app and [208:36] when you boot up Windows 8 you are [208:38] presented with a tablet tiled start [208:42] screen and if you're looking for a [208:44] desktop oh you got to go find it now in [208:48] practice it's not hard to find the [208:50] desktop you learn it in like five [208:51] seconds you're like okay I see the start [208:53] screen is actually the start menu but [208:55] full screen so if I click in the bottom [208:57] left corner I can collapse it I can [208:59] enter the desktop mode and then it's [209:00] like it doesn't even exist I can run my [209:02] 132 apps blah blah blah but there is a [209:04] learning curve there's also just the [209:06] shock value though of I bought a Toyota [209:09] Camry expecting it to be a Toyota Camry [209:12] and I don't even know what this is It's [209:14] like a scooter yeah and it is a little [209:16] confusing I used it for a long time when [209:17] I worked at Microsoft and figuring out [209:19] how to app switch between things that [209:21] are part of the Metro modern UI versus [209:24] the Legacy apps and what's sitting on my [209:26] desktop and what's sitting in the tablet [209:29] optimized app switcher it mixed two [209:31] metaphors now the question is why did it [209:33] mix two metaphors and it took me a while [209:35] to figure this out but what ended up [209:38] happening was the original vision for [209:41] the Windows 8 Touch thing that we're all [209:44] talking about these live tiles [209:46] That was supposed to only ship for [209:48] tablets as it was originally dreamed up [209:51] and there was a version of Windows 8 [209:53] that did not have that that was going to [209:55] ship for desktop PCS that was going to [209:58] look like Windows 7 probably word comes [210:00] down from on high Windows is Windows we [210:04] need to ship Windows across all devices [210:08] so what happens all this effort has gone [210:10] into and momentum and political capital [210:13] and betting your career has gone into to [210:16] this HTML 5 developer Community the [210:21] Metro [210:22] UI and so that is the desktop version [210:27] that ships I see yeah there can only be [210:29] one windows and so we got to put both of [210:33] these babies in here yep and what you [210:36] have is not as bad as Vista but man the [210:41] roll out was pretty bungled it's [210:43] confusing the reception was poor [210:47] interestingly not by the tech pundits [210:49] like the tech pundits who actually spent [210:51] some time and figured it out were [210:52] trained up pretty quickly but the cat [210:54] was out of the bag even before they got [210:56] to review it on people who were angry [210:59] what do the Microsoft people refer to [211:01] him as people that uh oh the basement [211:03] the basement yes the [211:06] 0.001% power users who are the loudest [211:08] of course on the internet and so that [211:11] kind of taints the product oems hate it [211:14] because frankly OEM [211:16] weren't signed up to make these touch [211:18] devices but now Microsoft's putting all [211:20] this energy behind touch optimized [211:23] operating system so there's this mixed [211:24] message to Consumers it's like are there [211:26] even good laptops available am I [211:29] supposed to use touch on my desktop they [211:31] have to run arm they can't run x86 they [211:35] got to run arm processors they got to [211:37] run mobile processors but you're asking [211:40] these devices in 2012 to also be able to [211:45] function as laptops and the technology [211:47] just wasn't there there's a reason why [211:49] the iPad was a scaled up version of the [211:51] iPhone not a scaled down version of the [211:54] Mac today I think it might be a very [211:56] different proposition and the public [211:58] might be much more ready to accept [212:00] something like this I wish Apple would [212:03] do this with the iPad oh yeah I don't [212:05] want to have a Macbook and an iPad I [212:06] just want an awesome paint of glass that [212:08] can do everything and apple silicon [212:10] totally can do everything yeah my dream [212:12] machine is you know those Lenovo yogas [212:16] that can flip all the way around yes [212:18] yeah yeah yeah my complete dream machine [212:20] is my 13in M3 MacBook Air that when I [212:25] flip it all the way around they do [212:27] something like universal binary with the [212:29] apps where all the same apps that I had [212:31] installed on my Mac they now run their [212:33] iOS counterpart they grab all the data [212:36] that's stored in the same places so all [212:37] my apps you know it knows which Google [212:39] sheet I'm looking for it has the YouTube [212:42] videos cached it you know does whatever [212:44] but it just turns into an iPad with an [212:46] iOS UI that is the dream I can't figure [212:50] out if I'm like a super nerd for wanting [212:51] that and most people wouldn't actually [212:52] want that but I travel with an iPhone [212:55] and an iPad and a Macbook and I think I [212:58] could just do two I think a lot of [213:00] people would want it yeah so one [213:02] takeaway may just be hey it was too [213:04] early the other takeaway might be look [213:07] it turns out that tablets should have [213:09] been a scaled up phone not a scaled down [213:11] PC that was certainly true at the time y [213:15] certainly at the time time so complete [213:17] commercial failure the ecosystem of [213:20] Windows 8 apps did not really Galvanize [213:22] and where we're left is the state of the [213:25] Windows app developer ecosystem in the [213:29] 20134 time period is right back where we [213:33] started no one's terribly interested in [213:35] targeting that as a developer platform [213:38] all the energy is actually just going to [213:39] go into the web app yeah and then all [213:41] the energy is going to go into the [213:42] mobile app yep none of which is in the [213:45] Microsoft [213:46] ecosystem okay yeah I mean this is the [213:49] death of Microsoft as a consumer company [213:52] no doubt undeniably Zoom failed Bing [213:55] small market share Windows phone lost to [213:59] Android yep lost Mobile Windows 8 you [214:03] know this whole thing didn't work the [214:06] stock price has languished hasn't moved [214:08] in 10 years stock stuck at 30 bucks this [214:11] is dark this 2012 13 time period this is [214:15] dark dark and at the same time well it's [214:18] funny it's dark and revenues and profits [214:21] have grown tremendously the Enterprise [214:24] motion of Microsoft has basically never [214:27] had a down year I mean 2008 in the Great [214:30] Recession but other than that like chug [214:33] chug chug chug chug even through the bad [214:35] Windows releases it's dark and yet the [214:38] light is shining so bright on the [214:40] financial statements of this [214:42] company and what is going on here [214:44] obviously it's the ENT price but it's [214:46] even more than that it's Azure it's the [214:48] cloud it's already humming it's going [214:51] Microsoft did reinvent itself Microsoft [214:55] did position itself to be at the [214:58] Forefront of Technology it just did it [215:00] all within the Enterprise context of the [215:03] company and the Azure story is [215:05] absolutely incredible and I think nobody [215:07] knows how it really happened yes as the [215:10] general public is concerned Steve Balmer [215:13] was obsessed with Windows he built the [215:15] prise business he left satanella came in [215:18] and launched Azure and Azure has been [215:20] great like not exactly that's not really [215:24] what happened so when Bill was planning [215:27] to fully retire from the company to [215:30] retire from his chief software architect [215:32] role this is all the way back in 2004 [215:36] 2005 he and Steve know that there needs [215:40] to be a successor in this role even [215:42] Steve would be the first person to tell [215:43] you he is not a technology he can't do [215:46] both roles he needs a bill he needs a [215:49] chief software architect and [215:52] realistically you can't replace Bill [215:54] Gates with one person so we need two [215:56] bills yeah so they're casting about [215:59] Craig Mundy becomes one of those two [216:00] bills internally and they also know who [216:04] probably the perfect person is to take [216:06] the other job and that is Ray Azie Ry of [216:10] course being the author of Lotus Notes [216:14] Ray is Alle legendary developer and he [216:18] has great relationships within Microsoft [216:21] because Ray built Lotus Notes not at [216:23] Lotus but at his own software Studio [216:26] startup and Lotus was just his publisher [216:28] so he's known all the Microsoft guys for [216:30] years this is so fascinating I never put [216:33] two and two together but Lotus 123 and [216:36] Lotus Notes were not like peers together [216:39] Lotus 123 was developed by lotus Lotus [216:42] Notes is actually raise company it's [216:45] almost like the the way a game Studio [216:46] Works Ry and his company are building it [216:49] their publisher is lotus but Ry can have [216:54] agreements with Microsoft where he's [216:55] privy to information that Lotus is not [216:58] and so Ray is like really close in the [217:00] fold with the Microsoft folks I think he [217:03] was even a contractor working on maybe [217:06] the project was landman but he was [217:08] actively contributing to other Microsoft [217:11] products right cuz he had his own [217:13] software company yeah he was almost like [217:15] Swit in the middle yeah so by this era [217:18] now 2004 2005 Ry has a new startup [217:21] called Groove networks and Microsoft [217:24] just acquires the company and they get [217:27] ray so in June 2006 when Bill announces [217:32] his coming retirement Ray gets named as [217:36] his successor in the official Chief [217:38] software architect role and essentially [217:40] what's going on is Bill and Steve kind [217:42] of look at Ray and they say you figure [217:45] out the vision we've got all these [217:47] assets we've got a Killer Business got [217:49] all this great talent in the coming [217:51] world the next generation of Technology [217:55] why don't you figure out how Microsoft [217:57] fits in and what our play is so Ry [218:01] writes the internet services disruption [218:03] memo in October 2005 and to quote from [218:07] it this is Ry writing the environment [218:10] has changed yet again this time around [218:13] Services Computing and communication [218:15] technologies have dramatically and [218:17] progressively improved to enable the [218:19] viability of a services-based model the [218:22] ubiquity of broadband and wireless [218:24] networking has changed how people [218:26] interact and they're increasingly drawn [218:28] toward the Simplicity of services and [218:30] service enabled software that just works [218:34] businesses are increasingly considering [218:36] what services-based economics of scale [218:40] might do to help them reduce [218:41] infrastructure costs or deploy Solutions [218:44] as needed and and on a subscription [218:47] basis wait David you're telling me that [218:50] businesses may want to basically rent [218:53] capacity from Big Data Centers to just [218:55] deploy their applications and not worry [218:57] about the capex of buying the servers [218:59] and racking them all and maintaining the [219:01] data center and handling the privacy and [219:04] blahy blahy blah I'm telling you that [219:06] and I'm telling you they might not even [219:10] just want to buy the infrastructure they [219:13] might just want to buy the solution as a [219:15] Serv [219:15] hosted by us all right so this is Ray in [219:19] 2005 and so in January 2006 Ray with [219:23] Steve bomber's full blessing goes and [219:26] starts recruiting for a secret project [219:28] within Microsoft incubated outside any [219:31] of the existing divisions and this is [219:34] super important this should have come [219:36] within server and tools like that whole [219:38] big new business that we talked about [219:40] that was like the key Lynch pin of [219:43] Microsoft's Steve bomber era prise [219:46] strategy of course azer should have come [219:48] from within there Microsoft has a group [219:51] that produces a product called Windows [219:53] server and that is an operating system [219:55] that runs on other people's servers and [219:58] that group is not the group that [220:00] produced Azure the cloud service that [220:02] runs at the time Windows Server yes the [220:06] reason for that is that this is [220:07] completely disruptive to the whole [220:10] Windows server and server and tools [220:12] business model their go to market and [220:14] their business business model is we sell [220:17] these solutions to be operated in your [220:20] data centers in your infrastructure [220:23] where Accenture and all the consulting [220:26] firms and all the value added resellers [220:28] they're all our partners they're all our [220:30] go to market they're all going to go [220:32] Implement that on Prem for you and so if [220:35] we were now to say like wait a minute [220:37] all of a sudden we're going to do that [220:38] as a service and we're going to sell it [220:39] to you [220:41] separately that is a huge issue risking [220:44] a lot of my go toet Market motion not to [220:46] mention these end Enterprises in some [220:49] ways are actually the oem's customers [220:53] yeah these Dell servers are running [220:55] Windows server but Dell sold a bunch of [220:57] servers probably through Accenture to [221:01] the End customer there's that whole [221:04] issue of upsetting the apple cart [221:06] there's also the internal rewards issue [221:09] and kpi issue everyone in Windows and [221:13] Enterprise land ultim their kpi is how [221:16] many copies of windows can we sell to [221:18] end customers and generate the licensing [221:21] revenue on Windows and this new thing if [221:23] we actually pursue a cloud strategy is [221:25] how can we spend a whole ton of money [221:28] building out a data center buying other [221:30] people's servers generating zero [221:33] licensing dollars and hoping people use [221:36] the servers so we can charge them later [221:40] yeah Windows is nowhere in this equation [221:43] yeah we're going to build out a gigantic [221:44] server farm and rent usage to people [221:48] that doesn't fit into anyone's current [221:50] kpi or compensation so Ray is recruiting [221:53] for this project code named Red Dog and [221:56] he brings in the biggest of big guns [222:01] that's right the legend himself Dave [222:04] Cutler no one builds [222:07] hardcore Enterprise ready close to the [222:11] metal code than Dave Dave was the [222:14] architect on Windows NT we talked about [222:16] him a lot on part one also another guy [222:19] named amitab Shava amitab had also come [222:22] from deck which is where Dave came from [222:24] Total Beast of an engineer he had [222:26] experience both in the Enterprise server [222:28] and tools products and amav was also a [222:31] big part of getting Vista out the door [222:32] with Brian Valentine before Brian left [222:35] for Amazon and the two of them recruit a [222:38] team and they build Azure Cutler builds [222:43] a new hypervisor [222:45] that Azure runs on from scratch without [222:48] using open source like [222:51] himself hypervisor of course is the [222:54] piece of software that virtualizes [222:56] underlying hardware and allows you know [222:58] multiple software tenants to run on a [223:00] single piece of Hardware it's like [223:02] VMware was a hypervisor company it was a [223:04] whole company building hypervisors and [223:06] Dave just like yeah I got this it's [223:09] crazy so great and Steve bomber [223:13] supported this whole thing pushed it all [223:15] through despite heavy pressure and [223:17] incentives from inside the company from [223:20] Windows from Partners from the whole go [223:23] to market motion that he built you know [223:25] Microsoft's Enterprise go to market [223:27] motion Steve he didn't get it right away [223:31] but he started talking to enough [223:33] customers and realizing that this was [223:35] the future of Enterprise Computing that [223:37] he just flipped a switch and said I'm [223:39] all in we're doing this whatever [223:42] resources we need I mean we're talking [223:44] billions and billi ions of dollars of [223:45] capital expenditure to build up these [223:47] data centers this is not just a like oh [223:50] some little incubation project you know [223:52] sure we'll see what happens this is like [223:54] no we kind of got to like bet the [223:56] company on this well it's funny I [224:00] disagree that it's a bet the company [224:02] move because of two reasons one it's [224:06] only money that they're spending cash is [224:09] never a resource constraint the bigger [224:11] concern is Ray Ai and Dave Cutler are [224:13] working on this that is why it would be [224:15] bet the company two in its initial [224:19] Incarnation Azure did not [224:22] threaten the windows Centric approach if [224:26] you remember when Azure launched it was [224:28] Windows Azure and it ran Windows server [224:32] and it was platform as a service and [224:34] Microsoft in no way changed its tune on [224:37] open source I mean to this point in [224:39] history Microsoft thought that open [224:41] source was a complete cancer right and [224:44] for for good reason I mean at the end of [224:46] the day basically Microsoft charged for [224:48] things that open source was giving away [224:50] for free from operating systems to [224:52] programming languages to development [224:54] environments to servers I mean [224:56] everything about it it was like oh my [224:58] God is there a future where everyone [225:00] just expects all of our value to be free [225:03] and they managed to combat that and [225:05] build a great business despite that but [225:07] they never embraced open source they [225:10] never at all wanted to be a part of [225:13] anything that open source Developers [225:15] were doing until couple years into Azure [225:17] not until [225:19] 20145 and why I would say this is bet [225:22] the company you're right they didn't go [225:24] full infrastructure as a service and [225:26] embrace open source and let people use [225:28] Azure to run Linux in the lamp stack on [225:31] top of it that was not day one but they [225:34] knew they had to and they were going to [225:36] and it was just uh hey we're not going [225:39] to do this right away because the [225:41] company would organ reject this so hard [225:44] but we are moving in that direction and [225:47] we will be AWS we will offer everything [225:52] they offer and more to our Enterprises [225:55] who trust Microsoft yep that's a fair [225:58] push back but Azure came in with a [226:01] aggressive point of view we are platform [226:03] as a service which was distinctly [226:05] different than AWS which was we are [226:07] infrastructure as a service now [226:09] interestingly office and applications [226:11] and software as a service actually came [226:13] pretty quickly thereafter too famously [226:15] they did a pilot program with Energizer [226:18] like the battery company selling just [226:21] sort of I don't think it was like the [226:23] office productivity site but it was like [226:24] SharePoint and stuff I think as a [226:27] service as software as a service right [226:30] so in 2010 Ray Azie actually leaves the [226:35] company but as he's doing he and Steve [226:38] Roll Red Dog by this point in time [226:40] renamed Azure back into the server and [226:43] tools business and two of them go to the [226:46] University of Washington and Steve gives [226:49] a speech at the University of Washington [226:51] we are all in and we are betting the [226:53] company on cloud and on Azure the [226:55] intended audience of course was [226:57] Microsoft internally of like hey we are [226:59] sending a message to the server and [227:01] tools team this is the future and after [227:05] that Steve [227:07] replaces the division head of the whole [227:10] server and tools division who was Bob [227:12] muglia at the time Bob would later go on [227:15] to be the CEO of snowflake before Frank [227:17] slutman came in so he did fine Bob was [227:21] great Bob was crushing it as head of [227:24] server and tools Revenue was growing I [227:26] don't know 30 40% a year it's a 12 [227:29] billion business but the reason that [227:32] Steve made the change was he said we [227:35] need a new leader who's going to come in [227:36] and change this organization and make it [227:39] a cloud first organization and not carry [227:42] the baggage of all the success from the [227:44] the previous iteration and the person [227:47] that Steve Taps to do that from Bing is [227:53] none other than SAA Nadella yep to come [227:56] in and Lead that transformation and from [227:58] what I can tell it's just as motivated [228:01] by Azure is the future and it needs a [228:03] new leader as it is SAA is a really [228:06] talented Rising executive in this [228:08] company and needs to be put on an [228:10] important project absolutely almost like [228:13] Bing's not enough for this Guy where can [228:15] we put him totally it was let's get this [228:19] guy the right exposure to the right [228:22] important things that he could be CEO of [228:25] this company someday in the not very [228:27] distant future yep to say that this goes [228:30] well is an understatement obviously but [228:33] just to put some numbers on this [228:35] Microsoft has three reporting segments [228:39] productivity and business process AKA [228:41] office and then includes Office 365 as [228:44] part of that segment the more personal [228:46] Computing segment that's windows and [228:49] surface and their Hardware efforts and [228:51] then the intelligent Cloud segment and [228:53] that's Azure and I think LinkedIn is [228:56] part of the office segment if I have [228:58] that right I think that's right I don't [228:59] think LinkedIn is in the cloud segment [229:02] intelligent Cloud today is by far the [229:07] largest segment in the company by both [229:11] revenue and profit and by very very far [229:15] the fastest growing within the company [229:18] Windows is declining it's the largest [229:20] business now and the fastest growing [229:22] largest business most profitable fastest [229:24] growing in fiscal 2023 intelligent Cloud [229:28] did $88 billion in Revenue wow crazy it [229:33] is worth not to Pork cold water at all [229:35] because I think the high level point [229:37] stands I was gonna say this too I know [229:38] where you're going intelligent Cloud [229:40] includes SQL server and Windows Server [229:44] so these are big Legacy businesses yes I [229:47] think that is both a especially in the [229:49] early days when Microsoft and Satia was [229:52] hyping up how much Cloud Revenue the [229:54] company was doing being able to report [229:56] the Legacy server business as part of [229:59] that Revenue helped a lot on the other [230:01] hand the counterargument to that is this [230:04] is actually Microsoft's competitive [230:06] Advantage versus AWS totally agree [230:08] Microsoft can go to Enterprises and say [230:11] we are hybrid Cloud less so today but in [230:13] the earlier days of [230:15] the Azure transition saying like hey you [230:17] need to be on cloud we have a world [230:19] class public Cloud for you and it works [230:22] great with our on-prem server offerings [230:25] and we can be hybrid for you yep totally [230:28] agree we'll talk about this a little [230:29] more in conclusion we have one more [230:31] chapter in Nokia and the end of Steve's [230:33] tenure to talk about here but it turned [230:36] out [230:38] actually that the cloud Market was so [230:42] big that nothing else really really [230:45] mattered all the missteps all the losses [230:48] it makes sense right Cloud Powers [230:51] everything Cloud Powers Tech Cloud [230:54] Powers all the consumer services they [230:56] all run on the cloud so every consumer [230:59] service that is not owned by Microsoft [231:02] or meta or Google or Amazon runs on one [231:06] of their clouds yeah and some portion of [231:08] that Revenue acrise to Microsoft and [231:11] increasingly the offline economy [231:15] is becoming some sort of cloud dependent [231:18] service I mean it's crazy to just see [231:21] cars rely on the cloud and restaurants [231:23] rely on the cloud like anything that you [231:25] interface with in the physical world you [231:27] expect to have some digital component at [231:29] the very least take credit cards and all [231:32] of these things Point of Sales Systems [231:34] all of these things are routed through [231:37] the cloud at some point and so David I [231:39] think you're making the same point about [231:41] the cloud today that I was about [231:43] Microsoft in the PC era Microsoft was [231:46] lucky to own 90% market share and in [231:48] Cloud they own you know meaningfully [231:50] less than that right but it's still [231:52] basically a tracker on the growth of an [231:55] insane secular Tailwind that is just an [231:58] inevitability in the world it's probably [232:01] a 30 40 year wave that they get to keep [232:04] riding and again this is outside the [232:07] scope of this episode but it's sure [232:09] looking like to the extent AI is the [232:12] next Computing wave that is also [232:14] happening in the cloud in the data [232:15] center so like that's just going to [232:17] turbocharge everything y so to review [232:21] how it came to be interestingly it was [232:23] Ray azi in an incubation group doing it [232:26] outside the bounds of the business units [232:29] beginning in 2006 beginning in [232:32] 2006 with Steve's buyin and the air [232:35] cover from Steve to make it happen [232:38] organizationally you look at where all [232:41] the talent came from Bing taught them [232:43] how to do distributed systems Xbox Live [232:46] was a always on cloud service Real Time [232:50] Zero latency with 40 million subscribers [232:53] MSN was a super high-traffic web [232:55] property with 750 million registered [232:58] users Hotmail was a web application that [233:00] hundreds of millions relied on they had [233:02] SharePoint and exchange there was [233:04] knowledge of how to do server-based [233:05] application software for the [233:08] Enterprise I mean there was some [233:11] conflict business model wise with [233:12] Windows Server since Azure would be [233:14] orthogonal business model but the [233:15] technical chops were there I mean these [233:17] are hardcore server OS people of course [233:22] that group is going to be capable of [233:23] doing things like hypervisors so I just [233:26] think the ingredients were remarkably [233:30] there from all these other things that [233:33] Microsoft had been doing over the years [233:35] they were kind of the only one who could [233:38] pull this off at this scale with this [233:40] set of Enterprise relationships to [233:43] migrate all these people to the CL Cloud [233:44] as they built out the product Suite I [233:46] mean really kind of like we got I don't [233:49] know halfway is through our research for [233:51] this and this just hit me of holy crap [233:54] this era for Microsoft that everybody [233:57] thinks of is like the loser [233:59] ERA this is the era where they won you [234:02] know or they built the foundation to win [234:05] yeah there was seven years before Steve [234:10] Balmer handed the Reigns to SAA where [234:13] azure development was happening under [234:16] him yes that is nowhere near the public [234:20] narrative and Steve is the one who [234:22] handpicked SAA to lead it and get all [234:26] the credit and the narrative and the win [234:28] and then become the CEO pretty wild that [234:31] is definitely not the public narrative [234:32] out there yep I could see if you were an [234:37] Azure doubter and you were sitting there [234:40] at the top of Microsoft enjoying the [234:42] windows Monopoly the tremendous business [234:45] that is Windows and office and thinking [234:48] why would I do anything to jeopardize [234:50] this I mean Windows has self-reinforcing [234:53] Network effects everywhere huge [234:55] switching costs for the Enterprise super [234:57] profitable high margin one of the [234:58] greatest businesses of all time and now [235:02] there's this idea that you want me to [235:04] spend the money to run [235:06] servers people can run their own [235:08] software on my servers even if it's open [235:10] source so it's not feeding into my you [235:12] know Windows Centric ecosystem there's a [235:16] chance they're not paying for Windows [235:17] licenses right there's a chance they're [235:19] not even paying for my enterprise [235:22] software services like exchange or like [235:24] you know Windows server or whatever like [235:26] you're saying Ben they might be running [235:27] Linux on there or my competitor's [235:29] products it's not even as good a [235:31] business it's not zero marginal costs [235:33] running servers running these Big Data [235:35] Centers has huge costs and even if you [235:37] say oh those are fixed costs to Rack [235:39] them and you amortise them over the [235:40] course of but like energy has a real [235:43] cost it's kind of shocking that they [235:46] eventually did Embrace this very [235:49] unproven new business that could [235:51] potentially be way worse than their [235:52] current business totally and without [235:55] taking anything away from SAA because I [235:57] think he does absolutely deserve a ton [235:59] of credit for knocking it out of the [236:01] park on [236:03] execution I kind of think all of the [236:06] credit for the vision for it and the [236:09] championing it for the initial seven [236:11] years within Microsoft goes to Steve and [236:13] to Ray [236:14] yep now I will say the company stayed [236:18] the windows Centric company for too long [236:20] oh for sure yeah no doubt I mean Azure [236:22] was being built so it was successful [236:24] enough that it sort of erases everything [236:26] else but a lot of listeners know this [236:28] but I worked at Microsoft from 2011 to [236:31] 2014 my internship in 2011 was on the [236:35] word web app in the Office 365 Suite but [236:39] before it was called 365 and then my [236:41] real job was I worked on when I came [236:43] back off Office for iPad which was super [236:46] secret at the time it was really counter [236:49] strategy because we were the windows [236:50] company but at the same time What users [236:53] wanted in this world in 2012 was I want [236:57] to access my documents on any device [237:00] that I'm on we have moved to a world [237:02] where I have multiple devices I just [237:04] want to be able to use your application [237:07] on my device please and absolutely 100% [237:12] something that happened is all 200 of us [237:16] worked for multiple years to get these [237:18] things ready we had a ship date well we [237:20] had what we thought was a ship date [237:22] actually what happened was we were told [237:25] that actually we're going to shelf it [237:27] and instead of a ship party we had a [237:28] shelf party because the product got [237:30] cancelled o oh that's brutal oh I'm [237:33] sorry canceled what time frame is this [237:36] 2013 okay and basically it was hey we [237:41] just released Windows 8 we just released [237:44] the surface and we want the marketing [237:46] message for those things to be that [237:49] office is first and best on Windows and [237:52] the only tablet in the world that can [237:55] run real office is the surface and I of [238:00] course I'm too biased and too personal [238:02] to really think through this but I was [238:04] like oh this company has its head in the [238:07] sand this is ridiculous what users want [238:10] is we have a good version of Word Excel [238:14] PowerPoint that people can run on their [238:16] iPads and we've decided not to ship it [238:18] to try and Advantage surface and other [238:22] Windows 8 devices yes the year later we [238:25] did ship it actually right after SAA [238:27] became CEO that was one of the first [238:28] things he did so ultimately that [238:31] decision didn't happen that much later [238:33] than it would have otherwise and kind of [238:35] an open question of whether it was a [238:37] mistake like did Microsoft ever lose a [238:39] dollar for deciding to hold Office for [238:41] iPad another year probably not yeah [238:44] at the time I held this belief we have [238:46] stayed the windows company for far too [238:49] long and need to embrace users where [238:51] they are now with all this hindsight I [238:55] understand why you wouldn't make the [238:57] decision when you feel like the iPad [238:59] could be the end of you why would we go [239:02] all in on that now and put our finest [239:05] products to Advantage that thing when we [239:09] don't know if that thing is going to [239:10] kill us or not so there's the big [239:12] downside there's [239:14] not much upside to launching it what am [239:17] I going to renew a few more Enterprise [239:19] agreements because of it probably not [239:21] perhaps young Ben working at Microsoft [239:24] at that period of time failed to [239:28] understand how important it is to think [239:30] like an incumbent when you are the [239:33] incumbent and this was a low upside to [239:36] doing it right away plenty of downsides [239:38] to doing it right away really no risk on [239:40] sitting on it it really it was an easy [239:42] win for Satia during his first year to [239:44] say culture has changed here we are [239:47] shipping office on iPad we have shifted [239:49] from a devices and services company to a [239:53] cloud first mobile first company I [239:55] believe that was the message yes that [239:57] was the message and that was a great [240:00] supporting point to example of the [240:03] message yep well speaking of Transitions [240:05] and transitioning I think it is time to [240:08] wrap up our history of this period of [240:11] Microsoft and mobile and everything and [240:14] Steve's [240:14] tenure and talk about Nokia as we end [240:18] things here who bought Nokia oh that's a [240:21] good [240:23] question [240:25] okay in [240:26] 2011 after Microsoft had released this [240:30] Windows phone which like we said was [240:33] really kind of doomed to fail against [240:35] Android like you just couldn't compete [240:38] with free I think Bill Gurley had a blog [240:39] post about Android back in the day of [240:42] the less than free business model why [240:44] you can't compete with it yeah in fact [240:46] it's not that they were giving away for [240:47] free they were willing to pay people to [240:50] take it I mean if you think about it I'm [240:52] sure there was money that they spent on [240:54] the Droid marketing campaign I'm sure [240:56] there was money that they paid to the [240:58] carriers to pay to their sales people to [241:00] incentivize people to buy it versus the [241:02] iPhone and stores that was a common [241:04] practice in the mobile industry so I [241:06] think less than free is actually the [241:08] correct way to frame Android totally so [241:12] there was one phone OEM that was willing [241:18] to play ball with Microsoft and Windows [241:21] phone and that was Nokia well sort of I [241:25] mean Nokia basically had Symbian as its [241:29] OS they tried to start another OS [241:32] because Symbian was sort of reaching the [241:34] end of its life that wasn't going well [241:36] and so they were kind of left without a [241:39] platform and so they either needed to [241:41] pick Windows phone or Android as [241:44] platform of the future despite being [241:46] what used to be the dominant phone maker [241:50] for all cell phones [241:52] and the then CEO of Nokia was a guy [241:56] named Steven elop folks will almost [241:59] Surly remember that name he was a former [242:01] Microsoft guy and he had come over to [242:04] run Nokia so there were deep [242:05] relationships there in February 2011 [242:09] Nokia agrees to adopt the Windows phone [242:12] operating system as its primary [242:15] smartphone OS for its [242:18] devices like you said Ben it didn't have [242:20] a lot of options and it wasn't willing [242:22] yet to go Android pretty quickly though [242:25] as we get into 2012 [242:27] 2013 it's clear windows phone ain't [242:31] really working and Android is the future [242:35] so as we get into 2013 Nokia comes to [242:39] Microsoft and [242:41] says hey [242:44] we got to [242:46] talk we're going to go Android unless [242:49] you make it worth our while or you know [242:51] something happens here and changes and [242:54] as Steve put it to us it was only money [242:57] that is actually the right way to think [242:58] about it we've been joking about its [243:00] only money but honestly what was [243:02] Microsoft's market cap at this time call [243:04] it 300 billion and what was the [243:07] forthcoming acquisition offer for Nokia [243:10] seven so that is 2.3% of the company [243:14] you're willing to give up 2.3% of your [243:17] company for some particular bet I [243:21] actually think that's a very reasonable [243:22] way to think about this aquantive Skype [243:27] which we didn't talk about which [243:28] actually was a pretty good deal [243:29] especially because of the tax treatment [243:31] you know Yahoo Facebook you should think [243:34] of these things as a percentage of [243:35] market cap and sometimes things could go [243:38] really really right it's even less [243:41] consequential than that it's not even a [243:42] percentage of market capit all [243:45] Microsoft's operating income in 2013 was [243:48] $27 billion so it's $7 billion out of [243:51] $27 billion in just cash that they don't [243:54] know what to do with and that they are [243:56] getting credit for but your cash is [243:58] valued as a part of your market cap I [244:01] mean sure but Microsoft stock is in the [244:03] dumps here the cash flow geyser is not [244:06] appreciating the stock price here Wall [244:07] Street does not appreciate what's going [244:09] on it's funny I delivered you a [244:11] technically correct answer and you [244:12] delivered me back a very pragmatic one [244:14] yes right right which is that's not acre [244:17] to your market cap anyway so you may as [244:19] well spend it and I think it's probably [244:20] in Steve's mind here of like I'm not [244:22] getting any credit for this all this [244:24] cash I'm generating like right f it when [244:26] it doesn't cost you focus or your best [244:28] people or whatever the scarce resources [244:31] are it only costs you cash then you [244:33] should totally think about it as am I [244:35] willing to bet 2.3% of my company or [244:38] whatever percent of the cash that I'm [244:39] not getting any credit for if something [244:42] could go really right [244:44] it's a venture capital bet you know y [244:46] well and Nokia is basically holding a [244:49] gun to my head right there's actual [244:50] downside to it also there's actual [244:52] downside here right so that's how the [244:55] deal comes together it was super [244:58] controversial within the company and on [245:00] the board obviously at one point SAA [245:04] talks about in the book there's a straw [245:06] poll taken of all the division heads all [245:10] the top leaders in the company whether [245:12] they're for or against the acquisition [245:14] the majority are against the acquisition [245:17] SAA is against the acquisition the board [245:21] basically says to Steve like clearly [245:23] there's not support for this and so my [245:26] understanding of the timeline is there's [245:27] a seven half billion dollar offer on the [245:29] table Steve mulls it over plays with it [245:31] is for it proposes it to the board and [245:34] exactly the board comes back and says [245:36] hey there's not support for this okay [245:39] after that happens a series of [245:42] discussions [245:44] start and culminate we're in kind of [245:47] late summer early fall 2013 when all [245:49] this goes down and on August 23rd [245:53] 2013 Microsoft and Steve bomber [245:56] announced that he is retiring within the [245:58] next 12 months and that the board and [246:01] the company have started the search for [246:03] a successor as CEO that was August 23rd [246:08] 2013 yep on September 3rd so 10 days [246:12] later Microsoft agrees to buy Nokia's [246:16] mobile unit for S billion doar we heard [246:19] a bunch of different stories how it went [246:22] down we don't know exactly but the fact [246:25] pattern is Steve announced that he was [246:28] leaving 10 days later Microsoft agreed [246:31] to buy Nokia the question kind of [246:34] remains who bought Nokia [246:37] right we don't really know in any event [246:40] here's what happens next February 4th [246:42] 2014 Satia Nadella is introduced as the [246:46] next CEO of Microsoft Steve bomber steps [246:49] down on that same day Bill also steps [246:52] down as chairman of the board and John [246:54] Thompson becomes chairman of the board [246:56] so it is a wholesale Changing of the [247:00] Guard within Microsoft Bill Steve the [247:04] original folks were retiring we're done [247:07] it is a new day and that needed to [247:10] happen The Office for iPad discussion we [247:13] had a minute ago you know I think it was [247:14] emblematic like there is truth to what [247:16] Satia wrote in his book that we said at [247:18] the very beginning of the episode of hey [247:21] this company culture needed a reset you [247:23] it's kind of like a bigger version of [247:25] Brad Smith's presentation to the board [247:27] of it's time to make peace yeah like [247:29] it's time to make peace internally and [247:32] there just needed to be a reset yeah [247:34] there was a lot of baggage I mean it's [247:36] just what 40 Years of baggage Bill Steve [247:42] old Wars anti trust bad releases of [247:45] Windows like you just got to get it out [247:47] to move on and it was Bill and Steve [247:50] leaving in one Fell Swoop to clear the [247:53] path and at the same time everybody knew [247:57] hey there is a huge win that we are [247:59] sitting on right here like a huge huge [248:02] huge win in Azure and it is going to be [248:06] really good for everybody's personal net [248:08] worth if nothing else if we can just let [248:11] that be appreciated and let a new day [248:13] Dawn here so on February 4th 2014 on [248:17] that day Microsoft stock price was [248:21] $30.50 and as we said a minute ago the [248:23] market cap was I don't know call it $300 [248:25] billion slightly [248:27] below today 10 years later whoa whoa [248:31] whoa this is volume three David don't [248:32] yeah yeah yeah well you know just to [248:34] foreshadow show that this was the right [248:37] decision 10 years later stock is at [248:41] $465 market cap is $3.5 trillion dollar [248:47] probably like I don't know I haven't [248:49] done a sum of the parts analysis but I [248:50] think you can say probably the least [248:53] half is azure propping up that market [248:56] cap they are currently the most valuable [248:58] company in the world The Once and Future [249:01] King [249:04] Microsoft that's our story for part two [249:07] we still have a lot to talk about in [249:09] analysis I'm sure someone's looking down [249:11] at their podcast player right now like [249:13] why why are they acting like they're [249:14] done there's so much time after this are [249:15] they just gonna like play some music or [249:17] lot to talk about okay I have got some [249:20] start and finish stats on Steve's tenure [249:24] as CEO oh great this episode we started [249:27] a little bit before Steve took over [249:29] because we wanted to put the internet [249:30] chapter in and the antitrust chapter in [249:32] but I think everyone kind of feels it by [249:34] this point the question really is like [249:36] what happened when Steve was running the [249:38] company and so here are the numbers and [249:41] this is the time frame from 20 000 when [249:43] he was announced as CEO until 2014 was [249:47] SATA was announced so a 14year period [249:51] Revenue went from 23 billion to 84 [249:54] billion that's a 3 and a2x over 14 years [249:58] operating income went from 12 billion to [250:01] 30 billion so almost a [250:04] 3X important to pay attention to is the [250:08] price to earnings ratio when Steve was [250:11] announced as CEO was a 7 [250:14] 5x that's high it was real close to an [250:17] all-time high which was in the month [250:19] prior at an 80x so it is worth pointing [250:23] out it still has not been that high to [250:26] this day even today with all the [250:29] excitement around Microsoft AI [250:31] everything going on 40x right so Steve [250:34] comes in at an all-time high multiple [250:36] and right before the doj verdict and the [250:40] breakup of the company and the dot [250:42] bubble is exploding [250:44] yes and you're taking over from Bill [250:47] Gates all the things essentially if [250:49] you're doing an analysis of what [250:51] happened in Steve's tenure and you're [250:53] trying to grade that you are implicitly [250:56] saying did Steve make a good investment [250:59] to be honest I think Steve took one for [251:01] the team in taking over as CEO in that [251:03] moment he was handed a bit of a [251:06] impossible situation garbage [251:08] sandwich inheriting something when it is [251:12] valued that highly not to mention as we [251:14] talked about during that period all the [251:17] frankly going on at the company [251:20] completely so I worked at Microsoft [251:23] during this period I was a big open [251:25] source guy I was a big Apple guy I was [251:27] all these things and I hated Steve's [251:30] Windows strategy and frankly I didn't [251:32] like using any Windows products I felt [251:34] like they were all crap and it is still [251:38] true that it's totally insane to [251:41] evaluate how did someone do with an [251:44] asset that they were sort of forced into [251:46] buying at 75x earnings y so at the end [251:51] of his term it was 14x the PE multiple [251:54] went from 75x to 14x the market cap when [251:58] he was announced went from 600 billion [252:01] to when he left at 330 billion a lot of [252:05] that is basically the price to earnings [252:08] multiple rationalizing in that first [252:10] year and then after it did that the [252:12] stock price was basically flat for his [252:14] entire tenure no matter how much the [252:16] revenue or the profits grew and so one [252:19] crazy stat on this is you could have [252:21] bought Microsoft in 2009 for [252:25] 2.1x annual revenue oh my God I mean [252:29] everything was on sale back then but [252:31] like wow listeners are slight of hand [252:33] here we switch from earnings to revenue [252:36] but David I thought that too I was like [252:38] 2009 come on in [252:39] 2013 you could have bought Microsoft [252:42] stock for 3 x-ray [252:44] Revenue wow and so the question is why [252:49] why did investors give Steve zero credit [252:54] for any of this growth cut off that [252:56] first year when the multiple was coming [252:58] down why is it that effectively what [253:01] happened from 2001 to 2014 is for any [253:05] gains that they got in revenue or [253:07] profits it was offset by multiple [253:09] compression coming down and saying the [253:11] asset's still worth the same thing [253:13] one is very legitimately the investors [253:16] had little belief in Microsoft's [253:18] long-term relevance not the place for [253:20] user excitement not the place for [253:22] developers they doubted that there was [253:24] real vision from leadership I mean you [253:26] went from Gates this guy who created it [253:28] all to someone that everyone was [253:30] chalking up to be the sales and [253:31] marketing guy and there's the product [253:33] strategies all over the place and [253:35] windows isn't getting any more relevant [253:36] we trying all these new things that are [253:38] failing search passes You by social [253:40] passes You by blah blah blah but the [253:42] interesting thing is investors basically [253:45] didn't think windows and office [253:47] businesses were sticky and they were [253:49] only valuing the newer bets which was [253:52] super wrong windows and office have [253:54] proven to be these ridiculously durable [253:56] franchises generating more Revenue today [254:00] than ever so I mean it is ultimately on [254:03] the CEO to help shareholders understand [254:05] where the value is but shareholders [254:07] obviously did not price in the retention [254:09] and growth within the existing windows [254:11] and office customers through a new era [254:14] of technology I think people were just [254:15] betting that Microsoft would lose it and [254:18] they didn't they held on to these [254:19] durable franchises you know it's funny [254:22] when you asked this question a minute [254:25] ago I hadn't prepared for it ahead of [254:27] time because as listeners know we don't [254:29] share notes the first thing that popped [254:31] into my mind [254:33] about why Wall Street did not appreciate [254:37] the revenue and profit growth during [254:39] this time was just simply like Microsoft [254:42] did not do a good job telling it story [254:45] and I think you're saying the same [254:46] version here like it's so funny I mean [254:48] it's part of why I love doing acquired [254:51] part of why I think the show resonates [254:53] with people telling stories is the most [254:56] important thing if you cannot tell your [254:59] story right and in a compelling [255:01] fashion this is what's going to happen [255:03] to your stock price even if you triple [255:05] revenues and profits and build Azure and [255:08] all these things yeah I mean consumers [255:10] had no idea what Microsoft strategy was [255:13] and neither did developers so neither [255:14] did investors yep cios probably did sort [255:19] of but they were probably like what's [255:20] going on in Search and what's going on [255:22] yeah sorry what's Zoom is it winning [255:24] against iPod oh it's losing oh mobile [255:26] what losing too huh and it's for another [255:29] episode but it really was brilliant What [255:33] SAA did and the company did when he came [255:35] in of they got the story right the [255:38] messaging reset this is a mobile first [255:39] Cloud first company yep that was it that [255:42] was the key just saying those words over [255:45] and over and over again yeah anyone [255:47] who's listening who's a leader at a [255:48] company right now knows that the right [255:50] amount of repeating yourself to do is [255:52] about 10 times more than you think it is [255:54] you need to just keep delivering the [255:56] same message over and over and over [255:57] again and uh that wins the other way to [256:02] sort of look at Steve balmer's tenure is [256:06] comparing against what else was going on [256:08] in technology from 2000 to 2014 so on [256:12] the one hand like we've been talking [256:13] about you have the rise of Google in [256:15] Search and you have social networking [256:16] with Facebook and yes you absolutely can [256:19] compare a CEO to these category defining [256:22] startups that are in adjacent Fields but [256:25] that's a little bit of an odd way to [256:27] evaluate a CEO like how did you do [256:29] against they aren't even really [256:32] competitors of yours in your exact [256:34] market and by the way they created the [256:36] best businesses in history that were [256:37] also the fastest growing and capital [256:39] efficient how did you do versus those [256:42] two particular sort of related outliers [256:45] I think this is sort of a funny measure [256:47] even though this is the measure we kind [256:48] of all use but if you actually just look [256:50] at the peer set what other big companies [256:53] were there in 2000 in Tech you had Yahoo [256:57] AOL the whole cable and media sector you [257:01] had HP [257:03] Nortel I mean so many of the Great [257:05] companies of the previous era completely [257:08] fell apart the three who actually [257:11] survived and potentially thrived where [257:13] Microsoft Dell and only Apple after [257:17] Microsoft bailed them out and Steve Jobs [257:19] came back personally yeah I would throw [257:22] Oracle in there too but yeah yeah Oracle [257:25] but I mean surviving puts you in the top [257:28] 5% against the PE set of that era so [257:31] even if you overlook all the revenue and [257:33] profit growth and you just look at pure [257:36] Enterprise Value and relevance there is [257:39] actually a success in that the core [257:40] asset was preserved this whole whole [257:43] notion you have David that SAA came in [257:45] and we were great and then we sucked for [257:47] a while and then we were great again [257:48] even just setting up we preserved the [257:51] talent asset and that we had continuity [257:54] in our businesses for another 15 years [257:57] on what is already a 30-year-old [257:59] business I don't know that's way better [258:01] than anybody else does yep totally so [258:05] anyway this is all kind of analyzing the [258:07] tenure from a business perspective I am [258:09] very amenable to the idea that products [258:11] completely languish like I had no [258:13] interest in using any Microsoft products [258:15] during this period despite being an [258:17] employee of the company yeah I'm very [258:19] amenable to arguments of like yeah but [258:21] they didn't make anything good yep and [258:24] that is I think particularly resonant to [258:28] me at least to my history because they [258:29] used to they totally used to they used [258:31] to be the consumer technology leader [258:33] Windows 95 Windows XP everything we [258:35] talked about at the beginning of the [258:36] episode Internet Explorer the browser [258:38] Wars they were the leaders and they did [258:40] make some good you know Xbox is good [258:42] actually thought Windows Phone [258:44] particularly Windows Phone 8 was a [258:46] beautiful new crack at what do a phone [258:48] look like I thought it worked well but I [258:51] guess what I'm saying is the products [258:54] that ended up being their big profit [258:55] drivers were never their good products [258:59] right well they were their good products [259:01] just the Enterprise products they [259:03] weren't the good consumer products right [259:04] they weren't good for me as a user they [259:06] met the needs of customers yes all right [259:09] move into analysis great let's do it [259:12] seven power course so listeners this is [259:14] the part of the show analysis broadly [259:17] where we analyze the business sort of [259:19] after we've completed the story and the [259:21] first one is a section called seven [259:23] Powers which is named after Hamilton [259:25] helmer's book and the question that he [259:28] sort of poses is what is it that enables [259:31] the business to achieve persistent [259:33] differential returns or put it in [259:35] another way to be more profitable than [259:37] their closest competitor and do so [259:40] sustainably and there are seven [259:42] different Powers sort of categories that [259:44] it can fall into there's counter [259:46] positioning scale economies switching [259:50] costs Network economies process power [259:54] branding and cornered resource and I [259:57] think on part one we said Microsoft in [259:59] that era had all of these right I feel [260:02] like there were one or two that I was [260:04] shaky on but most yeah well in this era [260:07] they definitely don't have counter [260:08] positioning that's for sure right that's [260:11] the interesting thing once you're an [260:12] incumbent you can almost never have [260:13] counter positioning yep actually I would [260:17] say they had some of it in the [260:19] development of azure because they could [260:21] say to Fortune 500's we will do hybrid [260:25] cloud with you and we can be your [260:26] trusted partner in a way that AWS [260:29] couldn't counterposition against AWS [260:32] yeah yep but broadly as a company like [260:36] no way yeah they were getting [260:38] counterposed in Mobile I mean Google was [260:40] saying we'll give it to you free yeah [260:42] less them free y yeah all right perhaps [260:46] the single greatest asset they have is [260:48] scale economies with the number of users [260:51] and customers they have any investment [260:53] that they make gets amortized over such [260:56] a massive user base that it's worth it [260:59] if they can charge a dollar more on EAS [261:02] they should do almost any amount of [261:03] incremental R&D or Acquisitions right [261:07] and that translates directly into the [261:10] cloud era too yeah the cloud even more I [261:13] think there's crazy returns to scale on [261:15] cloud economics yep I think process [261:20] power I would argue they actually lost [261:22] during this era I mean the blackc [261:25] longhorn Vista thing illustrates that [261:29] yeah they went from knowing how to ship [261:31] the most beloved operating system of all [261:33] time with Windows 95 managing to pretty [261:35] much do it again even during the [261:37] antitrust thing with Windows XP with I [261:40] think zero blunders in between I mean [261:41] they had me but that wasn't a blunder as [261:43] much as like a I don't know Fresh coat [261:45] of paint that wasn't really real and [261:48] then yeah with Longhorn and Windows 8 [261:50] separate problems but completely forgot [261:54] why those franchises have economic value [261:57] yep I think they also definitely lost [262:00] branding power in the consumer world [262:04] yeah the question is that they become [262:05] more trusted by the Enterprise where if [262:07] you're offered the exact same service [262:09] from Microsoft and AWS are you more [262:12] willing to pay Microsoft for it [262:14] definitely so yes they gained it in the [262:16] Enterprise world yeah Microsoft has [262:19] unbelievable switching costs the EA you [262:22] just can't switch now you have to switch [262:24] sometime in the future and when that [262:26] time comes you're probably not going to [262:27] switch then either because that XDA is [262:29] going to offer even more value it's [262:32] funny to the extent that the doj and [262:34] governments were concerned about [262:36] Microsoft being a monopoly when it came [262:38] to product tying on the consumer [262:41] side the uh really should have been [262:44] concerned about product tying on the [262:45] Enterprise side where you pay us a [262:47] dollar amount per device on your network [262:49] and you get all of our software for sure [262:52] there would be way better Point solution [262:53] software for any one of the hundreds of [262:57] things that Microsoft is providing for [262:58] you but there's no way you're going to [263:00] switch yep you're so right it's so funny [263:02] you reflected it back to the doj it's [263:04] been so long now since we covered that [263:06] hours ago I kind of forgot about it in [263:08] that era there were literal switching [263:10] costs to getting a different browser the [263:13] one that came with your computer was the [263:15] one you were going to use because a [263:16] different browser was going to take like [263:18] 5 to 24 hours to download over your [263:20] internet connection in the dial up days [263:22] it was good for consumers to receive a [263:24] browser with their computer because [263:26] getting another one was almost [263:28] prohibitively difficult yep Network [263:31] economies it's actually a little thin [263:34] yeah I mean they had the great Network [263:35] economies with windows that we talked [263:37] about last time but that starts to a [263:39] road here as interoperability becomes a [263:41] thing as file format standardized and I [263:44] can open the same documents on Macs and [263:46] PCs it's like okay file formats stop [263:48] being a network economy that acrs only [263:51] to the operating system yeah I'm trying [263:54] to think are there any other like if you [263:55] were an Enterprise and become a [263:57] Microsoft customer and I'm an Enterprise [263:58] and I become a Microsoft customer do we [264:00] really benefit from each other being I [264:02] don't think so yeah I don't either that [264:05] just leaves cornered resource yep no I [264:08] don't think they have that meaningfully [264:09] no I don't think so I guess this power I [264:12] mean the fact that we came up with They [264:13] Don't Really counterposition they have [264:15] great scale economies they have a lot of [264:17] switching costs and that's kind of it [264:21] that's pretty illustrative of why you [264:23] kind of feel like this is the Lost era [264:25] of Microsoft and I think it also [264:27] illustrates that sounds like an [264:28] Enterprise company to me yep okay [264:31] Playbook Playbook well the first one [264:33] that talks to mind that I want to [264:35] address a little bit more specifically [264:37] is this idea of a cultural shift because [264:39] we've mentioned it many times on the [264:40] episode oh at the doj there was a [264:43] cultural shift oh with the new [264:44] leadership there was a culture shift but [264:46] what does it actually mean and how do [264:48] you go about quantifying that the thing [264:50] that I kept hearing in all the research [264:53] interviews we were doing was that when [264:56] the stock price was flat and flat for a [264:59] long time people became convinced it's [265:02] just going to stay flat Yeah so [265:05] basically whatever the cause of that was [265:08] it created a zero sum environment [265:11] nothing I do is going to make the [265:13] company more [265:14] valuable so therefore my value the only [265:17] way to grow value accur to me is to win [265:22] at the expense of someone else at this [265:24] company I'm going to get promoted over [265:26] them my product's going to eat their [265:28] product my teams going to eat their team [265:31] I get kudos and at the expense of them [265:34] looking like an idiot that's the kind of [265:37] incentive this is the cartoon org chart [265:40] of all divisions of Microsoft guns at [265:42] one another right and of course it's [265:45] Amplified by Stack ranking which I don't [265:46] have a problem with stack ranking [265:48] generally but famously at the company [265:51] everybody was ranked one to five every [265:54] manager was only allowed so many ones [265:56] and so many twos and so ultimately it [265:58] was an environment where everyone every [266:01] six or 12 months sometimes over midye [266:04] check-ins was kind of being baked off [266:05] against your immediate peers in your [266:08] group and because the company wasn't [266:10] growing in value you had to outcompete [266:13] your friends to win right the P was not [266:16] growing right so why was the P not [266:19] growing we talked a lot about that [266:21] there's a lot of reasons you could argue [266:24] why it wasn't but the culture is an [266:26] effect of that there's a big one I've [266:28] been thinking on which is how to go [266:30] about placing your bets for the future [266:33] as a company and I think in the 2000s [266:37] Microsoft was viciously trying to fight [266:39] against the tide there was open source [266:41] there was the WB there's always things [266:43] people wanted to do ultimately over time [266:47] you cannot fight what people want to do [266:50] as a company you can put up all these [266:51] barriers you can steer them back into [266:53] your ecosystem but ironically The [266:56] Playbook that Satya is now running is a [266:59] return to a classic Microsoft One [267:01] Embrace and extend yes rather than fight [267:04] what users are doing I want to use open [267:07] source software whatever I want to make [267:08] web apps I want to host a web app you [267:11] just figure out what people want you [267:13] embrace it and then you figure out what [267:16] product you can build with a business [267:18] model that extends that existing user [267:20] Behavior but it does require you to be [267:23] clever and outcompete a lot of other [267:25] people to invent that new business model [267:28] that you know is created on top of new [267:30] user Behavior I just want to double [267:32] underline and highlight this one because [267:35] I think [267:36] also this same Dynamic played out in the [267:41] building and evolution Microsoft's [267:42] Enterprise business really it just [267:45] wanted to control the network and [267:47] prevent users from messing it up right [267:49] eventually when the iPhone came out that [267:52] dam broke and it could no longer hold [267:55] back users within their company from [267:58] doing what they want and having what [267:59] they wanted and this is where the shift [268:02] to the cloud was another reason it was [268:04] so strategically important shifting to [268:07] the cloud is what enabled it to say okay [268:11] and become a partner to their users in a [268:13] way that they you know paid lip service [268:15] to before but they were really [268:17] antagonistic to their users yep and it [268:19] works exactly with what you're saying [268:21] for Microsoft as a company and its [268:22] products too like you want to use an [268:23] iPhone great you want to use open source [268:26] great we can still serve you yep and the [268:29] trick is figuring out how to make money [268:31] when you lean into what people want [268:33] because ultimately like if you just [268:34] reduce it all to economics what people [268:36] want is free value but you can't [268:39] actually build a business on giving away [268:40] free value yes you know I can give you a [268:44] dollar for 90 but ultimately I'm just [268:46] going to go out of business and so I [268:47] need to figure out some way that you're [268:49] happy from value creation and willing to [268:51] pay me more than it cost me to anyway [268:53] the trick is getting the business model [268:54] right yes there's this other one of what [268:57] was going on given that the ideas were [269:00] good and this is going to sound harsh [269:03] but timing implementation and taste at [269:06] Microsoft from call it Windows 98 on [269:12] were just terrible or maybe put another [269:14] way they had the right ideas but late [269:17] timing and bad execution strategically [269:19] correct but tactically misguided I mean [269:23] bill was super right that touch [269:24] Computing was going to be a thing he [269:26] referred to this idea of a natural user [269:28] interface very often bill was super [269:30] right that interactive TV was going to [269:32] be a thing I mean think about how I [269:33] watch Netflix I will watch Netflix [269:35] tonight after we record on my Apple TV [269:38] upstairs bill was right on mobile that [269:40] that was going to be a huge part of of [269:42] the Computing landscape and yet all of [269:44] these started at Microsoft 5 to 20 years [269:47] before the tech was actually ready and [269:49] they would often bet on the wrong [269:51] standard or Paradigm I mean touch [269:53] Computing ended up being capacitive not [269:55] resistive with a stylus tablets have [269:57] proven to be a cousin of phones scaled [269:59] up not PCS scaled down interactive TV [270:02] came after the internet not before and [270:06] only once there was a tremendous amount [270:07] of bandwidth I mean think about how much [270:10] more bandwidth it consumes for all of us [270:11] to ad hoc start NetFlix streams versus [270:14] there's one broadcast happening and we [270:16] all just tune in when we tune in and we [270:18] just catch whatever part of the [270:19] broadcast is over anyway not to mention [270:22] interactive TV looked like YouTube and [270:24] Netflix and not like you know a layer on [270:27] top of Comcast totally mobile was 5 [270:30] years early and it was more akin to [270:32] embedded devices than it was to scale [270:34] down PCOS so something was off in [270:37] Microsoft's ability to [270:40] leverage their [270:42] future predicting into creating the [270:46] right products right which is weird [270:48] because historically they have been good [270:50] at it yeah well they at least employed [270:54] the one Microsoft employee referred to [270:56] it as bracketing you basically develop [270:59] two products concurrently one aimed [271:01] below what the current technical [271:03] capabilities are and one aimed above and [271:06] as you get closer to shipping or as you [271:08] get closer to like letting the market [271:10] play out you kind of pick what whether [271:12] you're going to make the low-end one [271:14] better or you're going to sort of start [271:16] reducing functionality of the high-end [271:17] one and so in the IBM days you know you [271:19] had windows and os2 and in the internet [271:22] era you had like the web browser versus [271:25] all the interactive TV stuff or Longhorn [271:28] which was supposed to be little and [271:30] iterative versus black home which was so [271:32] ambitious it actually got cancelled yeah [271:34] the problem was during this era that [271:36] sort of optionality in multiple bits [271:39] kind of collapsed down to like no we're [271:42] make one bet in each of these yeah or [271:44] like the bets somehow couldn't continue [271:46] to flourish internally I don't really [271:48] know why but it seems like for some [271:51] reason bracketing worked well for a [271:53] while and then [271:55] eventually their ability to take a good [271:59] idea and implement it at the right time [272:01] the right way fell apart yep my next one [272:03] is the idea of positive sum leadership [272:06] this one's a little bit more personal [272:09] than our Playbook themes typically are [272:12] but I think it's an important takeaway [272:14] Bill Gates plus Steve Balmer in the [272:17] right roles with the right level of [272:19] respect for each other and who made [272:21] which decisions when that was all [272:23] humming that was way more valuable than [272:26] Bill alone or Steve alone y it was like [272:30] one plus one equals five yes they were [272:33] so great together this is actually [272:35] pretty common among teams the most high- [272:39] performing teams are so much better [272:43] together than they could possibly be a [272:45] part yeah well hell look at you and me [272:48] right I totally agree I was going to [272:49] make that analogy but is too much there [272:51] is no way we could do this on our own [272:54] yes too much Naval gazing but yes Bill [272:57] alone at least in this era was totally [272:59] at risk of getting too excited about [273:01] theoretical Technologies like win FS [273:04] that's the perfect illustration of this [273:07] and you know Steve needs a great [273:08] technology partner and one who kind of [273:12] has the extreme loyalty of the thousands [273:15] of brilliant engineers at the company so [273:17] they'll sort of align and follow the [273:19] vision and Steve also needed someone [273:22] willing to change their mind in the face [273:25] of new data bill was constantly [273:28] processing new information and as new [273:30] things came in he would say I don't care [273:32] how in motion things are if you're right [273:35] which is rare usually Bill's right but [273:37] if you're right and you're arguing [273:38] something to me like screw it we got to [273:40] change everything new data new thing the [273:43] internet Title Wave and Steve was much [273:45] more like we have to align an entire [273:48] aircraft carrier in the company and then [273:51] all the aircraft carriers outside the [273:53] company so we are going to make a [273:54] decision and then we are going to [273:56] implement and execute and I think [273:59] together there was some magic where [274:02] there was just the right amount of stick [274:04] tutiven versus adaptability yep my next [274:08] one is being extremely partner focused [274:11] is a gift and a curse Microsoft is an [274:13] extremely partner oriented company there [274:16] are far more profits who have accured to [274:18] Microsoft's independent software vendors [274:20] resellers Retail Partners than just to [274:23] Microsoft itself but it basically makes [274:26] it impossible to reset I mean Apple when [274:28] jobs came back hit a full reset all new [274:31] developer tools all new products all new [274:34] software all new platforms but when you [274:36] have all these externalities depending [274:38] on you you actually can't really hit a [274:41] set button to adapt for a new era you [274:44] have a whole ecosystem to preserve and I [274:46] think this is the more nuanced view of [274:48] the idea that well if you miss one wave [274:52] then you're actually well suited for the [274:54] next wave people often say the only [274:56] reason Apple was able to win in Mobile [274:58] is because they totally lost in desktop [275:00] or whatever and I think really what the [275:03] answer is is the more externalities you [275:05] have depending on you the more difficult [275:07] it is to reset and usually a Next [275:09] Generation the more switching costs you [275:12] have right a Next Generation technology [275:15] requires you to hit a big reset button [275:19] that's all I had for my playbook great I [275:22] have just one big one but I'm going to [275:23] save it for take away and Landing the [275:25] plane let's do that now listeners we've [275:28] been trying out this new way of ending [275:30] episodes how do we land the plane what [275:33] is the one takeaway that is really [275:35] sitting with me after having done all [275:37] this research talked through it with [275:39] David hardened our Thinking by counting [275:41] ideas off of each other what is the [275:44] thing that you can't get out of your [275:46] mind so for me this only came to me just [275:50] a few minutes ago but I think it's the [275:53] right and most complete version of what [275:56] I've been feeling about this part of the [275:59] Microsoft story for a long time since [276:01] we've been doing the research and the [276:03] feeling started with as we were talking [276:07] to people and digging in we were just [276:10] like this story is not understood right [276:15] and this narrative about these were the [276:19] losing years of [276:21] Microsoft yes there were a lot of L's [276:24] during this time but that's not complete [276:27] by any stretch of the imagination and as [276:31] we were preparing I really felt a lot of [276:33] weight on this one of like man we really [276:37] have a responsibility to try and get [276:38] this right here and I think what I [276:41] realized ized a few minutes ago as we [276:43] were talking is this was the biggest [276:47] failure of the company during this [276:50] period they did not tell their story [276:52] right and so much of what we think of as [276:56] the losses from this time frame and [276:59] certainly everything baked into the [277:00] stock price not moving was because of [277:03] that yes Steve came to the CEO role at [277:07] an all-time high multiple and like it [277:09] was the tech bubble and you know all [277:11] that stuff sure that's a big thing but [277:14] why did the stock price stay in the 30s [277:16] for his whole tenure they just couldn't [277:19] tell the story right and there are all [277:20] sorts of reasons for that but the story [277:23] does not have to be so negative because [277:28] there is so much positive that happened [277:30] during this time and yet the narrative [277:33] became this [277:35] self-reinforcing Microsoft sucks [277:37] narrative yeah irrelevant failing can't [277:41] do consumer the counterfactual is Amazon [277:44] if they had a consistent message that [277:46] they went forward with such as we are a [277:49] company who invents and wanders Amazon [277:53] has failed at so many things so many [277:55] things publicly huge bets that have [277:58] totally failed and yet huge consumer [278:00] failures the phone people are like what [278:03] a beautiful thing that Jeff Bezos has [278:06] imbued into this company this idea that [278:08] we invent and wander we make these bold [278:12] bets we Embrace failure I mean Kindle [278:15] Fire or I guess Kindle Fire kind of [278:17] counts certainly the phone I mean at [278:19] this point I feel like standing here [278:21] today we can say Alexa maybe if llms [278:24] hadn't become a thing I'd be with you it [278:26] turns out it might be good distribution [278:27] for a good llm if they actually yeah [278:29] sure but the thing itself anyway it's an [278:31] option on llms there's so many there's [278:33] too many to count there's grocery Amazon [278:36] Go Local I mean all the restaurant stuff [278:39] it's a narrative problem and yet the Nar [278:41] Nar about Amazon is H what a beautiful [278:44] thing Jeff andb and the company yes and [278:46] the narrative about Microsoft was they [278:48] can't get anything right yep you're [278:51] right I think yours is better than mine [278:54] now that's my new land the plane oh do [278:57] you want to hear what mine was before [278:58] yeah tell me what was on your mind all [279:01] right ultimately Bill Gates is Right [279:04] technology companies are always [279:08] extremely at risk of disruption even if [279:11] you won the battle today even if you're [279:13] the most dominant company today it is so [279:17] easy for you to lose and become [279:19] irrelevant tomorrow you may keep a great [279:22] business because these things are sticky [279:25] as we know IBM made a lot of money for a [279:26] long time but even without the whole doj [279:29] thing Microsoft probably would have [279:32] visted themselves Microsoft almost [279:34] certainly would have missed mobile [279:36] because there's no chance they would [279:37] have realized that the business model [279:40] that Google had meant that they were [279:43] going to win in Mobile when they came in [279:45] from the side and gave away the software [279:46] for less than free Microsoft was going [279:49] to have these huge Downstream misses [279:52] because technology moves so fast and is [279:54] such a dynamic landscape yep and I think [279:59] this is why I feel so adamant that [280:03] Microsoft during this era and Steve [280:05] deserve so much more credit than they [280:08] get because Microsoft is not IBM today [280:12] it is not large but irrelevant it is [280:15] very relevant and what they have done [280:17] with Azure and incloud and now with AI [280:19] is I mean hell they're the most valuable [280:21] company in the world if all this era did [280:26] was give them a free option to play in [280:29] the cloud and AI era or even just say [280:31] the AI era that would have been great [280:33] but also what they did was they tripled [280:35] revenue and profits right yeah they did [280:37] that while building this whole new great [280:40] business yeah [280:41] it's a great takeaway carve outs let's [280:45] do it okay I have two Hardware [280:47] technology products one is a recar out [280:50] from you past carve out the Rayband [280:53] metas oh yeah finally got a pair they're [280:55] great they're awesome the use case of [280:58] the audio ambient audio in my ears [281:01] without earbuds is great particularly [281:04] for a baby monitor when I am talking to [281:09] my wife or other family members or [281:11] friends or whatnot and I want to be able [281:13] to hear what is going on in the baby [281:17] crib and not wear earphones in my ears [281:21] great that is great they're also just a [281:23] great product in general another related [281:25] Hardware carve out is they startup [281:27] called Oslo and the Oslo sleep buds in [281:31] the last couple years I have slowly and [281:34] then pretty much every night gotten into [281:37] using some form of audio to help fall [281:40] asleep or if I wake up in the middle of [281:42] the night get back to sleep and I used [281:45] airpods for years and years and they're [281:48] great but you know if you sleep on your [281:50] side or do anything we like you know you [281:52] got the airpod jamming into your ears [281:54] these are little sleep buds that are [281:57] made for sleeping and if you lie on your [282:00] side on your ear they don't stick out [282:01] and so you can lie text me a link I'm [282:03] buying this immediately yeah they're [282:05] great so this is the team that was at [282:08] Bose that made the Bose I don't even [282:10] know what the product was called but [282:11] Bose had this product they killed it the [282:14] team left started a startup and so it's [282:17] all like Bose engineering anyway it's [282:19] great I love them I'm buying this as [282:21] soon as we get off yeah they're [282:23] fantastic awesome I have two and they're [282:26] like the most absolutely basic products [282:28] of all time and I'm okay with that the [282:31] first I mentioned earlier M3 MacBook Air [282:34] it's the finest computer I've ever owned [282:36] which I say every time I get a new Mac [282:38] if only you could turn it around and [282:40] touch the screen I know I'm rocking the [282:43] M1 MacBook Pro at home and it's a 16inch [282:48] and gosh that thing is just a beast to [282:50] fly with and so for all the travel we've [282:52] been doing recently David it has been [282:54] awesome to have this incredibly [282:56] lightweight incredibly fast just [282:58] beautiful machine for flights and all [283:01] sorts of travel stuff nice so it's my on [283:05] the go and then sticking with this theme [283:08] of staying incredibly basic and [283:09] predictable the Tesla Model Y is an [283:12] awesome car oh yes that's right you are [283:15] finally joining the club yeah we just [283:18] took a weekend and drove up to orcus [283:19] Island and it was so sweet I never once [283:23] charged it you know drove multiple hours [283:25] took a ferry were on an island that's [283:27] sparsely populated hung out for the [283:29] weekend drove all over the island did [283:31] the whole thing back got back with 18% [283:33] battery and had you needed to charge it [283:35] there'd be a supercharger Network yeah [283:37] and it's unbelievably fast and fun to [283:40] drive and I finally get the it's an [283:42] iPhone with wheels like it just feels [283:45] whenever I drive my other car it feels [283:47] kind of icky and this one feels clean [283:49] it's perfect yeah it's amazing I get it [283:52] I get it Tesla people all right we got a [283:54] lot of thank yous here yeah we have a [283:56] huge set of thank yous first of course [283:58] is our sponsors JP Morgan payments [284:00] service now and pilot you can click the [284:02] link in the show notes to learn more and [284:05] tell them that Ben and David sent you [284:06] when you get in touch I was trying to [284:08] count David it's definitely over 20 [284:10] people that we talked with this time so [284:13] on my end thanks so much to Brad [284:14] silverberg Brad LED windows for a while [284:17] notably the development of the Windows [284:19] 95 product and that team Thomas Reen who [284:24] was one of the original team members on [284:27] Internet Explorer and actually went on [284:29] years later to start control Labs which [284:31] sold to meta and as a part of meta's [284:34] effort now to do the um neural interface [284:36] you can sort of twitch your hands and I [284:39] don't know we haven't actually seen a [284:40] product yet so we'll have to to see what [284:41] that looks like but that was Thomas rear [284:43] in in his next ACT Steven sinowski who [284:46] led windows in Windows 7 and8 and led [284:51] office before that and David you read [284:54] quite a bit of Steven's words to prepare [284:56] for this yeah his blog hardcore software [284:59] he published in book form it's a [285:01] thousand page book it's like a textbook [285:03] sitting on my desk it's awesome we [285:05] talked to Stephen for a few hours he's [285:07] great he's a board partner in in Jason [285:08] horlitz now that was super fun [285:11] yep Julie Larson green was also great [285:14] she worked closely with stevenh in [285:16] office and also on Windows my old [285:20] cooworker anen rajas swaren who worked [285:22] with me on Office for iPad helped [285:25] refresh my memory on what the blowby [285:26] blow was like in those days where we [285:29] almost shipped the product then didn't [285:31] then got a new CEO then did huge thank [285:33] you to Fritz lman who worked in some [285:36] strategy and Corp Dev roles at Microsoft [285:40] just an awesome guy actually now he is [285:42] the CEO of the combined company class [285:47] pass and mindbody and talk about a [285:50] person with multiple lives and careers [285:52] he was great someone that he worked [285:54] closely with at Microsoft Charlie [285:56] songhurst Charlie's one of the smartest [285:58] people I've ever met I mean that could [285:59] go for a lot of people on this list [286:00] didn't Charlie do a great invest like [286:02] the best interview with Patrick a couple [286:04] years ago did he I got to go listen I [286:05] think he did yeah absolutely brilliant [286:07] mind expanding person John Rubenstein [286:10] who LED engineering at Apple and [286:12] actually went on to lead Palm it was fun [286:14] talking with him about what was it like [286:16] from the Apple side from the competitive [286:18] perspective competing against Microsoft [286:21] all these years huge thank you to Ray [286:23] Azie Who David and I both spoke with [286:26] yeah Ray is so damn delightful he's such [286:29] a legend delightful the right way to put [286:32] it Ray is now running another startup [286:35] new one called Blues Wireless and was [286:37] just so generous to give us a couple [286:39] hours he had some amazing you know [286:42] things that belong in a museum in his [286:44] office that he showed us over Zoom like [286:47] old computers and Hardware from the 70s [286:50] the 80s the 90s it was super cool yeah [286:55] more on my side to Rob Glazer who worked [286:57] at Microsoft in the 90s and founded and [286:59] ran real networks to Joe belfiore who [287:03] played a large role in Windows and [287:05] Windows phone and actually he demoed [287:07] Windows XP in the way back when 2001 the [287:11] launch announ reg Filman that's like how [287:13] the second half of the keynote works as [287:15] Joe is demoing features to Regis it was [287:17] J Leno for Windows 95 and rist for [287:20] Windows XP yes so Joe's got this [287:23] beautiful long history at Microsoft and [287:26] was just really great to VI Vara who was [287:31] at Microsoft deeply involved in sort of [287:33] the comms and legal stuff during the [287:35] antitrust era of course to Steve Balmer [287:38] being very generous with his time and [287:40] helpful in helping us sharpen some of [287:43] our thinking and especially being [287:45] generous with his time as we were [287:47] entering free agency [287:49] here he's got a busy day job these days [287:51] at the clips that's right our good [287:54] friend who runs the science of hitting [287:56] it's a great substack that does [287:57] investment analysis and just was very [288:00] generous sharing a large spreadsheet of [288:02] historical data from Microsoft it's very [288:04] easy to parse and look things up Live [288:06] while we're doing the episode and [288:08] finally last one from me to Todd Bishop [288:10] at Eire Todd has these unreleased [288:13] recordings from when he was covering [288:16] Microsoft at the Seattle PI way back in [288:20] the early 2000s and he sent me the raw [288:24] recordings you know when he was standing [288:25] there with Bill and Steve just being a [288:27] reporter and it's very fun to hear their [288:30] voices in ways that I don't think were [288:32] ever released or publicly heard yeah [288:35] super cool yeah a few more on my end [288:38] Terry Myerson the CEO of tretta Seattle [288:41] Terry ran windows and Windows phone at [288:43] Microsoft for a long time to s s sear [288:47] who is a managing director at madona but [288:49] is a legend in the server and tools [288:52] business at Microsoft so we talked about [288:55] him on part one but he's wonderful to [288:57] Mary Joe Foley it was so fun to talk to [289:01] Mary Joe Mary Joe dedicated her career [289:04] probably the last 20 plus years to [289:07] solely covering Microsoft and she is the [289:10] best in the business today she's the [289:12] editor and chief at directions on [289:13] Microsoft it's a research firm like [289:16] Gartner except it only covers Microsoft [289:20] she was super kind and generous and then [289:23] the last one Dave Mart it was so fun to [289:27] talk to Dave Dave was a [289:30] 33-year board member at Microsoft the [289:33] only outside Capital the only outside [289:35] Capital into the company before IPO from [289:38] tvi and then Dave went on to co-found [289:41] August Capital where he's a partner [289:42] Meritus these days I think Dave was the [289:45] longest serving Microsoft board member [289:47] besides Bill Gates three decades plus [289:50] wow oh and I have one more thanks to [289:53] good friend Arvin at worldly partners [289:56] for some of the research that he [289:57] provided as well so thanks arvand well [290:01] with that you should check out our [290:03] previous episode Microsoft volume 1 if [290:05] you've already heard that check out our [290:06] AWS episode we also recommend our Nvidia [290:09] series part one intersects nicely with [290:12] this era of Microsoft and we tell some [290:14] of the PC gaming story from the Nvidia [290:17] angle lastly of course if you are [290:21] sitting around right now and you're [290:22] thinking oh no what should I do next the [290:25] answer is acquired. [290:28] fmsf we cannot wait to see you at the [290:30] chase Center Mark freaking Zuckerberg is [290:32] going to be there it's going to be the [290:35] event of the century in the acquired [290:36] world so if you've always thought like a [290:38] I've always wanted to go to something [290:40] like Omaha you know for the Berkshire [290:43] hathway annual meeting where I've just [290:44] wanted to celebrate with other business [290:46] and technology nerds who also like [290:48] acquired this is going to be the [290:50] greatest way you could ever imagined to [290:51] do that yes if you are wondering what [290:54] you should be doing on September 10th [290:56] 20124 there is only one acceptable [290:59] answer and that is to be in San [291:00] Francisco at the chase Center [291:03] celebrating with us it's going to be [291:05] awesome Yep and with that listeners [291:08] we'll see you next time we'll see you [291:10] next time [291:11] who got the [291:13] truth is it you is it you is it you who [291:16] got the truth now huh [291:20] [Music]