Zuckerberg Locks In Voting Control to Shield Meta From Platform Wars

扎克伯格锁定投票控制权，以护航Meta抵御平台之战 Meta依靠创始人的绝对控制权与开源AI基础设施，绕开苹果的硬件限制并确保企业生存。 该内容按时间顺序详尽梳理了Meta二十年的发展历程，从Mark Zuckerberg童年对软件的热爱以及2004年在哈佛推出thefacebook.com讲起。内容追踪了公司早期的架构选择、从桌面端向原生移动端信息流的关键转型，以及通过收购Instagram和WhatsApp来对冲平台脆弱性风险的战略举措。最终，这一梳理重点突出了Meta出于生存驱动，向开源AI基础设施和先进增强现实技术演进，以此作为抵御封闭操作系统生态的长期防线。 Meta拥有前所未有的全球规模，月活跃用户达40亿，日活跃用户超30亿，旗下全套应用矩阵覆盖了全球约半数人口。 (1:11) Mark Zuckerberg童年时期对《席德·梅尔的文明》、计算机编程和古代历史的热爱，塑造了他的战略思维，并为ZuckNet等早期编程实验铺平了道路。 (7:19) thefacebook.com在哈佛的推出利用了高密度封闭社区冷启动策略，并通过验证.edu邮箱的限制确保真实身份，为该平台赋予了声望卓著的品牌光环。 (45:39) Friendster等竞争对手因计算全局关系连接而陷入N平方级扩展瓶颈并崩溃，而Facebook则通过将数据隔离在各大学网络容器内规避了这一问题。 (61:06) Shawn Parker从一开始就构建了Facebook的公司治理结构，以确保Mark Zuckerberg始终保持对董事会的永久投票控制权，避免其被职业经理人取代。 (77:04)

Acquired hosts Ben Gilbert and David Rosenthal argue that Meta's trajectory from a Harvard dorm room to four billion users was not accidental virality, but the result of relentless growth engineering, calculated monetization, and defensive hedges against third-party platform gatekeepers. Systematic Growth and Internal Process Power (335:25) The hosts challenge the notion that Facebook expanded through passive virality, framing its scale as an engineered outcome. Early growth teams identified an operational retention threshold: connecting new accounts with roughly seven to ten friends within their first ten to fourteen days to ensure feeds had immediate utility. This approach evolved into institutional 'process power'—a testing deployment system capable of serving multiple concurrent experiments, tracking statistical attribution, and executing automated code rollbacks when negative metrics triggered. Meta later decoupled critical operations from individual apps into centralized 'centers of excellence' overseeing infrastructure, growth, monetization, and safety under Head of Product Chris Cox. The hosts note that this optimization machinery faces natural physical limits: Meta commands four billion monthly active users out of an estimated 4.46 billion addressable internet users worldwide (excluding China's 940 million users, where Meta is banned). Monetization as a Data Relevance Discipline (236:09) Facebook initially lacked targeting sophistication, relying on an outsourced advertising partnership with Microsoft to reach profitability, with annual revenues climbing from $9 million in 2005 to $153 million in 2007. Under Sheryl Sandberg, the company built an in-house self-serve ad infrastructure, utilizing external embedded 'Like' buttons as tracking beacons across third-party websites to fuel behavioral ad targeting. Revenue surged to $3.7 billion by 2011. When mobile transition threatened advertising formats, product leader Andrew 'Boz' Bosworth approached monetization as an algorithmic relevance problem rather than an intrusive user tax. By optimizing ad targeting to function as content, Meta saw users begin utilizing platforms like Instagram specifically for product recommendations. Meta even observed user engagement drop in test markets where ads were temporarily deactivated, helping lift US and Canada average revenue per user from $11 at its initial public offering to $227. Platform Vulnerability and Defensive Acquisitions (254:49) The desktop-to-mobile shift eliminated Facebook's core technical advantage. Unlike the open web where Facebook pushed rapid updates, mobile environments subjected the company to two-week review cycles and structural platform control by Apple and Google. Furthermore, consumer preferences shifted away from public broadcast feeds toward private group messaging and ephemeral content on apps like Snapchat and WhatsApp. To counter platform erosion, Mark Zuckerberg adopted an explicit acquisition strategy revealed in unsealed 2012 correspondence. By acquiring emergent networks like Instagram ($1 billion) and WhatsApp ($19 billion), Meta bought twelve to twenty-four months of runway to systematically replicate competitors' novel mechanics at scale before rival network effects became entrenched. Meta's dependence on mobile platforms was underscored in 2022 when Apple's App Tracking Transparency framework erased an estimated $10 billion in advertising revenue, triggering a single-day $232 billion market capitalization drop. Open-Source AI and Hardware Independence (341:38) Meta's long-term technical strategy focuses on shielding itself from future software gatekeepers. The foundational machine learning capabilities developed by the Facebook AI Research division starting in 2013 enabled Meta to quickly reconstruct its feeds around un-connected creator content, neutralizing TikTok's algorithmic threat via Reels. In generative AI, Meta distributes its foundational Llama models as open source. The hosts compare this to Google's Android strategy: commoditizing foundational models prevents Meta from becoming subordinate to closed APIs run by rival providers. Concurrently, Meta's Reality Labs division—despite steep operating losses—developed the Orion augmented reality glasses prototype, which the hosts describe as an effort to establish proprietary control over the next major physical computing platform. While absolute founder control and aggressive technical adaptation have allowed Meta to survive repeated platform resets, the hosts conclude that its next existential challenge may be societal rather than competitive, centering on mounting public scrutiny and regulatory consensus concerning the impact of social media on youth mental health.

《Acquired》指出，Meta 的崛起与演进是一部通过极致工程化流程、绝对创始人控制权以及对平台依附危机不断突围的商业史。节目认为，Meta 屡次在平台迁移与监管风暴中存活，核心在于其将用户增长与广告变现塑造成高度工业化的“流程权力”，并依靠防御性并购、组织解耦以及对开源 AI 与硬件的重金布局，试图彻底终结受制于第三方操作系统的宿命。 创始人绝对控制权与重大战略博弈 (90:05) Facebook 在创立早期便奠定了由创始人完全主导的治理结构。早期参与者 Sean Parker 在搭建公司治理架构时，确保了 Mark Zuckerberg 拥有不可剥夺的绝对投票控制权。这一制度设计在后续的关键抉择中起到了决定性作用：2006 年夏季，面对雅虎因自身股价下跌 20% 而将收购报价从 10 亿美元下调至 8 亿美元的举动，尽管董事会和高管团队极力主张接受收购，Zuckerberg 依然凭借投票权强行拒绝。随后在 2007 年，Meta 获得了微软 2.4 亿美元的投资，对应估值高达创纪录的 150 亿美元，当时 Facebook 年营收仅约 1.5 亿美元且刚刚打平，这一高估值彻底颠覆了硅谷投资界的认知。 节目指出，绝对控制权给 Meta 带来的巨大收益具有强烈的幸存者偏差。Zuckerberg 在学习速度、面对错误时的极低虚荣心以及追求真相的务实态度上展现出罕见特质。然而，这种将整个庞大商业帝国的命运完全系于单一领袖个人判断的模式，在商业史上属于不可复制的极端个案。 工业级增长机制与相关性变现哲学 (187:13) 节目破除了 Meta 用户规模仅仅源于“自然病毒式传播”的神话，强调其连接数十亿人口的过程依赖于高度精密的工程干预。早期增长团队明确了关键的留存阈值——即新用户必须在注册后的 10 至 14 天内添加 7 到 10 位好友，以此为算法提供充足数据来构建高粘性的信息流。这种体系构成了 Hamilton Helmer 所定义的“流程权力”：新入职的工程师能够立即承接自动化代码部署、故障自动回滚以及复杂的跨实验统计归因系统，持续优化转化漏斗。 在商业化变现方面，时任高管 Andrew Bosworth 摒弃了将广告视为对用户“征税”的传统思维，将其重新定义为“相关性与数据优化问题”，确立了“广告即内容”的工程哲学。在 Sheryl Sandberg 建立的自助式广告体系及全网追踪机制推动下，Meta 的变现效率实现爆炸式增长：其在美加地区的用户平均收入（ARPU）从 IPO 时的 11 美元跃升至现今的 227 美元，全球新兴市场平均 ARPU 也达到 44 美元，验证了高相关性广告对平台价值的拉动。 移动端平台依附与防御性重构 (254:49) 从桌面互联网向智能手机移动端的迁移曾对 Meta 构成致命威胁。在开放网络时代，公司能够自由推送代码、快速试错；而在移动端，代码更新受制于苹果与谷歌长达数周的审核周期与分发限制。更为严峻的是，桌面端能够将各种功能集成于单一门户的网络效应被打破，手机屏幕上的独立应用使竞争格局彻底重置。 2012 年 2 月公开的内部邮件证实，Zuckerberg 的核心策略是通过收购 Instagram 等新兴应用来为自身争取 12 至 24 个月的战略窗口期，防止竞争对手形成网络效应，随后再利用 Facebook 的庞大体量消化其产品机制。面对用户从公开“城市广场”向私密“起居室”（如 Snapchat 与 WhatsApp）的分享习惯转变，以及 TikTok 算法推荐引擎的冲击，Meta 在产品负责人 Chris Cox 的主导下彻底重构了组织架构，打破传统独立应用孤岛，建立由 Facebook AI Research (FAIR)、基础设施、商业化变现等跨业务“卓越中心”统一支撑的运作体系，从而成功利用沉淀多年的算法底座迅速推出 Reels 予以抵御。 苹果生态封锁下的底层自主突围 (304:45) Cambridge Analytica 数据泄漏事件不仅重创了 Meta 的公众声誉，并在 2018 年 7 月引发单日 1190 亿美元市值的暴跌，更让苹果找到了切入隐私战役的战略切口。2021 年苹果正式推出应用追踪透明度（ATT）政策，直接导致 Meta 在 2022 年损失约 100 亿美元的广告收入，并在该年 2 月的财报公布后引发单日市值暴跌 2320 亿美元（跌幅达 26%），创下当时商业史上的单日市值蒸发纪录。 这一重创迫使 Meta 全面转向谋求底层计算自主权。在用户规模逼近天花板的背景下（除中国以外全球约 44.6 亿网民中，Meta 月活已达 40 亿），公司实施了激进的技术布局：一方面，Meta 投入巨额资本开支打造 Llama 系列开源大模型，采取类似 Google Android 的开源策略，防止未来 AI 时代被专有闭源生态敲诈或封锁；另一方面，Reality Labs 虽累计承受数百亿美元的巨额运营亏损，但公司仍坚定押注 Orion 等轻量化 AR 眼镜原型，试图抢先定义并掌控摆脱智能手机约束的下一代计算平台。 节目总结认为，Meta 的驱动力兼具高度理性的资源分配与对技术连接的偏执追求。然而，其未来面临的最严峻挑战或许不再是平台税或算法竞争，而是来自社会维度的伦理审判：若科学界与公众对社交媒体实质性损害青少年心理健康形成广泛共识，这种根本性的价值质疑将构成 Meta 历史上未曾遭遇过的生存级风险。

[10:47] Mark Zuckerberg: five whole hours to just sit and play on my computer and write software [24:50] Adam D'Angelo: Synapse was cool and we had a bunch of people downloading it but this feels like something more on the order of AIM itself. [27:28] Adam D'Angelo: Mark gets way too lucky. [44:50] Mark Zuckerberg: I don't really know what the next big thing is because I don't spend my time making big things. I spend time making small things and then when the time comes, I put them together. [96:30] Michael Moritz: Never have we paid so much for so little. [137:37] Mark Zuckerberg: Calm down, breathe, we hear you [144:34] Josh Elman: if you're a developer and you can get someone to bother 10 friends to get one more user to join, you're very happy. [168:31] Mark Zuckerberg: thought ads were gross [224:40] Cheryl Sandberg: Mark, nobody can fire you and only you can fire me so if you're in, I'm in. [250:50] UK Government Report: While the models showed some success in correctly predicting attributes on individuals whose data was used in the training of the model, the real world accuracy of these predictions when used on individuals whose data had not been used... was likely much lower. [303:15] Steve Jobs: thank you, we're just not going to let somebody else build a platform on top of our platform. [329:40] Host: Meta addressable users who aren't yet users is less than 6% of humans. Just wild. [344:39] Mark Zuckerberg: We must ensure that we always have access to the best technology and that we're not locked into a competitor's closed ecosystem where they restrict what we build.

[0:00] all right so I was up late last night [0:03] late for me as a dad is like 11: p.m. [0:05] but I'm sitting here at my computer in [0:07] my dark basement you know pulling notes [0:09] together David what music did I have on [0:12] you have one guess uh oh oh I know [0:15] exactly what you had on you had Trent [0:17] resner social network [0:20] soundtrack it makes anything you're [0:21] doing feel uh you know twice as [0:23] important and twice as revolutionary and [0:26] uh it just felt very apt for this [0:28] episode oh man so I have been listening [0:31] the last 24 hours to 50 Cent into club [0:34] because that came out my freshman year [0:36] of college same year as the facebook.com [0:39] and like man 50 c [0:43] facebook.com can't get any better than [0:45] that perfect match made in heaven well [0:48] I'll check out our wall to- wall from [0:49] the old days and see if there were any [0:50] posts about [0:52] that all right let's do it let's do it [0:55] who got the [0:57] truth is it you is it you is it you who [1:01] got the truth [1:02] now is it you is it you is it you sit me [1:06] down say it [1:08] straight on the [1:11] wayu welcome to the fall 2024 season of [1:15] acquired the podcast about great [1:17] companies and the stories and playbooks [1:19] behind them I'm Ben Gilbert I'm David [1:21] renthal and we are your hosts today we [1:24] are studying a company whose products [1:26] are used by More Humans than any other [1:29] compy company in history meta of course [1:33] formally known as Facebook so I figured [1:36] I would contextualize these numbers a [1:38] little bit meta has 4 billion monthly [1:41] active users and daily active users are [1:44] over three billion yeah nuts there are [1:48] only 8 billion humans on Earth so as I [1:51] started to brainstorm what the closest [1:53] competitors could be to serving half of [1:56] the humans I thought surely I can find [1:58] it in Empire or governments from the [2:01] past where there is some larger [2:03] percentage Yeah makes sense nope the [2:06] Roman Empire at its peak was only 40% of [2:10] humans tops you know the data is a [2:11] little bit hard to actually find from [2:13] that period of time but the British [2:15] Empire which we have a little bit better [2:16] handle on at its peak was only 23% of [2:19] the global human population so no [2:22] government tech company utility Etc has [2:25] ever addressed so much of the world it's [2:28] just wild there's no other way to put it [2:30] in the over 20 years since its founding [2:33] meta truly has connected Humanity [2:35] through its apps Facebook Instagram [2:37] WhatsApp messenger and now threads So [2:40] today we're going to study how they did [2:42] it there's been a lot of ink spilled [2:45] writing about Facebook over the years [2:46] and for its first 10 years most of that [2:48] writing focused on the many benefits to [2:51] society with breathless exuberance over [2:53] Milestone after Milestone and for the [2:56] past 10 years it's kind of seemed like [2:58] meta could do nothing right reporting [3:01] focused on its many stumbles the massive [3:03] mistakes the incredible controversies [3:05] surrounding the company and while we [3:07] will of course discuss these events as [3:09] part of our story and Analysis our goal [3:12] today on this podcast episode is really [3:15] to understand how it is that meta became [3:18] the dominant fabric that connects the [3:20] human race and why they've been so [3:22] successful at continuing to win over and [3:25] over again no matter what you think of [3:27] the company it is undeniably one of the [3:29] most most important institutions in the [3:31] world and their global scale is no [3:34] accident it is the result of careful [3:36] actions from some of the most motivated [3:38] and brilliant people in the world who [3:41] believe in one Mission connecting as [3:44] many people as possible I had thought [3:46] about this till now but almost sort of [3:49] tautologically this is like the most [3:51] important episode we'll ever do to try [3:53] and understand this like it's the [3:55] biggest company that has ever existed [3:58] it's the biggest thing that's ever [3:59] existed [4:00] Yeah by these measures that we're [4:02] talking about right now right now of [4:04] course the story of how we got here is [4:05] nuts it is the perfect acquired stew [4:07] like you're saying David they speedran [4:09] their startup phase they swerved through [4:11] multiple disruptive technology waves [4:13] they battled Fierce competitors they [4:15] invented or maybe discovered one of the [4:17] greatest business models ever and [4:19] they're now trying to pull forward the [4:20] next technology generation through sheer [4:22] force of will with ar VR and AI so [4:26] finally listeners we tackle one of the [4:28] greatest corporate stor stories ever [4:30] Facebook the Mark Zuckerberg production [4:34] so listeners we have one big [4:36] announcement for you today it is time [4:37] for our annual acquired survey if you [4:39] have 3 to 5 minutes please click the [4:41] link in the show notes or go to [4:42] acquired. fm/ survey to take it we'll be [4:45] raffling off a pair of shiny new meta [4:47] rayb bands and giving away a bunch of [4:49] acq dad hats as well this is really our [4:52] one big ask of you all year and it helps [4:54] us immensely with making the show better [4:57] to hear your suggestions and also to [4:59] help our sponsors understand just how [5:01] impactful the acquired audience is so go [5:03] to acquired. fm/ survey and David and I [5:07] are both eternally grateful after this [5:09] episode discuss it in the slack and [5:11] check out acq 2 our second show where we [5:13] just had CLM Delong the CEO of hugging [5:15] face on to talk about his view on how [5:18] the open source AI ecosystem will play [5:20] out and before we dive in we want to [5:22] briefly thank our presenting partner JP [5:24] Morgan payments yes just like how we say [5:27] every company has a story every company [5:29] story is powered by payments and JP [5:31] Morgan payments is part of so many [5:33] companies Journeys from seed to IPO and [5:35] Beyond so with that this show is not [5:38] investment advice David and I may have [5:39] investments in the companies we discuss [5:41] and this show is for informational and [5:43] entertainment purposes only David Roman [5:47] Empire like what's our starting place [5:49] here no man we got to start in 1984 [5:51] we're never going to finish this [5:53] episode and really I mean where else can [5:56] you start I can make up you know Roman [5:58] Empire classic escalus whatever t-shirt [6:02] saying Mark has this day of the week but [6:05] Facebook has always started and ended [6:08] with Mark Zuckerberg so we start in May [6:11] of 1984 when Mark Elliot Zuckerberg is [6:14] born as the second of four children to [6:17] Karen and Ed Zuckerberg in DOS fairy New [6:20] York small little Suburban Town suburb [6:23] of New York City yep and his mom is a [6:27] psychologist and his dad is a dentist is [6:29] that right that's right Ed studied [6:31] Dentistry at NYU he was always like [6:35] really into math and computers though [6:37] but growing up the son of Jewish [6:39] immigrants in New York City you know the [6:41] expectation was you're going to be a [6:42] doctor or a lawyer so he goes to [6:44] Dentistry School becomes a dentist he [6:47] meets Karen they get married after Ed [6:49] finishes dental school they move up to [6:51] dobs fairy in Westchester County and [6:54] Mark and his three sisters grow up as [6:56] prototypical middle class suburban [6:58] American Kids in the '90s Life Is Good [7:01] your grandparents worked really hard to [7:03] make a bunch of sacrifices for your [7:04] parents your parents are now [7:05] professionals they're also working hard [7:08] to give opportunity to you and the [7:10] family is kind of leveling up in America [7:12] it's the classic story here you can play [7:15] sports and you can hang out with your [7:16] friends and play video games indeed and [7:19] for Mark and for me growing up and maybe [7:22] you too like it's kind of amazing like [7:24] there's no reason not to just explore [7:27] your passions and for Mark those [7:30] passions become three-fold one [7:33] turn-based strategy video games most [7:36] particularly Sid Meers [7:38] civilization two programming computers [7:42] and three as you mentioned and everybody [7:45] who watched our Chase Center show knows [7:47] ancient Greek and Latin history classics [7:50] the classics and pretty much in that [7:52] order of importance both to him [7:54] personally and the future course of [7:56] Facebook and meta so let's start with [7:59] civilization [8:00] September 1991 David I toyed with the [8:03] intro to this being and we are your [8:05] hosts civilization is a video game upon [8:08] which but I decided not to right I mean [8:11] it's kind of laughable but like no this [8:12] is really important here so September [8:16] 1991 when Mark is 7 years old [8:19] civilization or Civ as its fans call it [8:22] comes out for PCs and young Mark gets [8:26] hooked and for those of you who haven't [8:28] played Civ it is a very particular type [8:31] of turn-based strategy game and is [8:33] probably the foremost example of what is [8:35] known as the fourx genre that's the [8:39] number four letter X and 4x stands for [8:44] explore expand exploit exterminate and [8:49] despite that sort of Nefarious sounding [8:51] nature to the last one there especially [8:54] Civ and 4x games were kind of this [8:57] radical branch of video games at the [8:59] time where the action of playing and the [9:02] way to win wasn't just about like [9:04] killing all the enemies on screen it's a [9:07] lot more like a board game or like Risk [9:09] if you're familiar with that it's about [9:11] strategy it's about growing and [9:13] marshalling your resources effectively [9:15] and most importantly there are multiple [9:17] ways to win the game most other games [9:20] out there you can win by defeating all [9:22] the other players but you can also win [9:25] by completing a technology development [9:27] tree and reaching like the peak of [9:30] technology which is in the game [9:32] launching a spaceship to go colonize a [9:34] new planet so that's another way to win [9:36] and in modern versions of the game you [9:38] can also win by diplomacy by getting [9:42] elected the leader of the UN you can [9:45] also win by dominating all the other [9:47] players on either culture or religious [9:50] Dimensions too so it's kind of like this [9:54] amazing simulation of what it's like to [9:57] run one of the top 10 Mark cap Mega [10:00] capitalization companies in the world [10:02] today but to your point this idea that [10:06] it's turn based you try to amass [10:09] resources figure out how to deploy your [10:11] resources have multiple concurrent [10:13] strategies so that you as the role of [10:16] the dice of the universe happens and [10:18] things unfold before you and your and [10:20] other players make their moves yep right [10:23] to be able to effectively react to it [10:25] all and use your resources to win in [10:27] whatever way winning means to you you [10:29] it's kind of funny this is way too [10:31] simplistic of an analogy and you know [10:33] not fair to mark on any Dimension but [10:35] the whole construct here sort of reminds [10:37] me of the book Enders Game you know [10:39] where like it's these kids growing up [10:40] playing this game and they didn't [10:43] realize all along it was actually the [10:45] real [10:47] game H so playing Civ and other video [10:50] games when he's young PC games leads [10:52] young Mark to want to learn how to [10:54] program himself so when he's 10 he has [10:57] his parents take him to the local Barnes [10:58] & Noble one day [11:00] and begs them to buy him C++ for dummies [11:03] and he dives into the book learns [11:06] everything in there and he's like I I [11:08] need more I need more so after that his [11:10] parents hire a local tutor to come and [11:12] tutor him for an hour a week in [11:15] programming and learning how to write [11:16] software and Mark has this quote that he [11:19] says later I'd go to school and I'd go [11:21] to class and come home and the way I'd [11:23] think about it was awesome I have five [11:26] whole hours to just sit and play on my [11:28] computer and write software and then [11:30] Friday afternoon would come along and it [11:32] would be like okay wow now I have two [11:34] whole days to sit and write software [11:37] this is amazing it really is astonishing [11:40] as a 10-year-old to be learning C++ this [11:42] is not basic this is a language where [11:45] you are thinking very close to the metal [11:48] you have to be very aware of the [11:50] constraints of your system of manually [11:53] managing memory it's impressive for 14 [11:57] or 16 year olds to be learning [11:59] simplistic languages this is a whole [12:01] different ball game as it always is with [12:03] Mark so while he's a kid here you know [12:06] in DBS fairy He codes up all sorts of [12:08] little projects including a family chat [12:11] tool for all the networked computers in [12:13] the house and also his dad's Dentistry [12:16] office is attached to the house so he [12:18] builds this network called Zuck net [12:21] which is family chat between all the [12:22] computers very pre here and then [12:26] speaking of family and friends chat in [12:29] 1997 when Mark is 13 years old AOL [12:34] launches aim AOL Instant Messenger oh I [12:38] remember it well many many many hours [12:41] lost to aim and uh warning my friends [12:45] and other dynamics of the aim system to [12:48] gang up on people and to have other [12:50] people gang up on me and oh man punting [12:53] people off aim and yes having to create [12:55] a new screen name and log off and log on [12:58] and message and pretend you're someone [13:00] else I mean the whole this was Middle [13:02] School this was middle school for all of [13:04] us and this was middle school for Mark [13:06] too the crazy thing about aim to go on [13:09] just a quick little digression here I [13:11] had no idea till doing the research it [13:13] was actually like a rogue Skunk Works [13:16] project within AOL did you know this hm [13:19] no yeah so it was a couple Engineers [13:21] that just went off and built this almost [13:24] against the wishes of management it was [13:26] not originally intended to really be for [13:28] social social use or for kids at all it [13:31] was more like the idea they had was it [13:33] was going to be kind of like slack like [13:35] chat for workplaces they ended up using [13:38] it internally at AOL to communicate and [13:42] that's actually where away messages come [13:44] from so they came up with away messages [13:46] because people would be in meetings or [13:48] you know this is the days of desktops [13:49] there's no cell phones barely even [13:51] laptops people would go out to lunch and [13:53] they'd have to put up a message so that [13:55] when their colleagues were trying to [13:57] communicate with them be like no no I'm [13:58] in a meeting [14:00] that's the origin of messages that's [14:03] fascinating and it is interesting how [14:05] you ended up kind of using messages [14:06] after a while even while you were there [14:08] just to indicate status it was the [14:10] ultimate emo teenager thing it was like [14:12] I am expressing my whole self through [14:14] this on line away message 100% so I mean [14:18] for people all of our age Mark's age at [14:21] the time I mean that completely changed [14:23] our social experience when that came out [14:26] so the other important thing about aim [14:29] when it came out that I had no real [14:31] conception of at all as a kid was that [14:34] it was free so my family was an AOL [14:37] family like we were AOL subscribers we [14:40] had the you got mail Etc but you didn't [14:42] have to be a paying AOL subscriber to [14:46] download and use aim and join the [14:48] network aim goes on to become this total [14:53] viral thing in North America I mean kind [14:56] of concurrent with Napster a little bit [14:59] before Napster it's the first piece of [15:01] consumer viral software and so by 2001 a [15:06] couple years after launch it has 36 [15:08] million users including as we've been [15:11] talking about basically every middle and [15:13] high schooler with access to a computer [15:15] and you know 36 million users doesn't [15:18] sound like a lot today but you got to [15:20] remember the internet was way smaller [15:23] back then way way way smaller not just [15:26] in terms of number of people had access [15:28] but also number of people who were [15:30] literate in how to use it totally it's [15:33] the early Broadband era is that era [15:35] we're talking about and you're right [15:37] tens of millions of people at least in [15:38] the US had internet access at this point [15:41] in time yep and basically all of them [15:45] and certainly all the young people were [15:46] on Aim so the other important thing [15:49] about aim and AOL here is it was pretty [15:52] easy to hack but like oh man there was [15:56] just a field day that kids and teenagers [16:00] had with like aimbots punting software [16:04] Wares hacking programs if you were [16:08] someone like me you were just [16:09] downloading this stuff off of FTP [16:12] servers on the internet and installing [16:14] it on your machine and it let you kind [16:16] of be a punk and like kick people off of [16:20] aim by bombarding them with messages you [16:22] know you could do some like Teenage Punk [16:24] stuff but it also let you extend the [16:27] functionality of what AOL and aim was [16:30] you could add graphics you could add [16:32] emojis you could draw pictures add links [16:35] in your profile yeah oh the asky art in [16:37] your profile and your away message [16:39] totally and like this becomes such an [16:42] important part of teenage culture on the [16:46] internet as we're growing up it's like [16:47] all this personalization of like hey I [16:50] don't want just what AOL gives me I want [16:53] to control this so for me I was [16:56] downloading and using this stuff you [16:58] might have been writing it although you [16:59] might have been a little young no I [17:01] didn't learn PHP [17:03] until I was 15 or 16 so call it 2005 [17:10] okay so yeah a little later yep Mark is [17:12] one of the kids that's writing this [17:14] stuff and you know I think that's where [17:16] he got his started he was doing Zuck net [17:18] he was doing all these little projects [17:20] but like you know here's AOL this sort [17:22] of unintentionally infinitely hackable [17:25] and extensible canvas that he can play [17:27] around on so that's number two for Mark [17:30] you know passion number two computers [17:32] and programming and then passion number [17:34] three is Classics so Mark when he enters [17:38] High School he'd gone to public school [17:39] all through Elementary and Middle School [17:41] in there in dobs fairy in Westchester [17:43] County he does two years at the local [17:46] public high school at Ardsley high [17:48] school and after two years at Ardsley [17:52] Mark is like well you know a couple [17:55] things I'm not sure that this is the [17:57] right high school for me for the rest of [17:59] my high school career one I think I [18:03] would enjoy just being in a sort of like [18:05] better school like a higher level of [18:08] competition higher level of learning you [18:10] know he's definitely at the top of the [18:12] class at Ardsley two is you know I'm [18:14] really into Latin and Classics and like [18:17] I'm going to tap out on that real fast [18:19] here at Ardsley there's got to be a [18:21] place I can go where I can really study [18:23] this and that takes us to Philips exitor [18:27] Academy in New hampsh right yes in New [18:30] Hampshire and so exitor is one of these [18:33] private boarding high schools that have [18:35] existed basically as long as iy League [18:37] colleges have and they're there just [18:39] kind of as feeder schools into Ivy [18:41] League colleges so Mark applies to exter [18:44] gets in and the summer before he's about [18:47] to go the school hosts an event in New [18:51] York City in the city for incoming two [18:53] years to meet each other get to know the [18:56] school get ready Etc and then there mark [18:59] meets another kid who's a lot like [19:02] himself who also is really into [19:05] computers really into aim and everything [19:08] going on there and Mark's like yeah man [19:11] I really made the right choice like [19:13] everybody here is going to be just like [19:15] the two of us and that person is Adam [19:19] D'Angelo future CTO of Facebook amazing [19:23] they met in high school now the great [19:24] irony of this story is that Mark and [19:27] Adam ended up being like the only two [19:28] kids who are really into computers in [19:30] the class it was just the two of them [19:32] they just happened to meet in New York [19:34] City at the prechool event which is [19:36] probably one of the last few years that [19:38] that would be the case I'm sure if you [19:39] went to this school today it'd be 80% of [19:41] the classes or 100% of the classes [19:43] really into computers yes I think in [19:45] large part because of Mark and Adam it's [19:47] quite different these days yep so the [19:50] two of them when they get to exiter [19:52] become fast friends they work on all [19:54] sorts of computer projects together [19:56] software projects and [19:59] by the end of their senior year to [20:01] graduate from exitor you have to do a [20:02] senior project and you know most people [20:05] do this like a mini college thesis or [20:07] they do something in the community etc [20:09] etc Mark and Adam are like now we're [20:11] going to build some software this is [20:13] what we do so mark one day is listening [20:16] to music on his computer I think [20:18] probably on Winamp which is how [20:21] everybody listened to MP3s that you got [20:23] from Napster back in the day I think I [20:26] listened on sound Jam but it's just [20:28] because I had a mac oh yeah future seeds [20:32] of iTunes yep that's right it got [20:34] acquired by Apple and then eventually [20:37] they I think built off the same code [20:38] base and turned it into iTunes but [20:40] everyone used winam that was the aim of [20:42] music players and in fact AOL I think [20:44] would acquire Winamp a couple years [20:46] later ah bring it all into the fold so [20:50] Legend has it Mark is listening to a [20:52] playlist in Winamp one day during his [20:54] senior year and he gets to the end of [20:57] the playlist and he's like [20:59] why did the music stop I listen to so [21:01] much music all the time on Winamp it [21:03] should just know what I like and [21:05] automatically pick the next song it [21:08] should just be a DJ and then inspiration [21:10] hits and he calls up Adam and he's like [21:12] I've got our senior project we are going [21:15] to build an aid DJ plugin for Winamp and [21:21] they call this synaps and like they did [21:24] it these two high school kids they build [21:25] an AI you know AI wasn't called AI at [21:28] the time but like a you know a [21:29] recommender system that looks at the [21:31] data of what songs you've consumed on [21:33] Winamp and predicts what song you're [21:37] going to like to consume next I mean [21:38] it's basically Spotify today what year [21:40] is this this would have been spring of [21:43] 2002 yeah real early days so they build [21:47] this thing and then they release it to [21:49] the internet and people start using it [21:51] it starts getting some adoption and um a [21:54] couple companies including Winamp itself [21:58] reach out to Mark and Adam and are like [22:00] hey do you want to come work here like [22:02] can we buy this yeah and I think [22:04] Microsoft was one of them too it was for [22:05] like a million dooll acquisition offer [22:08] or something like it wasn't you know [22:09] crazy but these are high school kids [22:11] right in 2002 yes but yeah this is [22:14] really important because they're about [22:16] to graduate Adam is going to go off to [22:19] college at Caltech Mark is going to go [22:21] off to college at Harvard in the fall [22:23] but Mark is not just like any student [22:25] coming into Harvard he's like I can [22:27] build these things and they have value [22:29] and big companies are going to be [22:31] willing to pay me for them I don't need [22:33] to be that worried about getting a job [22:35] right he sort of always knew he had a [22:37] great fallback plan so it kind of let [22:39] him be risk on so Mark graduates he goes [22:41] off to Harvard it's the Natural Choice [22:44] on many fronts I mean one mark of course [22:47] Aces his SATs I think he gets a 1600 on [22:50] the SATs there's this great legend about [22:52] um you had to take sat 2s to get into [22:55] college back then I don't know if you [22:57] still do but I think you had to have a [22:58] minimum of three Mark doesn't study or [23:01] prepare for them at all he walks in the [23:03] day of the test says I'm going to take [23:05] all three that I'm going to take on the [23:06] same day he gets 800s on all of them [23:09] College admissions was not going to be a [23:11] big deal and his older sister Randy was [23:13] already there at Harvard so clearly G to [23:17] go to Harvard if he wants yes and I [23:20] think he wants enough like I I don't [23:22] know that it's that important to him but [23:23] he's like yeah sure Harvard it's the [23:24] best I'm going to go there yep in fact [23:27] there's a video of that that's right [23:28] have you seen this his dad sort of [23:30] filming his reaction and he's pretty [23:32] measured about it it's very yep got into [23:35] Harvard and then kind of goes back to [23:37] doing something else on his computer yep [23:39] and I think for most kids for most [23:42] people in the world you would look at [23:45] that and be like what arrogance and look [23:47] it's not like 18-year-old Mark is not [23:49] arrogant but I think it's also this is [23:52] why all this stuff we've been talking [23:53] about is important it's just not that [23:56] important to him you know he likes [23:58] building stuff he likes making stuff he [24:00] knows it has value he knows people will [24:02] use it whether he goes to Harvard [24:05] doesn't go to Harvard goes to college [24:06] doesn't go to college it's just not that [24:09] big a deal to him I don't think yeah [24:11] reading in between the lines it seems [24:12] like the thing that he was most excited [24:13] about is oh good a challenging [24:15] environment where I will encounter other [24:16] really smart people like myself who are [24:18] ambitious yes I think that is totally [24:21] right so Mark gets to Harvard in Fall of [24:24] 2002 and you know like you said he goes [24:27] to class he does well but he wants to [24:30] meet other people and be around other [24:32] smart folks he ends up joining the AI [24:34] fraternity there like my co-host here my [24:36] esteemed colleague Ben indeed also my [24:39] father while he was in college ah I [24:41] didn't know that yeah that's right [24:43] that's right a pip Adam meanwhile of [24:45] course goes off to Caltech you know [24:47] 3,000 miles away in the opposite corner [24:49] of the country also can I just say how [24:51] crazy is it that Adam D'Angelo would [24:54] become Facebook's first CTO they don't [24:57] go to college together right well I was [25:00] going to get into this in a sec Adam has [25:02] a very different experience at Caltech [25:03] he ends up graduating from Caltech I [25:05] think um they probably do not have the [25:07] same generous leave policy there that [25:09] Harvard does because Harvard's is weird [25:11] right it's like you can leave and if you [25:12] ever want to come back you just it's [25:14] like you never left oh we're going to [25:16] come back to that in one sec put it been [25:18] in that but for Adam at Caltech he still [25:20] finds time while he's there to keep [25:22] working on these projects so during his [25:24] freshman year he builds something called [25:27] buddy Zoo you know kind of building on [25:29] synapse and the success that they had in [25:32] their senior project buddy Zoo takes [25:36] your buddy list from Aim from AOL and [25:41] effectively turns it into a social [25:43] network so you upload your entire buddy [25:46] list to Buddy zoo and then you get your [25:49] friends to do so too buddy Zoo then [25:52] analyzes and gives you a list of who are [25:55] the friends of your friends how many [25:57] degrees of separation [26:02] [Music] [26:04] your are of are it measures popularity [26:08] among users this is starting to sound [26:10] pretty familiar here and this is pretty [26:12] advanced stuff for 2002 [26:14] 2003 very advanced stuff especially for [26:17] a side project for a very busy Caltech [26:19] student yep so buddy Zoo pretty quickly [26:23] in the months after Adam launches it [26:26] gets a couple hundred thousand users [26:28] there's gravity to this thing it's [26:31] becoming fairly big especially given how [26:33] small the internet was at the time yeah [26:35] and Adam is like whoa okay this is [26:38] interesting so this is kind of what [26:41] scale looks like synapse was cool and we [26:43] had a bunch of people downloading it but [26:45] this feels like something more on the [26:47] order of aim itself like this is really [26:49] interesting and the important thing with [26:51] synapse was it was a single user [26:53] application so it could analyze my songs [26:56] and tell me what to listen to this is [26:58] the first time Adam Mark you know [27:01] particularly Adam in this case is [27:03] discovering a very different type of [27:05] application that gets better as more [27:07] people join it yes now we say Adam and [27:10] and Adam definitely built and launched [27:11] it but it's not like Mark wasn't [27:14] involved they're staying in touch the [27:16] whole time and collaborating and working [27:18] through coding challenges together on [27:21] what else on AIM it's kind of serving [27:24] its original use case like they're [27:25] collaborating through it across the [27:27] country so that summer 2003 after their [27:30] freshman year Adam comes to Boston and [27:32] gets an internship at the MIT media lab [27:35] he discovers this thing called [27:36] Friendster another social network out [27:39] there and Adam is like whoa this is a [27:42] dedicated Standalone kind of version of [27:45] what I just built with Buddy Zoo this is [27:47] like really interesting and it's all on [27:50] the web now friender was based out here [27:54] in Silicon Valley had raised money from [27:56] Kleiner Perkins and benchmark right oh [27:59] Benchmark 2 I didn't know that [28:02] Y is what tied their hands and why they [28:05] could not pursue an investment in [28:08] Facebook that's funny not that I think [28:11] that was a real possibility anyway [28:12] because things were kind of moving so [28:14] fast but you never hear them in any of [28:15] the discussion about who was pursuing [28:17] Facebook and who was iced out they just [28:19] they had a competitive investment that's [28:21] right it's funny I never even think [28:23] about that because obviously Matt kler [28:25] would go on to become such an important [28:27] partner at Benchmark and of course Matt [28:29] was an early executive at Facebook yes [28:32] exactly interesting so here's frster [28:35] this example of like a real Standalone [28:37] company social network growing virally [28:41] funded by Venture capitalists based in [28:43] the San Francisco Bay Area Adam's like [28:46] oh maybe that's sort of what we should [28:49] be focusing on and by the end of the [28:52] summer in 2003 frster had over 3 million [28:55] users so like really interesting growing [28:59] really quickly frster had real scale and [29:02] grew really fast that is something that [29:04] has kind of been lost to history frster [29:06] is like the butt of a lot of jokes but [29:08] it exploded out of the gates true so [29:11] Adam's playing around with frenster and [29:13] frenster had this feature called [29:17] testimonials which was kind of like a [29:19] Facebook wall but in the long run one of [29:22] the core problems with Friendster was it [29:25] positioned itself as a social network [29:26] and about friends but really it was kind [29:30] of intended to be a dating site [29:32] Underneath It All absolutely that was [29:35] wink wink not a dating site it was this [29:37] idea that like there's all these other [29:39] things that are dedicated to dating so [29:41] there's this weird stink about it but if [29:43] you just naturally meet a friend of a [29:45] friend online through a thing that's [29:47] totally not a dating site then great [29:50] there's no Stigma around it yep so uh [29:54] specifically when it comes to [29:55] testimonials the purpose of testimonials [29:57] was supposed to be like why you should [29:59] date somebody or why they would be a [30:01] great partner which is funny so Adam [30:04] while he's playing around with this and [30:05] he's telling Mark about it he actually [30:08] posts on Mark's frster profile a [30:11] testimonial and he says that quote Mark [30:15] gets way too lucky and he posted this in [30:19] the summer of 2003 and Adam was telling [30:22] me the story and he was like you know I [30:23] have thought about that post for 20 [30:26] years since then [30:28] because on the one hand it was like a [30:30] stupid throwaway you know double on [30:32] Tandra that I put on there because it [30:33] was a you know a dating site on the [30:35] other hand it is both absolutely true [30:40] about Mark that he always gets way too [30:42] lucky and Facebook and meta always gets [30:44] way too lucky but it's not luck yeah [30:47] it's actually Mark there's something [30:49] very different about the way he operates [30:52] and all the reasons that meta and [30:54] Facebook get lucky over the years Mark [30:57] was always that way and it's always [30:59] because of Mark Mark is unbelievably [31:01] good at positioning himself for luck to [31:03] do its thing and specifically at this [31:05] point in time you know what Adam talked [31:07] about is like he was always looking [31:09] around he was looking at frster he was [31:12] looking at what we did with synaps he [31:14] was looking at what I did with Buddy Zoo [31:16] he was looking about the projects at [31:17] Harvard he did that we're about to talk [31:19] about and he was always thinking like [31:22] how do I incorporate all of this and [31:24] make whatever I'm doing better this was [31:27] the stew that Facebook is about to come [31:29] out of so this brings us to fall [31:31] semester 20032 2004 Mark sophomore year [31:36] and the famous Kirkland House Suite [31:39] h33 with Mark Zuckerberg Chris Hughes [31:43] Dustin moskovitz and Billy Olsen as the [31:46] four roommates there three of the four [31:48] of those names are names you may know as [31:50] the founders of Facebook yes so Mark is [31:53] focused heading into his sophomore year [31:55] at Harvard and he's not focused on [31:57] classes he's focused on projects and [32:00] he's done like 10 I mean this is [32:02] something that again is kind of lost to [32:03] history people talk about oh there's the [32:05] one face smash thing that he did that [32:06] led to Facebook he did like 10 side [32:09] projects I mean I remember being in this [32:11] era of my life where you feel like you [32:12] have superpowers as a programmer and [32:14] you're looking around and you're like oh [32:16] I can make a website for that I can make [32:17] a website for that I had the same thing [32:20] where the world wasn't saturated with [32:22] apps yet and so you could just like make [32:24] things that made your life better or [32:26] other people's lives better or cool ways [32:28] to connect people and I distinctly [32:30] remember feeling like how come nobody [32:32] else realizes you can just do this I [32:34] know what that sort of feels like and I [32:36] know that mentality he was in there were [32:38] 10 things that he had worked on even [32:40] before face Mash yep so the first of [32:43] which right as they're getting back to [32:45] campus that fall is Mark builds a tool [32:47] called course match oh yeah that's right [32:50] and this is as students are shopping for [32:52] classes for fall semester you can upload [32:55] which classes you're thinking of taking [32:57] or which class you've signed up for and [32:59] you can see who of your friends have [33:02] also signed up for that class or are [33:04] planning to sign up for that class this [33:06] is just text all this is is a list of [33:09] classes and a list of people who sign up [33:12] and it takes Harvard by storm I mean [33:16] think about it right like it's obvious [33:17] in hindsight a bunch of college [33:19] undergraduates yeah you want to take [33:21] classes you're interested but like [33:23] really you want to make sure you take [33:24] the classes with the people that you [33:26] want to be around I was literally in my [33:28] friend yesterday who was at Harvard at [33:30] this time and she said oh yeah people [33:32] totally chose their classes based on who [33:34] was in them of course of course so Mark [33:37] sees this and he's kind of like you know [33:39] again it's one of these learnings an [33:41] observation about the world he's like [33:42] wow people are spending a lot of time on [33:46] this tool it's literally just lists of [33:47] people you click on a class name and [33:49] people would spend a lot of time just [33:51] combing over that list of people totally [33:53] there's nothing visually compelling here [33:55] so he's like well okay now it's time for [33:57] my next project you know people have [33:58] registered for classes what if there was [34:01] something visually compelling for people [34:03] to spend time on what if I overstepped a [34:07] little bit and uh yeah and this is face [34:10] Mash well chronicled in all the many [34:13] books Stories the Steven Levy book The [34:16] David Kirkpatrick book movies you know [34:19] what Happ movies movie movie Mark codes [34:23] this up in one evening in his dorm in [34:25] Kirkland house and it's basically the [34:27] website hot or for Harvard with a couple [34:30] like actually interesting mechanical [34:31] twists on it now the two big problems [34:35] with it are one this is probably just [34:37] not a good idea or not like kind [34:40] generally it is a head-to-head voting [34:43] mechanic on which picture is quote [34:45] unquote hotter than the other picture [34:47] the other thing that is really not great [34:49] is rather than having users upload their [34:52] own pictures Mark hacks into the [34:55] Kirkland house servers and downloads the [35:00] Facebook photos of all the students to [35:03] populate the website lowercase face [35:06] space lowercase book yes so we should [35:10] probably spend a minute and talk about [35:11] what a Facebook was before Facebook so [35:15] my big question for you is did Ohio [35:17] State have a Facebook it did not it [35:19] probably would have been like a [35:21] telephone booksize thing if they did I [35:23] mean I went to college in the fall of [35:25] 2007 and so you could look up people's [35:28] email addresses if you had their first [35:30] and last name you might be able to get [35:32] some other maybe major but it was a text [35:35] blob it wasn't a photo well I think most [35:38] if not all Ivy League schools back in [35:40] the day had these Princeton where I went [35:42] certainly had these when I showed up on [35:44] campus in fact Phillips exitor had one [35:47] the private high schools did too that's [35:49] right that's right so you had to give [35:51] the University or in exter's case the [35:54] high school a school photo of you when [35:57] you were coming in before you [35:58] matriculated and then they would [35:59] literally make a book like a printed [36:01] book with the names and class years and [36:05] photos of all the incoming students from [36:07] the freshman class and I think if I [36:10] remember right at Princeton at least [36:11] they would republish the book every year [36:14] you know for your class because some [36:15] people would leave some people would [36:16] join etc etc all right so these are [36:19] photos stored digitally that Mark then [36:22] while plugged into the campus network [36:24] from his dorm room is access [36:28] downloading and then putting up on his [36:30] own website that he is hosting yes [36:33] called face Mash now couple interesting [36:35] things about face Mash one gosh if Mark [36:38] was blown away by the engagement of the [36:41] Harvard population with course match [36:43] face Mash takes that to such a whole [36:46] another level that it actually melts [36:48] down the servers in Kirkland house and [36:50] the IT department at Harvard has to shut [36:52] off internet within hours of it going [36:55] live to all of Kirkland house L this [36:59] like continue to spread virally within [37:01] Harvard and take down like all of the [37:03] servers oh so maybe he didn't actually [37:05] rehost the images maybe he was just [37:06] pointing at the URLs that were hosted by [37:08] Kirkland house oh I didn't think about [37:10] that yeah that might have been how it [37:12] worked but either way dude knows how to [37:14] make something that gets engagement yes [37:17] yes he does so the other interesting [37:19] thing from Mark's perspective about face [37:22] Mash is that it was the first time he [37:25] was building a rich web app as one of [37:29] these projects you know fully built on [37:31] the lamp stack and he had probably used [37:34] the lamp stack Linux Apache MySQL PHP [37:38] for course match and so to be clear [37:41] that's an operating system a web server [37:44] with Apache MySQL the free open source [37:47] database PHP the free open source [37:49] programming language yes but this is the [37:52] first time that he's building like a [37:53] real Rich consumer grade web app on top [37:57] of the lamp stack Rich consumer grade [38:00] it's the technical requirements of [38:03] course match but with photos uh photos [38:06] and voting sure yeah you have like [38:08] probably an additional table of [38:10] information in the database or something [38:12] but the big takeaway here is a of course [38:14] the use case shouldn't have done it B [38:16] wow it's engagement C this is training [38:19] wheels of how to use these new [38:21] open-source web [38:23] Technologies if you go back two three [38:26] four years you're going to have to go to [38:28] Oracle you're going to have to go to [38:30] Microsoft you're going have to buy like [38:31] Enterprise grade super proprietary [38:34] systems to do this and this is all just [38:36] like free and something you could Cobble [38:38] together and upload onto some Linux web [38:40] hosting and boom there had to be a [38:42] hundred times more a thousand times more [38:44] web applications created by the Mark [38:47] Zuckerberg's everywhere in this period [38:49] of time doing whatever random little [38:51] project that they thought would be fun [38:52] or funny or useful yep it's such a [38:55] perfect example despite its problem atic [38:57] nature of this is like the first time in [39:01] history where you need neither money nor [39:04] permission to launch an application like [39:07] this on the internet totally $10 domain [39:10] $100 a year web hosting that includes [39:12] the database no licenses required PHP is [39:16] all free maybe the bandwidth would have [39:18] been an issue but this is on the order [39:19] of $100 to do this yeah well clearly the [39:21] bandwidth was an issue because they shut [39:24] down internet to Kirkland house so David [39:27] this feels like a thing you should get [39:28] kicked out of school for doing yeah [39:30] Harvard as you imagine is not pleased [39:32] about this they haul Mark in front of [39:34] the administrative board which is a [39:36] disciplinary committee for students when [39:38] things like this happen he and his [39:40] fraternity brothers are pretty sure that [39:42] he's actually going to get thrown out of [39:44] school for this so the night before the [39:47] ad board meets to make their final [39:48] decision AI the fraternity throws a [39:52] goodbye Mark party that night and Legend [39:55] has it that this is where Mark and [39:58] Priscilla Chan meet for the first time [40:00] and Priscilla of course is now Mark's [40:02] wife yep so thead board meets the next [40:06] day and decides not to throw Mark out of [40:08] school but to put him on disciplinary [40:10] probation for the rest of the year [40:12] basically don't do this again slap on [40:14] the wrist slap on the wrist so Mark [40:17] doesn't end up getting kicked out of [40:20] Harvard but he does kind of end up [40:22] becoming like the Harvard computer [40:25] celebrity even before Facebook [40:27] right I mean this a big deal everyone on [40:30] campus knew about face Mash a lot of [40:32] people already knew about course match [40:33] anyway so he's now the guy who can make [40:36] websites and web applications here at [40:38] Harvard that people use Y and it just so [40:41] happens that there is a group of upper [40:43] classman at Harvard led by two twins the [40:46] winkl Vos twins who have an idea to [40:50] digitize Harvard's Facebook and build an [40:54] online version of it that they're [40:56] thinking of calling the Harvard [40:57] connection then they think maybe we [40:59] could actually take that to other [41:00] schools too and call it connect you as [41:02] sort of the broader [41:04] version this was not at all a unique [41:06] idea not at all lots of people at lots [41:09] of schools had this exact same idea in [41:13] fact Harvard itself had this idea the IT [41:16] department at Harvard had been promising [41:18] for years and talking about how they [41:20] were going to build a digital version of [41:22] the Harvard Facebook and frienders out [41:25] there and I looked it up St for had Club [41:28] Nexus columia had CU Community Yale had [41:31] Yale Station I'm sorry there's a whole [41:33] movie made about the drama of how novel [41:35] this idea is it's not totally not I'm [41:38] pretty sure Princeton had one too when I [41:40] was there pre Facebook yeah totally not [41:43] a novel idea lots of people have it no [41:46] the winkl vast twins though you know [41:47] Mark is now this celebrity on campus [41:50] they need somebody to be a programmer to [41:52] write the site so they talk to markk and [41:54] Mark agrees that he's going to help them [41:55] out and help code this site that they're [41:58] thinking about also again the hutzpah to [42:01] say we have this idea you are a [42:04] programmer you will program our idea [42:08] clearly Mark can come up with his own [42:10] ideas that get people excited enough to [42:13] use the stuff he builds now so far it's [42:15] been a little unsavory but he doesn't [42:17] have a problem coming up with ideas with [42:19] product Market fit and executing them [42:21] end to end the only thing he hasn't done [42:23] so far is made money made anything as a [42:25] business but what other pieces of the [42:28] puzzle does he really need from some [42:29] people with an idea none it's funny I [42:32] was going to save this for a little [42:33] later but let's talk about it now in [42:36] addition to technically because of Open [42:38] Source and the lamp stack and also [42:41] socially because of aim and everything [42:42] going on the world sort of for the first [42:45] time something like Facebook being able [42:48] to be built for no money and no [42:52] permission also this was the first time [42:55] that a technical person [42:57] because of this could just do everything [43:00] like Mark didn't need anybody Mark [43:01] didn't need a non-technical co-founder [43:03] Mark didn't need a CEO he was the CEO I [43:06] remember thinking the same thing when I [43:07] shipped my first app to the App Store [43:09] co-founder Ian and I in [43:12] 2009 I think made something called sees [43:14] the day got over a million downloads two [43:16] programmers uploaded something to the [43:18] App Store it's a slightly different era [43:19] because that's mobile not web but I [43:21] remember looking around being like Whoa [43:23] We like didn't need a business guy right [43:26] that's the crazy thing you can make [43:28] stuff and you can put it in the world [43:31] totally so this takes us to Harvard's [43:33] Christmas break now at the time Harvard [43:35] did final exams for the fall semester [43:38] after Christmas break in January so the [43:41] idea is all the students go home for [43:42] Christmas break you come back in January [43:44] and then there's a week of reading [43:46] period where you ref familiarize [43:48] yourself with all the material from your [43:49] classes then you take final exams then [43:52] there's a little break there's a course [43:53] shopping period for spring semester and [43:55] spring semester starts in February so [43:58] Mark has a few weeks to program a new [44:00] idea exactly to not study and program a [44:03] new idea and during Christmas break he [44:06] had actually come out to Silicon Valley [44:08] to the Bay Area to hang out with some [44:11] friends out here and seen you know the [44:14] physical embodiment of all these great [44:16] tech companies Yahoo early Google Etc [44:19] drove by them all and so he comes back [44:22] and he's like extra extra motivated he's [44:25] like you know this digital Facebook idea [44:28] I think this is something that a lot of [44:30] people want I bet I can code this up in [44:33] a week and just launch it and again not [44:35] a business just like a project he has [44:38] this gut feeling that people will use [44:40] this if he makes it totally now [44:43] importantly though all the stuff he'd [44:45] been doing throughout the year he's like [44:48] well I can incorporate that into this [44:51] project too it'd be pretty boring if [44:54] it's just a digital Facebook like I [44:57] don't think people want that but I've [44:58] made all this other stuff that people [45:00] really really like and engage with and [45:02] when you say just a digital Facebook you [45:03] mean like essentially the profile page [45:05] would just be a photo and a name right [45:07] just like the physical version of like [45:09] here is a list of all the students in [45:11] the class and here is their you know [45:12] stock profile picture y so later after [45:16] the facebook.com blew up Mark told the [45:19] Harvard Crimson in an interview I don't [45:21] really know what the next big thing is [45:23] because I don't spend my time making big [45:25] things I spend time making small things [45:28] and then when the time comes I put them [45:29] together that's the kind of stuff I do [45:31] small little projects and eventually [45:33] they all fit together and that's what [45:36] the facebook.com ends up being so Mark [45:39] gets back from Reading period instead of [45:41] studying he Dives in he IM AMS his [45:43] friend Andrew McCullum who he knows from [45:45] CS classes and asks Andrew to do the [45:48] page design Andrew says yeah sure thing [45:51] I'll help you out Andrew goes and fires [45:53] up Photoshop he finds an image on the [45:56] internet of some guy who looks kind of [45:58] cool and puts it in the header of the [46:01] design for the site behind a cloud of [46:04] ones and zeros because it's cool it's [46:06] like the Matrix yes it's gradiented you [46:09] know it's sort of white on the left blue [46:10] on the right and for like you know a [46:13] year or two however long the Facebook [46:15] guy was at the header everybody's like [46:17] who is the Facebook guy people think [46:19] it's alpacino people like it's student [46:22] do you know who the Facebook guy was I [46:24] totally do I looked this up I spent like [46:27] trying to figure this out really yeah [46:29] it's [46:30] amazing cuz everyone always said it's [46:32] alpino I'm like it doesn't really no [46:33] alpino must have looked really different [46:35] when he was younger and it's not all [46:37] right tell me your answer I'll tell you [46:39] if it matches mine all right so there [46:41] was an 80s song that got really big [46:44] called centerfold oh yes this is it and [46:47] the imag is of Peter Wolf yep who is the [46:50] lead singer of the J GES band we'll link [46:53] to the image there's this great Kora [46:54] post which we should say Adam D'Angelo [46:57] founder and CEO of Kora after Facebook [47:00] there's a great Kora post about this [47:02] where there's unmistakably the photo of [47:05] Peter Wolf that Facebook guy is based on [47:07] they just sort of pixelated it and you [47:09] know made it Duo tone instead of the [47:11] original photograph so great so great so [47:15] on the evening of February 4th 2004 [47:19] after final exams right when students [47:22] are coming back getting into the mode [47:25] for the new semester starting to shop [47:27] for spring semester [47:29] courses Mark Zuckerberg launches the [47:34] facebook.com and right off the bat from [47:36] the moment somebody logs in the [47:39] functionality is like awesome so the [47:43] registration page reads you can use the [47:46] Facebook to search for people at your [47:48] school find out who are in your classes [47:51] course match yep look up your friends [47:53] friends buddy Zoo see a visualization of [47:57] your entire social network buddy Zoo [48:00] plus graphics and then at the bottom of [48:03] every page is the famous copyright 2004 [48:06] the Facebook on Mark Zuckerberg [48:10] production awesome I feel like it's like [48:12] one of the most infamous screenshots is [48:14] the screenshot of that original homepage [48:16] so so great so importantly things that [48:19] did not exist here yet messages wall [48:23] posts even I don't think pokes were in [48:26] the original [48:27] very first version that's a good [48:29] question if they weren't in the very [48:31] first version they launched pretty [48:32] quickly thereafter yep but no Facebook [48:35] events no photos other than your profile [48:37] photo I mean no I don't even think [48:40] status updates were in the very first [48:42] one I don't think status updates were [48:45] there yet no but all the lessons from [48:48] all the other projects all the elements [48:49] of them are coming together here so once [48:52] again registration is limited to email [48:55] addresses on the heart [48:57] edu domain that was part of the magic of [49:00] what had made all the other projects so [49:03] successful is like it's not Rand in here [49:06] it's other people at Harvard it's your [49:08] classmates it's the people you care most [49:10] about yep and by the nature of that it's [49:12] their real identities and that's what [49:14] makes it different than all these other [49:15] social networks Myspace existed frster [49:19] existed but anyone could sign up for [49:21] these and in part that meant that they [49:23] had more explosive growth cuz there was [49:25] no governor on the growth and we'll talk [49:28] about all the problems that sort of come [49:30] from anyone anywhere can sign up at any [49:32] time but the very core thing here is [49:35] authenticated people the people who are [49:38] signing up for Facebook at the start are [49:41] people you know go to the same College [49:43] as you and have to use their real name [49:45] matched against a university issued [49:47] email address identity an authentic [49:50] identity is a part of the company from [49:52] its first moment yep coupled with super [49:56] learning number two from face Mash it's [49:59] all User submitted content so no [50:03] pre-population with anything from [50:05] Harvard you know no scraping from [50:06] Harvard servers Etc and like yeah on the [50:09] one hand that prevents mark from getting [50:10] into trouble on the other hand though it [50:13] sets the [50:14] expectation and the requirement and [50:17] trains the behaviors of the users of you [50:20] got to enter all the content into this [50:22] site yourself you put in your aim screen [50:25] name you put in your cell phone you put [50:26] in your interest you put in your classes [50:28] also isn't it crazy that they displayed [50:30] cell phone numbers for the longest time [50:33] right well again think about the [50:35] functionality and the utility to the [50:37] users on the one hand you would never [50:39] disclose your cell phone number to the [50:41] general public on the other hand other [50:44] students and your friends at Harvard [50:46] knowing your cell phone number well [50:48] that's actually pretty useful for [50:50] getting in touch for going out for going [50:52] to parties for planning things etc etc [50:54] these are all people you would give your [50:56] email and phone number and maybe even [50:58] birthday too if they just asked totally [51:01] you would probably also give them your [51:03] relationship status which is in [51:06] there formalizing your relationship [51:08] status for the first time here on the [51:10] facebook.com you can add a photo of your [51:13] own choosing as your profile picture so [51:16] big difference from Old School Facebooks [51:18] where it's whatever the university [51:20] chooses or it's a stock school photo or [51:22] whatever like no you get to express your [51:24] personality you get to photoshop it as [51:26] much as you want and you can update it [51:28] whenever you want and you can update it [51:29] whenever you want so it is amazing how [51:32] much of the next billion plus People's [51:37] Source of value comes from this founding [51:40] moment the Facebook grows tremendously [51:43] in functionality over time but [51:45] everything is like a natural outcropping [51:47] out of authentic identity User submitted [51:50] content trust that the people who are [51:53] seeing this content are people in your [51:55] sort of private Network the whole thing [51:57] is here in this first few weeks of [52:00] coding that got done and then threw up [52:01] the landing page I would even go one [52:04] step further than that the actual nature [52:07] of the seaing of the network here is [52:11] maybe the most important thing and [52:13] here's what I mean by that once Mark [52:16] launches this it takes off like wildfire [52:19] within Harvard of course he's already a [52:21] celebrity already known for launching [52:23] these projects people are tuned in [52:25] waiting for the next Mark Zucker [52:26] production within the first 24 hours 650 [52:30] people join there's only 2,000 people in [52:33] a class at Harvard so I think there's [52:35] about 7 or 8,000 undergraduates total [52:39] within 2 weeks over half of the entire [52:42] undergraduate population is active on [52:45] Facebook so now here's what I mean and [52:48] why the seating is so important this [52:51] initial network is super dense super [52:55] engaged and super acve that not only [52:58] sets the norms for how you're going to [53:00] behave on Facebook going [53:02] forward it also means that as the [53:04] network grows virally it radiates out [53:07] from this like nuclear reactor of a [53:11] base contrast that with an approach of [53:14] something like frster that's default [53:17] open to everybody backed by Silicon [53:19] Valley venture dollars trying to grow [53:21] they're doing marketing okay let's say [53:23] somebody in Ohio here about Friendster [53:27] somehow maybe they see an ad for it they [53:29] sign up when they join the service [53:32] there's no friends there there's nothing [53:34] to do right there's no functionality [53:36] it's not alive when you join the [53:39] Facebook even though it's limited to [53:41] just Harvard [53:43] students you sign up and you're like [53:45] Blown Away by how alive this site is [53:49] right almost nobody can use the Facebook [53:51] but for the people who can it's an [53:53] unbelievably great experience as they [53:56] expanded that stayed true it was either [53:59] you're not allowed to use it or if you [54:00] are it's Blow Away Great instantly [54:03] that's basically always true and the [54:06] moments when Facebook stops growing is [54:08] when that's not true and then they sort [54:10] of refocus on it and then that gets the [54:13] growth back yep here's another thing [54:15] that I think was really important about [54:17] this seating of Facebook was actually [54:20] that it happened at Harvard Harvard is [54:24] the top Global brand [54:26] among universities whatever you [54:28] personally think about Harvard it's [54:30] undeniable especially back then that it [54:33] was like the top if you were a kid or a [54:36] family pretty much anywhere in the world [54:39] you knew what Harvard was this was [54:41] really important as Facebook expanded [54:43] because it had this patina to it it was [54:45] like Elite especially as they grew there [54:48] were clones there were lots of other [54:50] Facebook copies out there and even if [54:53] they were as good as what markk had [54:55] built they didn't start at Harvard it's [54:58] kind of like Civilization nobody else [55:00] could ever invade Facebook's Turf but [55:03] they could invade other people's Turf so [55:06] somebody put it to me in the research [55:07] like there was a fairly big competitor [55:10] in Germany I think maybe called steady [55:12] Vice or something like that Facebook [55:15] could really easily go into Germany and [55:17] people in Germany would be like oh [55:19] Facebook I've heard of that this is [55:21] interesting study Vice could never go [55:23] into Boston like ever hm [55:27] it's interesting yeah that's a great [55:28] Point okay so like weeks go by and [55:33] they've saturated Harvard and not only [55:35] that you start seeing this thing happen [55:37] which just to level set with listeners [55:40] this stayed true for several years this [55:42] is a crazy stat 70% of people who ever [55:46] signed up were active that day there are [55:51] not user engagement retention metrics [55:54] better than that in the world the fact [55:56] that as they grew it didn't matter how [55:58] many more people they added it was still [56:01] the case that 70% of users ever were [56:03] daily active and so you have this crazy [56:06] situation at Harvard where it's just [56:08] like to your point nuclear reactor [56:10] levels engaging in fact this term [56:13] wouldn't get coined till later in the [56:15] summer but this is from the Facebook [56:16] effect David Kirkpatrick's book they had [56:18] coined a term for how students seem to [56:21] use a site Zuckerberg moscovitz and [56:24] Parker Sean Parker who we'll talk about [56:25] who comes in called it the trance once [56:28] you start combing through the Facebook [56:30] it was very easy to keep going it was [56:32] hypnotic you just kept clicking and [56:35] clicking and clicking from profile to [56:36] profile viewing the [56:38] data and every single one of those [56:41] clicks is a page view so yeah this hits [56:46] Harvard Like An Earthquake it's face [56:50] mask all over again except a good [56:52] problem not a bad problem it's all legit [56:55] the servers are melting down even Mark [56:58] is blown away by how big this thing gets [57:00] so quickly yep right away they add the [57:02] wall so there's a way to publicly post [57:05] uh things like testimonials on your [57:07] friends walls uh you could view a wall [57:10] to wall so you could see that public [57:11] post that you are making back and forth [57:13] to each other so that you've got your [57:15] profile information wall posts and [57:18] that's it yeah but to your point very [57:20] quickly Mark is adding features adding [57:22] functionality adding things to do to the [57:24] site the other thing that he does right [57:27] off the bat is say okay we're taking [57:30] this to other colleges too so he [57:32] recruits his roommates Dustin moskovitz [57:35] comes on to help maintain the site [57:37] Dustin was actually an econ major he had [57:39] no idea how to code he goes out and buys [57:43] Pearl for dummies and Mark's like uh [57:46] that's great that you read that the site [57:47] is written in [57:50] PHP so Dustin goes and learns php2 and [57:54] helps with coding the site and keeping [57:56] up with everything H you know Pearl you [57:57] can learn PHP it's much easier Chris [57:59] Hughes joins too Andrew McCullum joins [58:03] and then they need to keep buying [58:05] servers to keep up with the traffic or [58:07] else everything is going to melt down so [58:09] one of Mark's fraternity Brothers in aai [58:11] is a guy named Eduardo Savin and Eduardo [58:14] was I think part of the business and [58:16] investing Club at Harvard so they [58:18] recruit Eduardo in and Eduardo invests [58:21] $1,000 in infrastructure for the site [58:24] Mark invests another ,000 so together [58:26] there's $2,000 of investment in the [58:29] facebook.com and the deal they strike is [58:31] that Mark gets two-third of the company [58:33] and Eduardo gets onethird of the company [58:35] and and wardo sets up a business entity [58:38] to formalize all of this as a Florida [58:41] LLC Ed wardo was from Florida right away [58:44] as we were saying Mark's like great this [58:46] is not just a Harvard thing we need to [58:47] bring it to other colleges and he [58:49] specifically chooses colia because Ben [58:52] as you said there was an existing [58:54] competitor there and he's like I want to [58:56] see how Facebook competes and like I [58:59] want to know right away can we displace [59:02] it it's so interesting he doesn't go to [59:04] the schools where there's nothing he [59:05] goes to the schools where there is [59:06] something and wants to be better than it [59:09] which is so counterintuitive to how most [59:11] people would think they say like oh no [59:12] this is a race I'm going to go to the [59:14] white space I'm going to gobble [59:15] everything up and then wait to take on [59:17] competitors when I'm bigger and stronger [59:19] like Mark's like nope right away this is [59:21] a early mindset of we are going to be [59:24] globally dominant if you're okay [59:26] splitting the market you go for the [59:27] white space and then you say let's see [59:29] how much of the low hanging fruit we can [59:30] easily get but if you're Facebook you're [59:33] going to say let's go to the hardest [59:35] possible competitor extinguish them yes [59:39] and then and then we uh we'll have a [59:42] better shot at owning the whole market [59:43] and we can get to the low hanging fruit [59:44] later by the way their strategy once [59:47] they did win at a school like Colombia [59:49] was to go figure out all the schools [59:51] that were like closest to Columbia in [59:55] terms of network connection and then win [59:57] there too to kind of like build a around [59:59] their Victory yep so there we have the [60:02] initial crew Mark Dustin Chris Andrew [60:07] and Eduardo the five co-founders of [60:09] Facebook the other interesting thing [60:11] about going to Colombia which they do [60:15] three weeks after the launch at Harvard [60:17] 22 days it was the same month all within [60:21] February [60:22] 2004 they set up colia as a completely [60:27] different network than Harvard so if you [60:31] join the columia Facebook you can't then [60:34] go access all the profiles of Harvard [60:37] students right at first there was Zero [60:39] connectivity between the two which is [60:41] another sort of counterintuitive thing [60:43] you'd think like oh great I want to [60:45] build this as big as possible Mark [60:46] Clearly wants to conquer all these other [60:48] schools the best way to do that is to [60:51] have all this alive content come right [60:53] to them from Harvard I think he realized [60:56] that had he done that it would have [60:58] socially inhibited sharing within [61:00] Harvard yeah the scarce commodity is [61:03] trust that is the important lesson here [61:06] it's also a very convenient [61:10] infrastructure decision where if these [61:13] two systems truly don't need to talk to [61:15] each other that's great put different [61:17] servers in different data centers don't [61:19] worry about overloading the database [61:21] don't worry about number of wrs per [61:23] second and reads per it's great if every [61:25] school gets its own server rack you end [61:28] up not having some of these issues that [61:30] the other social networks have frienders [61:32] out there trying to compute second [61:34] degree friends of friends which is this [61:37] crazy hard computer science problem [61:39] especially as the global number of [61:41] people growing it's an N squared problem [61:44] and Facebook is over here or the [61:45] Facebook at the time in the land where [61:48] they're keeping n small so even anything [61:50] that's N squared is contained within [61:53] that school and their servers aren't [61:55] falling over and taking 20 plus seconds [61:57] to load pages the way that the frienders [62:00] of the world are this is such an [62:02] interesting point because technology [62:05] infrastructure Advantage has always been [62:08] a core part of Facebook even back to [62:11] this very beginning what's interesting [62:13] is that it flips at some point as [62:17] Facebook grows and eventually becomes [62:18] open to [62:20] everybody their competitive Advantage [62:22] becomes their Tech is so good and their [62:25] infra is so good that they can do [62:27] everything you were just talking about [62:28] that Friendster couldn't do and that [62:30] Myspace couldn't do but in these early [62:32] days their competitive Advantage is [62:34] actually like no we're scale out not [62:37] scale up and then Mark was totally [62:39] willing to just flip the bit later right [62:41] it's effectively counterposing they're [62:43] building the Facebook in such a way that [62:44] the use cases don't require incredibly [62:47] sophisticated technology to accomplish [62:50] those feature sets there's this funny [62:51] chicken of the egg thing here and I [62:53] think the answer like many chicken of [62:54] the egg is both Mark at the moment he [62:57] conceives of an idea thinks through the [63:01] technical requirements and the user [63:04] experience and they're sort of [63:06] co-mingled in the product development [63:09] process and so I really do think even at [63:11] age 19 he was sort of aware of the [63:14] scaling benefits in addition to the user [63:16] experience benefits of launching in this [63:18] way totally agree okay fun bit of trivia [63:22] about Harvard's leave policy that we [63:24] referenced earlier do you know what else [63:28] Mark does on the same evening that they [63:31] launched Colombia no idea he goes to [63:35] hear Bill Gates speak on campus that was [63:38] the same night the same night during [63:42] which speech Bill mentions that he [63:47] actually felt comfortable dropping out [63:49] of Harvard because he discovered that [63:51] Harvard had this really generous leave [63:53] policy that you could take an infinite [63:55] of absence and pursue something else and [63:58] come back anytime you want and Mark [64:00] would actually say later that bill [64:02] saying that at the event was part of [64:04] what got him comfortable doing the same [64:06] thing unbelievable it is crazy tying [64:09] this back to our Microsoft episode this [64:11] would have been what early 2004 and so [64:14] this would have been right after Bill [64:16] stopped being CEO handed the CEO Reigns [64:19] to Steve he was still the Chief [64:22] Architect and still chairman of the [64:24] board but it's sort of this like post [64:26] doj time for Microsoft where Bill is [64:28] just technically focused and can do [64:30] things like go speak at Harvard yeah it [64:33] is [64:35] amazing how much influence Microsoft had [64:38] on Facebook and on Mark we have a lot [64:41] more to talk about on that front so next [64:44] couple days after Columbia launches [64:45] Facebook launches at Stamford at Yale by [64:48] the end of the semester they're in over [64:50] a hundred schools in like you know what [64:52] three four months like the speed they're [64:54] moving with is crazy and they start [64:56] building these weit lists I mean this is [64:58] the other there's so many common startup [65:01] things that Facebook kind of invented [65:04] they built tremendous demand before they [65:08] would light up the network they knew [65:09] that as soon as you're in Facebook you [65:11] want to quickly get to Seven friends or [65:14] 10 friends or whatever the metric is to [65:15] create that sort of Magic Moment where [65:17] you're like oh yeah Facebook now works [65:18] for me so they wanted to wait until they [65:21] had sufficient demand to Boom open that [65:23] school and then once it's open everybody [65:26] should have the best possible experience [65:27] and so you're seeing Facebook basically [65:30] say Okay wherever there's really strong [65:32] demand that's where we'll open next and [65:34] we're not going to open anywhere where [65:35] we see like you know middling demand for [65:38] our product yep totally meanwhile here [65:42] we are at the end of the school year the [65:43] Winkle us twins take the whole connect [65:45] you situation to Harvard's president [65:47] Larry Summers to adjudicate this dispute [65:50] between them and Mark Larry says Hey [65:53] Harvard is not going to get involved [65:55] here this is a a business dispute [65:56] between students eventually the two [66:00] twins Sue Mark and Facebook they settle [66:03] for $65 million 45 million of which is [66:07] paid in pre-ipo Facebook stock obviously [66:11] that grows a large portion of that then [66:14] I think they sell and put into Bitcoin [66:16] in like 2011 2012 time frame so they end [66:21] up pretty pretty fine out of all this [66:24] they did great they did great [66:26] but here we are at the end of the school [66:27] year summer comes around and Andrew [66:31] McCullum has some connections out in the [66:33] Bay Area he had interned at Electronic [66:35] Arts Adam D'Angelo is going to be [66:37] spending the summer from Caltech so the [66:40] crew all decides like hey great let's [66:42] not get internships let's go rent a [66:45] house out in Palo Alto move to Silicon [66:48] Valley and work on Facebook there for [66:49] the summer and see what happens and yes [66:53] the house did indeed have a pool [66:56] yes they did indeed install a zipline [66:58] off the [66:59] chimney that was not fiction in the [67:02] movie that did actually happen and it is [67:04] important to know to this point it is a [67:08] project Mark and Dustin and Chris and [67:10] everyone who moves out they are there to [67:13] work on a project that they think is [67:15] cool and seems to be working and get [67:17] exposure to Silicon Valley and Venture [67:20] capitalists for when they start their [67:23] startup having a network that is [67:26] literally the mentality Facebook is live [67:28] at how many schools David 100 100 [67:31] Schools and they are going out not with [67:34] the intent to make this a company but to [67:37] contemplate what company they could [67:38] start and meet people that can help them [67:40] with that yep now when we say crew who [67:43] moves out that crew did not include [67:46] Eduardo Eduardo had an internship in New [67:48] York and decided that he was not going [67:50] to move out for the summer yeah tough [67:53] decision not a great decision on the [67:56] other hand he ends up with 2% of [67:57] Facebook so he's also fine yeah so to [68:00] make a long story short yes everything [68:02] you saw in the social network around [68:04] this the result is effectively correct [68:07] that Eduardo Savin goes from owning a [68:09] third of this Florida based LLC to [68:11] something like 2% of this C Corp that is [68:15] a Delaware SE Corp based in California [68:17] that goes on to become or is Facebook [68:20] Inc and the justification that they [68:23] effectively use in changing this [68:25] structure is hey a bunch of us moved to [68:28] California to start a company together [68:30] you stayed back and yes you sold some [68:32] ads in the meantime but like you didn't [68:34] come start the company with us listener [68:36] we leave it to you to sort of decide how [68:38] that should have played out and what's [68:40] fair none of us were there he was a [68:42] co-founder of Facebook and then ended up [68:45] with 2% and lived on a different Coast [68:47] okay so they're in California they get [68:49] really serious about the Facebook they [68:52] actually are still working on something [68:53] else called wirehog concurrent L within [68:56] the same team but they're starting to [68:58] realize okay this the Facebook thing [69:00] really really has legs and as they're [69:02] sort of contemplating their next move [69:04] they literally run into someone on the [69:07] street who will change everything David [69:09] who is this person that person is the [69:12] co-founder of Napster Shawn Parker who [69:15] would have a brief but very very large [69:19] impact on the company yes so before we [69:22] tell the Shawn Parker chapter of the [69:24] Facebook [69:25] now is a great time to talk about our [69:27] presenting partner JP Morgan payments in [69:30] these critical moments like we are [69:31] talking about now where a new paradigm [69:33] or technology shift happens David and I [69:35] typically focus on the founders and the [69:38] innovators who figure out how to build [69:40] something really great but unfortunately [69:42] it's not just the good actors who take [69:44] advantage of these opportunities oh yeah [69:47] there are always Bad actors on the other [69:49] side of the same new technology and this [69:51] is especially true whenever money or [69:54] value is exchanged whether it be phone [69:56] scammers taking advantage of [69:57] international networks in the 80s and [69:59] 90s viruses spreading through the early [70:01] internet or today's cyber crime like AI [70:04] powered identity theft as technology [70:06] gets smarter so will the fraud tactics [70:08] enter JP Morgan payments with the [70:10] digital payments ecosystem expanding [70:12] rapidly through Innovations like [70:14] real-time payments blockchain and [70:15] peer-to-peer networks the attack surface [70:17] has grown exponentially for a company [70:19] like JP Morgan who moves1 trillion doll [70:22] a day across 160 countries staying ahead [70:24] of Bad actors is critical for the entire [70:26] Global Financial system that's why [70:28] they've been investing in technology [70:30] like Ai and cyber fraud prevention for [70:32] years their AI models have been trained [70:34] on billions of transactions that is [70:36] unparalleled scale and expertise when it [70:38] comes to fraud intelligence and it's [70:40] only getting smarter and more reliable [70:42] given they're processing over 5,000 [70:44] transactions per second with [70:46] 99.9% accuracy regardless of your size [70:50] or industry you can leverage their [70:51] Enterprise grade Solutions like [70:53] validation Services which validates Bank [70:55] accounts individuals and businesses one [70:57] great example is a Fortune 500 Global [70:59] Media company they needed an automated [71:01] solution for the thousands of [71:03] influencers on their platform to do [71:04] rapid onboarding minimize errors and [71:07] ensure consistent validation of bank [71:09] account information they were able to [71:10] compress validation times from days to [71:12] minutes so they could scale they've got [71:14] a great payments developer portal it's a [71:16] trusted scalable platform built with [71:19] payment security and risk management in [71:21] mind developers can even integrate their [71:23] preventative fraud solutions to act [71:25] before an incident takes place sardine a [71:28] leader in financial crime prevention [71:29] collaborates with JP Morgan payments and [71:31] embeds its account validation Services [71:33] API to analyze and protect over two [71:36] billion devices for 300 companies with [71:38] AI to the highest standards of Fraud and [71:40] compliance safety David I actually met [71:42] the founder of Sardine at the Meetup [71:45] that we did after the chase Center event [71:46] with JP Morgan payments in San Francisco [71:48] last month that's right was awesome [71:50] ultimately every business benefits from [71:52] built-in fraud prevention whether you're [71:54] moving 100 $100 or a billion dollarss [71:56] and with the trust and Innovation from [71:58] JP Morgan payments whole ecosystem you [72:00] can drive growth while protecting your [72:02] business check out more Payment [72:05] Solutions and stories at jpmorgan.com [72:08] acquired okay David Shan Parker enters [72:11] the picture here we go so Shawn Parker [72:14] had of course started Napster with Shan [72:17] Fanning back in the late 1990s and after [72:20] the whole Napster Saga Parker ended up [72:24] starting the e contact list company [72:27] called plaxo which was kind of its own [72:30] Proto social network and pioneered email [72:34] address book exploitation and exporting [72:38] which is how every social company got [72:40] its start Facebook did a ton of this [72:42] LinkedIn did a ton of this WhatsApp did [72:44] this exactly it pioneered the first [72:47] bootstrap viral social network mechanic [72:50] hey let me see who all the people you [72:52] know are so that on this new thing that [72:54] you want to be connected to the people [72:55] you know you're connected to all the [72:57] people you know oh by the way we might [72:59] also invite all the people that uh [73:01] you've ever contacted yep so Shawn and [73:05] his co-founders at plaxo raise money for [73:08] plaxo from venture capitalist and not [73:11] just any venture capitalist the best [73:13] venture capitalist from Sequoia capital [73:17] and all is going well plao is going fine [73:20] Sir Michael Moritz is on the board of [73:22] the [73:22] company anyway after a couple years [73:26] Shawn Parker ends up getting kicked out [73:29] of the company by the board now Shawn [73:33] wasn't exactly the most like reliable [73:35] employee and uh definitely was um quirky [73:40] but the result of this is Shan's [73:42] Everlasting emnity for seoa and paranoid [73:47] distrust of all venture capitalists and [73:50] Company boards this is going to become [73:52] very very very important here in a sec [73:56] so rewind slightly back to the spring of [73:59] 2004 when the team is still at Harvard [74:01] Shawn is living in Palo Alto and one of [74:05] his housemates girlfriends who's I [74:08] believe a student at Stanford pulls up [74:11] the facebook.com on her laptop and it's [74:15] kind of like the scene where this [74:16] happens in The Social Network movie sha [74:19] is like holy crap this is it this is the [74:23] winning social Network and Sean I [74:26] believe had actually been an adviser to [74:28] Friendster as well so he was like [74:29] between plao and Napster Friendster [74:34] really belied social networking was [74:35] going to become a thing totally and it [74:37] is kind of crazy to think about unless [74:39] you had a student show you on their [74:42] computer this is what Facebook is [74:43] there's no real way for you to know [74:45] other than the signup page right which [74:47] you can't get past it and see how alive [74:50] the thing is yep so Sean being the [74:53] enterprising person that he is he cold [74:56] emails the company's business email [74:58] address and asks to set up a meeting and [75:01] Eduardo Savin replies to Sean is like oh [75:04] cool you're the co-founder of Napster [75:06] yeah they set up a dinner in New York [75:09] City this is while the spring semester [75:11] is still going on at Harvard and Mark [75:14] and Eduardo and Shawn all get dinner [75:17] together and Mark and Shawn really Bond [75:19] Marcus super pumped to meet him but [75:23] that's about it like there's no you know [75:25] discussion about you know what's cool a [75:27] billion dollars you know etc etc as [75:29] dramatized in the movie that doesn't [75:31] happen they kind of think that they're [75:33] just G to go their separate ways right [75:35] nice to meet you I don't think even Mark [75:38] knows yet that he's moving to California [75:39] this summer no totally not so here we [75:42] are now in June the Facebook crew has [75:45] moved out to paloalto they're walking [75:48] down the street one evening Shawn is [75:50] also walking down the same street in [75:52] Palo Alto sees them [75:55] and is like hey it's the Facebook [76:01] guys and at this point Shawn Parker was [76:04] living in a different house [76:07] specifically the house of his [76:09] girlfriend's parents not a good [76:11] situation and he makes a split-second [76:13] decision right there on the street to [76:16] insert himself into Facebook and is like [76:20] hey I'm actually looking for a place to [76:22] stay do you think I can crash with you [76:23] guys in the crash pad for the [76:26] summer and he joins the crew and this is [76:30] how it all starts that is wild yeah it's [76:34] like uh you ever play old school like [76:37] Super Nintendo role playing games like [76:40] Japanese RPGs like Final Fantasy or [76:42] whatnot uh it's been a while I mean it's [76:45] just like you meet somebody on the [76:47] street and they join the party and like [76:49] you know you always know the character [76:51] on the street that's going to join the [76:52] party because their pixel Sprite is more [76:54] highly detailed than the regular NPC pix [76:57] I mean this is exactly what it is like [76:59] here Shawn Parker had a very detailed [77:01] pixel Sprite yes so while he's there you [77:04] know living in the house uh crashing [77:06] with the college students he basically [77:08] takes Mark aside and he's like look [77:11] you've got something really special here [77:14] I have been in your shoes let me tell [77:16] you exactly what is going to happen this [77:19] summer you have created something [77:21] magical you're now out here in Silicon [77:23] Valley you are going to be the bell of [77:26] the ball all the Venture capitalists out [77:29] here all the big companies they're going [77:31] to whine and dine you they're going to [77:33] court you they're going to tell you [77:34] You're great and they're going to want [77:36] to invest in your company and then [77:39] they're going to turn around and they [77:41] are going to screw you they're going to [77:43] take control of your board they're going [77:45] to force you to bring in quote unquote [77:47] Professional Management they're going to [77:49] kill all the magic growth is going to [77:51] slow then they're going to blame you and [77:53] then they're going to fire you and then [77:54] they're going to put the company up for [77:55] sale and here's the thing to this point [77:58] in history he really wasn't wrong he was [78:01] absolutely right there wasn't Founders [78:03] fund yet there wasn't a16z yet there [78:06] wasn't like the notion of founder [78:08] friendliness what Venture capitalists [78:10] did is they invested in Founders [78:12] companies and then brought in management [78:16] to take them to the next level it's what [78:18] happened at eBay it's what happened at [78:20] Google this is a pattern that happened [78:22] over and over and over again and usually [78:24] they would try to do it with the sort of [78:26] help and support of the entrepreneur to [78:27] build this like really great team [78:29] together but often times that backfired [78:30] and they just hot swap the CEO with a [78:32] big grown-up manager and a nice suit yep [78:36] and that was not the wrong Playbook [78:42] until this moment in time because as [78:44] we've been talking about until now if [78:47] you wanted to scale something you needed [78:49] money you needed permission you needed [78:51] business people you needed deals you [78:53] needed contact [78:55] yeah that's true the open web kind of [78:57] changed the hard requirement on that yes [79:01] and I think Shan as bitter as he was [79:05] about his own experience I think he [79:07] probably also recognized that and [79:10] recognized that the Facebook and Mark [79:13] had the potential to start a new [79:15] paradigm of how to do this differently [79:18] so Sean says to mark look if you just go [79:20] out on your own that's what's going to [79:21] happen I am going to take you under my [79:24] wing and make sure that that does not [79:26] happen and I'm going to help you reset [79:29] up the company so that you permanently [79:31] control the board and no one can ever [79:34] fire you because you are what is [79:36] important here and I'm going to help you [79:40] go about fundraising but I'm going to [79:42] make sure we find the right people who [79:45] are cool with that stipulation number [79:47] one because if you go try and do it by [79:49] yourself like you don't know enough you [79:50] don't know these people you're not going [79:52] to find the right folks h it's kind of [79:55] wild I mean Shawn was also in a position [79:59] to take advantage of Mark and the [80:01] company and he didn't there was some [80:04] nice self-interest in all of this it's [80:06] not like this was like charitable he [80:07] ends up being the president of Facebook [80:09] anded he ends up with a pretty decent [80:11] chunk for a non-founder and he ends up [80:13] with a board seat now granted it's [80:15] technically Mark's board seat but you [80:17] know Shawn's the one sitting in it [80:19] totally but it would have been really [80:21] easy for him to do one of two things you [80:23] know one obviously take advantage of [80:25] marking the company and get a lot more [80:27] for [80:28] himself two I think he also could have [80:33] been so Jaded by the plao experience to [80:36] say we shouldn't raise Venture Capital [80:38] at all but actually what he counsels [80:40] Mark to do is the right thing no we [80:43] should raise Venture Capital we should [80:46] professionalize and the whole Silicon [80:48] Valley [80:49] Machinery really can help us but we're [80:53] going to do it in the right way on our [80:54] terms yep I think he just really wanted [80:57] to help Mark win the game I think you're [81:00] right the other thing that Shawn does by [81:02] all accounts is he makes sure that [81:05] everyone in the house that summer gets [81:07] equity in the company okay so what's [81:09] wirehog what's what's the second project [81:11] that's going on here yes I really would [81:14] love to know what Shawn thought when [81:16] Mark and the crew told them about [81:18] wirehog wirehog was basically going to [81:22] be a captive nap [81:25] to Facebook everybody on Facebook in [81:28] these private networks W Gardens what is [81:32] something that college students really [81:34] really wanted to do and was happening [81:36] internally on College servers all the [81:38] time share files share music share [81:42] movies wirehog was going to be a product [81:44] to make this happen and actually Legend [81:47] has it I think in Steven Levy's book [81:50] while they're talking about this sea [81:52] Parker actually comes up with a name for [81:53] the product uh it's like oh if you [81:56] launch this you should actually call it [81:58] Dropbox no way yeah like a year before [82:03] Dropbox the actual company gets founded [82:05] that's so funny but they really are sort [82:07] of thinking of themselves as this almost [82:08] like incubator lab I don't really [82:10] understand how this is true but they [82:12] really did convince themselves that [82:13] their current front runner for the [82:15] product they were most excited about was [82:17] wire hog and Facebook might be like a [82:19] good distribution vehicle for it or [82:21] maybe at some point they don't even [82:23] focus on Facebook anymore and they go [82:24] all in on wirehog and they're actively [82:26] talking about this Insanity while [82:30] speaking with investors about raising [82:33] Capital yes so to the investors that [82:37] Sean goes out and helps them find Sean [82:40] at this point knows basically everybody [82:42] in Silicon Valley and he's like okay I [82:45] need the pretty tight window of people [82:47] who a have money B know what they're [82:50] doing and C are okay with our terms that [82:52] Mark is going to control the board and [82:54] control the company so we need an [82:55] individual not an institution so he [82:58] calls two angel investors who he knows [83:02] pretty well and he thinks might be the [83:03] right fit one is Reed Hoffman who Reed [83:07] had been running LinkedIn and Shawn had [83:10] been in this early social networking [83:12] thing and he just started LinkedIn they [83:14] actually knew each other because Reed [83:16] was an early investor in Friendster so [83:18] Sean calls up Reed and he also calls up [83:21] Mark pinkis Mark had been Shawn's boss [83:25] at an internship that he had in high [83:27] school and then Mark had gone on to [83:29] invest in Napster when they started [83:31] Napster as a company and we should say [83:33] who Mark Pinkus is for anyone who [83:34] doesn't know Mark would eventually start [83:36] Zinga yes could become a very important [83:39] person and Company in the Facebook [83:41] ecosystem the maker of Farmville yes so [83:45] Pinkus and Reed meet up with Shawn and [83:48] Mark and they're just blown away by the [83:51] Facebook and then as you say Mark starts [83:53] talking about wire hog and apparently [83:55] Reed is like no no no no stop talking [83:57] about this wire hog thing like I'm [84:00] running LinkedIn I can tell you this is [84:02] special like do the Facebook thing not [84:05] only do the Facebook thing but like [84:08] music is Kryptonite as a space right now [84:11] because of what Napster did to the music [84:13] labels how is Shawn not the one being [84:15] like stay a thousand yards away from [84:18] music my God I mean amazingly a couple [84:21] years later you know Shawn being so [84:23] open-minded I mean he the one of the [84:25] ones he really helped Spotify get going [84:27] yep and really brokered the relationship [84:29] with Mark for Daniel and Mark to hit it [84:31] off and obviously Spotify to have a huge [84:33] amount of growth on Facebook totally so [84:36] okay Reed really excited he's like no no [84:39] the Facebook focus on the Facebook your [84:41] user numbers are what your engagement is [84:43] what your retention is what your Dow Mau [84:45] ratio is what oh my God so it reads like [84:48] look I'm running [84:50] LinkedIn Pinkus and I don't have that [84:52] much money together we'll give you some [84:54] we can't like fund a whole you know [84:56] round for you and they're kind of [84:58] conflicted both with frster and Linkedin [85:01] but he's like I think I know the perfect [85:03] person who can put this together I [85:06] worked with this guy at PayPal my old [85:09] colleague from PayPal who actually as [85:12] luck would have it is thinking about [85:15] starting his own Venture Capital fund [85:17] this could be a really good fit why [85:20] don't you meet Peter [85:23] teal and Peter yes indeed was just in [85:26] the process of starting what would [85:27] become Founders fund Ben you already [85:30] alluded to this but like Founders fund [85:34] was such a radical idea when Peter [85:38] started it the idea being baked into the [85:41] name that they will always side with [85:42] Founders and never push a Founder out of [85:44] a company I was shocked I mean it just [85:48] it's so easy to forget this but it was [85:50] just 20 years ago the idea that a [85:51] venture capitalist makes an investment [85:54] in a company and they are not the [85:56] controlling shareholder was Blasphemous [85:59] just 20 years ago totally Blasphemous [86:02] and of course PayPal has its own crazy [86:04] history where you can understand why [86:06] Peter would arrive at this conclusion [86:09] but he really was like maybe the only [86:12] person in Silicon Valley that hit that [86:14] VIN diagram that I talked about a minute [86:17] ago of like super legitimate would [86:19] actually help the company and help them [86:21] navigate Silicon Valley had money [86:25] because the PayPal exit had just [86:27] happened and was going to be cool with [86:30] Mark controlling the company yeah it's a [86:33] oneof one intro to make I mean it it [86:36] would have had to be someone else in the [86:37] PayPal Mafia that kind of fits that yep [86:40] but I don't think at that time you know [86:42] Elon wasn't doing investing I don't [86:44] think there was anybody else who really [86:45] could have done it yeah that's a crazy [86:48] counterfactual thinking about if Elon [86:50] had invested in Facebook instead of [86:52] Peter teal right what if Elon hadn't [86:54] started SpaceX he was in just as good a [86:57] position totally had ruoff gone to seoa [87:00] yet if he had he would have just started [87:02] this is effectively the pool of [87:04] financiers who could have done this deal [87:06] right but Sequoya wasn't going to do [87:08] this deal under those terms so ruoff [87:10] couldn't have done it yep it was maybe [87:12] one of two with the other one [87:15] potentially being Elon wow that would be [87:17] a really different world that we would [87:19] be living in today [87:21] y Okay so take the idea to Peter in the [87:27] meeting Reed brings along a young guy [87:29] who's working for him at LinkedIn named [87:31] Matt [87:32] ker and to come help you know talk to [87:35] Peter evaluate this from an investing [87:37] standpoint they strike a deal that [87:40] Peter's gonna lead a round of [87:43] $500,000 Reed and Mark pinkis are each [87:46] gonna invest [87:48] 37,500 Peter's going to do the rest at a [87:51] $5 million pre-money valuation so 5.5 [87:55] million [87:56] post which you know on the one hand is [87:58] like laughable now on the other hand at [88:02] this point in time for a bunch of kids [88:05] who started this a couple months ago [88:07] it's crazy that this would be worth $5 [88:09] million out of $5 million valuation the [88:13] multiple since then has been about [88:16] 250,000 [88:19] X so that takes us to the end of the [88:22] summer [88:24] so everybody now has to make a decision [88:28] do they stay in Palo Alto and keep [88:31] working full-time on Facebook or do they [88:34] go back to school so Eduardo never came [88:38] out in the first place so he did his [88:40] internship in New York he goes back to [88:41] Harvard and starts his junior year Chris [88:44] Hughes also goes back to Harvard starts [88:48] his junior year graduates in 2006 and [88:51] then rejoins Facebook after graduation [88:54] Andrew McCullum stays in Palo Alto for a [88:58] couple years and then goes back to [89:00] Harvard later and completes his degree [89:02] he takes the uh you know Bill Gates Road [89:04] not traveled option Dustin never goes [89:08] back stays with the company until 2008 [89:10] when he leaves to start AA and Mark I [89:14] think probably never really even [89:17] considered going back yeah at this point [89:19] the 70% stat keeps being true despite [89:22] the fact that they keep opening all [89:23] these new schools [89:24] it keeps being true that 70% of people [89:26] who have ever signed up are daily active [89:28] users so I think Mark realizes oh I have [89:31] created one of the most engaging [89:33] technology applications ever yes and [89:37] certainly at this point the combination [89:38] of [89:40] sha Peter Reed Mark [89:45] pinkis all advising and being part of [89:48] the company they know and they have [89:52] certainly helped him realize that yam [89:55] dangelo does go back to Caltech and then [89:58] through aim I guess remotely helps [90:00] collaborate with theany until he [90:02] graduates and in 2006 and becomes CTO [90:05] officially got it Shawn like we said his [90:09] impact on the company was huge I really [90:11] want to underscore this again Mark [90:15] having full control over the company as [90:17] we are going to see time and time and [90:19] time again as we go through the story [90:20] here makes all the difference in the [90:22] world and that is 100% because of Shawn [90:25] Parker yep so how long is Shawn at the [90:28] company not very long because at some [90:32] point in time the next spring so this [90:34] like nine total months at the company [90:36] yeah I think about nine total months [90:38] Shawn ends up leaving the company after [90:41] charges get filed against him as a [90:43] result of some drug incidents at a house [90:45] party during a trip to North Carolina [90:49] charges do eventually get dropped but in [90:51] the meantime the decision is made hey [90:54] this is not something the company's [90:55] goingon to get sucked into yep he [90:58] immediately joins where else found f [91:03] which Peter partially on the back of [91:05] this Facebook investment has now set up [91:07] and raised officially and like we said [91:10] sea goes on to find and fund and really [91:13] help nurture Spotify through his role of [91:16] Founders fund yep it's kind of amazing [91:18] thinking about Founders Fund Raising [91:20] Capital off on the back of this [91:21] investment one pitches we invested in [91:24] Facebook the flip side of it is we just [91:27] invested in a company we're currently [91:29] holding it flat cuz there's been no [91:31] markup yet from any other investors and [91:33] they're not generating any [91:36] Revenue well let's talk about uh other [91:39] investors by this point kind of towards [91:42] the end of spring semester 2005 Facebook [91:46] had now dropped thethe become just [91:49] Facebook that was one of Shawn Parker's [91:53] final imps on the company is he actually [91:55] negotiated the domain name purchase of [91:58] facebook.com I the company yeah I [92:00] believe they bought it for [92:02] $200,000 hm which actually was a good [92:05] chunk of that Venture money that they [92:06] raised right because it was 500k that [92:09] they raised before it was 500k yep and [92:11] they're actively spending that the [92:13] company has never consumed that much [92:15] Capital but they are having to spend it [92:17] on servers as they stand up new colleges [92:19] they are having to go into Data Centers [92:21] and rack servers or rent on a monthly [92:24] basis and the server bills are starting [92:27] to add up especially as they scaled to [92:29] how many schools were they at in the [92:31] fall of 2004 several hundred well Shawn [92:35] Parker introduced Mark to Western [92:38] technology investment WTI the debt fund [92:41] yeah yeah and so they basically invented [92:44] Venture debt and so they make a [92:48] $300,000 loan it's like I think like a [92:50] revolving credit line in the fall of [92:53] 2004 [92:54] and that comes with warrants that if the [92:57] company ever goes public that they can [92:59] then exercise they do a deal again the [93:02] next spring for another $300,000 so [93:04] they're in for [93:06] $600,000 I uh spoke with someone years [93:08] ago who told me this ended up being [93:11] phenomenally phenomenally successful and [93:13] the warrant exercise on this I think is [93:15] the greatest Venture debt deal of all [93:18] time has to be nice spring 2005 Excel [93:23] comes in leads of series a $12.7 million [93:28] at a $98 million post money valuation [93:32] that's sort of like crazy in and of [93:34] itself for the time and though still [93:37] being okay with Mark having board [93:40] control yes which you know one thing for [93:43] like okay at the angel round maybe like [93:45] that's still pretty crazy but whatever [93:47] this is Angel round but like no like a [93:49] real Venture Capital firm real series a [93:52] 2005 [93:54] five a100 million valuation B being okay [93:58] with that that was wild oh also C [94:01] getting less than 20% of the company [94:03] like rule of thumb Venture investing [94:05] back in the day was you need at least [94:07] 20% of the company ideally more like 25 [94:09] or 30 yes so Kevin ausy who ran down the [94:13] deal and did the diligence at Excel and [94:14] Jim Brier who was the partner and took [94:16] the board seat this is one of the [94:18] all-time great Venture Investments and [94:20] so it's worth talking about some of the [94:21] deal mechanics it was the very first [94:26] post.com institutional Venture Capital [94:29] deal where the founder maintained [94:32] control certainly the norm was as soon [94:35] as a VC gets involved it is a VC [94:37] controlled board and Company the other [94:40] thing that's worth noting is this is a [94:41] company getting a $98 million valuation [94:44] now the silly land that we live in now [94:46] we're like this happens all the time [94:48] this didn't happen and we were only just [94:51] coming out of the era so eyeball clicks [94:54] had just had four years of demonization [94:57] from everyone from the press to the [95:00] public to limited partners in fact Excel [95:04] had limited partners drop out of this [95:06] fund who were LPS previously including [95:09] your beloved Princeton including Harvard [95:12] I mean big and I think Stanford was one [95:14] of the only ones that really stuck with [95:15] them for this $400 million fund and [95:18] they're kind of looking at this [95:20] realizing this might be one of the [95:22] greatest companies of all time and we [95:23] are going to have to do the type of deal [95:26] that everyone got raked over the coals [95:28] for 5 years ago in the Mania for doing [95:31] and no one is doing this type of deal in [95:33] this [95:34] environment but it's Facebook so we're [95:36] going to do it I think the level of risk [95:39] and reputation risk that they took on [95:42] this cannot be underscored enough [95:44] totally I was trying to think of were [95:47] there any other deals like this that you [95:51] could even kind of point to as a [95:54] comparable from the past and the only [95:55] one I can think of was the Google series [95:57] a which was $25 million at a $100 [96:01] million post money valuation split [96:04] between Michael Moritz from seoa and [96:06] John door from Kleiner Perkins but it [96:09] was in the dotcom runup exactly exactly [96:11] it was a totally totally different [96:13] Paradigm and the legend has it I don't [96:16] know if this is apocryphal or not is [96:18] that after making the investment Moritz [96:21] told Don Valentine at seoa never have we [96:23] paid so much for so little yes and I [96:26] think he was actually referring to like [96:27] the small amount of equity that they got [96:29] by having to split the deal in Google so [96:32] yeah in the do bubble that happened but [96:36] like that this would happen in 2005 [96:38] Excel went way out on a limb for this [96:41] yeah so some other interesting deal [96:42] points $1.1 million of this $12.7 [96:46] million round was done by Jim Brier [96:49] personally wow yeah the shares were [96:53] acquired [96:54] at 4.5 cents per share Facebook just [96:58] recently hit $600 a share so Excel and [97:03] their limited partners and Jim for [97:05] anyone who's still holding the shares [97:06] after they distributed them that is a [97:09] 13,000 X [97:12] return another interesting thing on this [97:15] do you know the whole Don Graham Dynamic [97:17] with this deal yes the original deal was [97:21] that Don Graham from The Washington Post [97:23] was going to invest and it wasn't going [97:24] to be a VC deal that was at a $60 [97:26] million valuation Excel comes in over [97:29] the top and in part of the negotiation [97:31] to get it up to this $98 million deal [97:35] there was actually a secondary oh wow 3 [97:40] million went to Mark Dustin and Shawn as [97:45] a secondary in this deal wow I had no [97:50] idea yep why I mean add to the Litany of [97:53] things that were not done back then this [97:56] is Facebook having all the leverage and [97:59] full deal control in negotiating this [98:01] series a wow and still it was one of the [98:04] best Venture Capital Investments of all [98:06] time absolutely man that is freaking [98:09] wild [98:11] yeah wow okay well that is a lot of hype [98:14] to live up [98:16] to which brings us now to the summer of [98:20] 2005 where Mark motivated by perhaps I [98:24] hadn't quite thought about that [98:26] announces to the company his product [98:29] road map for the summer and it is a [98:34] Sixpoint product plan number one a [98:37] redesign of the site bye-bye Facebook [98:42] guy number two a photo application which [98:46] by the way photos got written by like [98:49] one guy in two weeks yes move fast and [98:52] break things yep [98:54] number three a personalized newspaper [98:57] based on all of your friends activity [99:00] number four an events feature number [99:03] five a local business product and number [99:05] six a sort of vague idea for a feature [99:09] called I'm bored which would let users [99:14] on the site who were bored consume media [99:18] and play games that's a pretty big [99:21] feature mandate this is freaking wild so [99:24] now okay of that I think only the [99:27] redesign and photos would actually ship [99:29] in 2005 maybe events did too I'm not [99:33] entirely sure but this is incredible [99:36] photos [99:37] Newsfeed platform which is really what [99:40] this I'm bored feature is Summer 2005 [99:44] it's all right there in the vision and [99:46] Mark thinks they can accomplish all of [99:48] this by the end of the [99:51] summer and the wild thing is they do [99:53] accomplish all of this in the next two [99:56] years which two big points on this for [99:59] any other company these are all like [100:02] multi-year long development processes [100:05] well and each one is its own company [100:07] exactly photos is flicker there are [100:09] independent game platform companies out [100:12] there mini clip Etc yeah flash games [100:16] what were some of the other ones events [100:18] Eventbrite yep personalized newspaper [100:21] well that becomes a true true innovation [100:24] in Newsfeed but I think at the time [100:26] there was a lot of Buzz and talk about [100:29] personalized portals and like Yahoo was [100:31] really big on this and Google was even [100:33] kind of big on this of like oh make this [100:34] your homepage oh I Google do you [100:37] remember I Google yes I Google exactly [100:40] and I think that might have been [100:42] somewhat of the inspiration although [100:43] obviously Mark was thinking way bigger [100:45] than that but like yeah again local [100:47] business product Yelp each one of these [100:50] under the old Paradigm was its own [100:52] company and Mark was like no this is all [100:54] part of Facebook crazy super crazy and [100:58] this really I think speaks to the genius [101:01] of Mark as a product [101:04] strategist it's one thing to say I'm [101:06] going to add a lot of random features to [101:09] my site and I'm going to throw spaghetti [101:11] against the [101:12] wall this was not a random grouping of [101:16] features [101:17] everything reinforced one another and [101:20] drove the engagement Loop of the site so [101:25] just as one example let's take from you [101:28] know this road map that Mark lays out [101:30] events let's start there events and [101:33] parties planned on Facebook get tagged [101:37] with who's going to attend the events [101:40] all of those tags get published out as [101:44] activity to Newsfeed that drives [101:47] interest in fomo among the friend [101:49] Network so more friends come to the [101:52] actual event at the event photos get [101:56] taken and now there's more people there [101:58] well those photos get posted back to the [102:01] site afterwards and they all get tagged [102:03] with the people who were there who are [102:05] in the photos all of that becomes [102:07] activity that then gets publish to [102:10] Newsfeed that generates more desire from [102:13] more people who see that on Facebook in [102:15] the newsfeed to either engage with the [102:18] photos and like and comment on them or [102:21] if they're hearing about it from their [102:22] friends and they're not yet on Facebook [102:24] to now register for Facebook and get [102:27] involved in this Loop all of that now [102:29] gets published back to Newsfeed so now [102:31] it's time for the next party and you can [102:33] see how this just grows and builds on [102:34] itself over and over and over again it's [102:36] funny as you're saying all this it [102:38] sounds like very old hat like Almost [102:40] Boring of course it does all of that at [102:42] the time it was so revolutionary [102:44] everything I just described was there [102:46] within the next 18 months so photos I [102:49] think was like late summer of 05 [102:51] interestingly photos did not originally [102:53] include photo tagging photo tagging was [102:55] like a pretty New Concept I think [102:58] flicker may have had it but the idea [103:00] that like you're tagging a person and [103:02] then you can go browse that person's [103:04] profile by photos they've been tagged in [103:06] that wasn't brand new innovation that is [103:08] a mechanic that was not thought of in [103:11] social networks before if you think [103:13] about what Friendster and Myspace were [103:15] you could upload a limited set of photos [103:18] which by the way you'd have to delete [103:19] one to add another yes so Myspace only [103:22] allowed you to have eight photos maximum [103:24] at the time and this is hey Ben what [103:27] you're describing of person tagging in [103:29] photos incredible Innovation right that [103:33] drives the whole viral Loop and this is [103:35] a through line through the whole episode [103:37] what social media or social networks are [103:40] definitionally changes every year and [103:43] often Facebook was the one in the early [103:45] days to push the onvelope and say this [103:47] is what it means to be a social network [103:49] in the later years it was Facebook's [103:51] competitors that then they had to sort [103:52] of adopt that functionality but the idea [103:55] that a social network includes an [103:58] infinite number of photos that you can [104:00] tag and give XY coordinates on the photo [104:02] to map to a specific person that is [104:03] another entity in The Social Network [104:05] that was actually a new component to [104:08] what it meant to be a social network and [104:10] as you're saying with Newsfeed that [104:12] wouldn't happen for another what was [104:13] that late summer of 2006 or something [104:16] September 2006 yeah so another year year [104:20] and a half after this that completely [104:22] turned what social media social [104:24] networking was on its head again yes oh [104:28] we're going to talk all about it in a [104:29] sec to this point a social network is a [104:32] set of static profile pages that you can [104:34] navigate to and if you go to home like [104:37] the root you know facebook.com/ instead [104:39] of SL profile PHP it's pretty useless [104:42] like the homepage is actually an [104:43] uninteresting place to hang out there is [104:45] no hey let me see what's going on in my [104:48] network yes until [104:51] Newsfeed okay but before we get get to [104:53] that 2005 to your broader Point here [104:56] though yeah for the first six or seven [104:59] years of its life Facebook was a change [105:02] maker defining what social networking [105:04] was and then after that it became a [105:07] change taker we'll talk about this when [105:09] it happens but like again the mental [105:12] flexibility of Mark and the company to [105:14] be like okay I'm not going to have pride [105:17] about that anymore I'm going to adapt [105:19] and still win was incredible very [105:21] microsof Ian very microsof Ian okay so [105:26] even this even though it takes two years [105:28] to roll out the full road map photos [105:31] events but especially photos the 20052 [105:35] 2006 school year is just off the charts [105:38] for growth sharing engagement among [105:41] college students Ben you found a like [105:43] incredible stat about Facebook's page [105:46] views right yeah this stat is Crazy by [105:49] November of [105:50] 2005 they were getting [105:53] 230 million page views daily which means [105:57] that they had passed Google in page [106:00] views wow Google the company started six [106:03] years earlier and the reason is because [106:06] when you're on Google you do like one or [106:07] two searches and then you go to your [106:09] destination on Facebook you get caught [106:11] in an hours long trance of looking at [106:14] everybody you know and what they're [106:15] doing right which was already the case [106:18] when it was basically text only but [106:20] adding photos I mean oh my God every [106:24] photo is a page view with such a small [106:27] user base at that point in time I think [106:28] what probably certainly less than 10 [106:30] million users to have 200 million plus [106:35] page views every single day and be [106:37] passing Google in traffic is wild I mean [106:40] let's say it's 10 million it's 23 page [106:42] views per day per person that's assuming [106:44] that every single person who is a [106:46] monthly user is accessing it every day [106:48] and loading a page 23 times right right [106:51] in our mobile age now that does doesn't [106:53] really sound like much I bet the number [106:54] of photos that someone Scrolls through [106:56] on Instagram is way higher than that but [106:58] for the time that interactivity was just [107:01] nuts I mean certainly flicker Myspace [107:05] Friendster couldn't hold a candle to [107:07] that type of Engagement yeah so here's [107:09] their user growth in June of 05 as you [107:11] mentioned they were at 3 million by [107:13] September of 05 they were at 5 million [107:16] users that was 10x their user base just [107:19] a year ago in September of 04 and and [107:23] almost a third of all us college [107:27] students were included in that 5 million [107:30] so by September of 05 they had a third [107:33] of us college students of their five [107:37] million users that they had about 18 [107:39] months in 70% were daily active 85% were [107:42] weekly active 93% were monthly active [107:46] wow those are just insane numbers yes I [107:51] could go on and on and on but one [107:53] interesting thing to also point out at [107:54] this period of time October of 05 they [107:57] were up to 8.3 million users they were [108:00] the 10th most visited site on the [108:01] internet but the important part is they [108:03] were doing a million dollars a month in [108:05] Revenue they had actually started [108:08] figuring out the advertising business [108:10] model so here we are 19 months after [108:12] founding they're no longer burning [108:14] Capital yes well when you say figured [108:17] out the advertising business model they [108:19] just had such a high volume of page [108:22] views that they didn't have to figure [108:24] anything out right you just plug in some [108:26] crappy ad networks and like print money [108:29] Fair yeah you're right actually I [108:31] completely misspoke I would say that [108:32] they were nowhere near figuring out the [108:34] advertising business model but they had [108:36] a traffic machine they were able to [108:39] through Alchemy turn page views into [108:41] Revenue yeah but they had not by any [108:44] means figured out their business model [108:47] and in this world where startups glorify [108:49] raising Capital burning huge amounts of [108:52] money [108:53] delaying monetization and then having [108:56] these amazing screaming returns when [108:58] they finally do turn on the money faucet [109:00] Facebook was just the opposite they lost [109:02] money and not that much money for like a [109:05] year and a half and then from that point [109:06] on they were just profitable yes so the [109:10] 20052 2006 Academic School year [109:13] incredible for the company everything we [109:15] just talked [109:16] about but there were a [109:19] couple yellow flags shall we say that [109:22] popped up during that year mark of [109:26] course was already starting to think [109:28] about well how can this keep getting [109:30] bigger people graduate from college this [109:34] is great in colleges but like I don't [109:36] want to just build a college site I want [109:40] this to be a lot bigger and obviously [109:42] that was what Excel was investing into [109:45] at that valuation so they launched two [109:48] sort of tests throughout the year one [109:51] was opening up [109:53] Facebook to high school students and the [109:55] other one was opening Facebook up to [109:59] workplace groups and both of them kind [110:02] of flopped and for different reasons [110:04] right the workplace ones they were like [110:06] the workplace thing is going to be [110:07] awesome because it's authenticated email [110:08] addresses the same way that the colleges [110:11] had authenticated email addresses most [110:13] high schools don't have authenticated [110:14] email addresses so we expect these [110:16] workplace networks to work better yep [110:20] but the problem was going into workplace [110:22] networks people didn't want to share [110:24] with their colleagues I don't want some [110:26] low reses party photos of me from last [110:28] night showing up on my workplace and I [110:30] think also just the density wasn't there [110:33] it was violating the principles that got [110:35] them there of like this thing needs to [110:37] be alive and every way we fractally [110:40] spread out needs to be bringing that [110:42] nuclear reactor of what's Happening [110:44] along with it going straight into work [110:46] places like you've got 60y olds and [110:49] they're like there's going to be nothing [110:50] compelling for them right now the [110:52] extreme [110:53] bease on Facebook at this moment in time [110:55] is you started with the group of people [110:57] who are the most social and the most [110:59] open to share in their entire lives [111:02] you've already saturated a third of them [111:04] every single cohort that you add from [111:07] here is probably going to be worse yep [111:09] and that was a very rational argument [111:12] now high schools as you say failed for a [111:15] different reason there wasn't the same [111:18] standardized email architecture in high [111:20] schools across the country it was just [111:22] way more fragmented so you couldn't [111:24] elegantly set up these private networks [111:26] in the same way Y and in fact what they [111:28] ended up doing cuz I was in high school [111:30] at this time it was technically a [111:32] different Facebook in fact I will quote [111:34] the homepage when you went to sign up if [111:35] you went to facebook.com in 2006 the [111:37] title was Facebook is an online [111:39] directory that connects people through [111:41] social networks at schools now there are [111:43] two Facebooks one for people in colleges [111:46] and one for people in high school the [111:48] site is open to a lot of schools but not [111:50] everywhere yet we're working on it and [111:51] so if you signed up as a high schooler [111:54] you got the crappy one you had to log [111:56] into [111:58] h.f. and it looked mostly the same it [112:01] had slightly different features but [112:03] there was this weird thing where [112:04] somebody who was already in college had [112:07] to invite you it's not like anyone could [112:09] sign up for High School Facebook you [112:11] needed a college student who went to [112:13] high school with you to basically vouch [112:15] for the fact that you are a real high [112:16] school student and thus eligible to sign [112:18] up for hsf.net [112:22] L remember this of my younger friends [112:25] who are still in high school asking for [112:27] this and the subd do meing thing was [112:29] that lasted for a long time because the [112:32] college Facebooks were subdomain as a [112:35] part of their like Tech infrastructure [112:38] you know harvard. face.com only went to [112:41] the set of servers that they had set up [112:42] for that it made it all really easy from [112:44] an infrastructure perspective I was on [112:47] hsb.com for a long time as a college [112:50] student right [112:54] yeah because there had to just be one [112:56] hsf.net [113:06] workk or like certainly the largest [113:08] subdomain was the high school one and so [113:11] I think like within a month or two of [113:12] allowing High School signups you know [113:14] high school as a whole was much much [113:16] much much larger and they had to solve [113:17] for the technology constraint of what do [113:20] we do with this it doesn't behave like [113:21] any of our networks y so the academic [113:25] school year in the core College user [113:28] segment went great but once you hit [113:31] summer 2006 all those kids go on break [113:35] and college growth stops for the moment [113:39] so summer 2006 was kind of a scary [113:42] moment for the company the initiatives [113:44] to expand outside colleges were not [113:46] going super well College growth had [113:48] stopped for the moment and yeah Ben like [113:50] you said like very rational argument [113:52] that like okay this thing is going to be [113:54] a really interesting Niche site for [113:57] college students and how valuable is [113:58] that well Viacom which owned [114:03] MTV thought that that was worth $750 [114:07] million which is what they offered to [114:10] buy the company for in the summer of [114:13] 2006 well there was another party that [114:16] really really wanted to own Facebook at [114:18] this point in time and that was Yahoo [114:22] and Yahoo was willing to top Viacom and [114:26] pay a billion dollars to buy the company [114:31] so here we are a billion doll offer on [114:35] the table and growth is slowing exactly [114:37] exactly there's this existential [114:39] question mark about the company growth [114:42] has slowed a lot you've got a billion [114:45] dollars on the table which is a lot of [114:48] money especially in those days it was [114:51] shocking six years later when Facebook [114:54] bought Instagram for a billion dollars [114:56] this would have made the careers of [114:58] everybody involved yep and you know Mark [115:01] of course controls the company controls [115:03] the board thanks to Shawn Parker he [115:04] doesn't have to sell but the whole [115:06] management team is like yeah we should [115:09] probably [115:10] sell and I think the board you know it's [115:12] hard to know they obviously know they [115:14] can't force him to sell but I suspect if [115:17] you had asked all of their opinions and [115:19] certainly Mark did they probably would [115:21] have been like like you know it's been a [115:24] great run we should probably hit this [115:26] bid so [115:28] Mark [115:30] actually agrees to the Yahoo [115:34] deal billion dollars it's going to be [115:37] all in stock yaho is going to buy the [115:40] company in the interim between when they [115:43] shake hands on the deal and while the [115:46] docks are being drafted Yahoo announces [115:49] their quarterly earnings and they have a [115:51] bad quarter and the stock drops 20% and [115:55] Terry seml who was the CEO of Yahoo at [115:57] the time said okay the deal is you know [116:00] for the same amount of stock so [116:02] commensurately the deal is now worth [116:04] $800 million not a billion [116:07] dollars and that was history turning on [116:10] a knife point and that was all Mark [116:13] needed to say you know what thanks but [116:14] no thanks and this part of the story [116:16] never gets told the fact that actually [116:19] at First Mark did not turn down the [116:21] billion dollar offer from y [116:22] Mark actually accepted or at least said [116:25] let me turn over another card and get [116:27] one inch closer to the negotiation being [116:29] final you know it wasn't an outright [116:31] rejection at first it was yeah come back [116:33] with the papers yep totally that is not [116:35] the way it gets talked about today but I [116:37] think actually makes the story like kind [116:40] of even more powerful it almost really [116:43] did slip away well it's certainly much [116:46] more realistic it's a simple and [116:48] Powerful story to say they just outright [116:51] rejected it but that's just never how [116:53] these things go it's like uh show me [116:54] you're real yep okay so analyzing let's [116:58] say he just did turn it down outright [117:01] well here's a reasonable way to look at [117:02] it if what you want to do is run a great [117:05] company for the rest of your life it [117:07] actually was totally rational Mark has [117:10] said publicly I mean this like 15 18 [117:12] years ago I didn't have any more ideas [117:14] as good as Facebook but that's like way [117:16] under selling it there aren't better [117:19] companies to start right yes it would [117:21] have been nice to have certainty on a [117:23] billion dollars or whatever percent of [117:24] the billion Mark owned at that point and [117:27] also a year before Myspace had gotten [117:29] bought for 580 million right by news [117:31] court so it's like gez okay twice as [117:34] much as Myspace but rationally if the [117:38] optimization function is I want to run [117:40] the best company I can for a very long [117:42] time almost nobody has started a better [117:44] company since so if you're looking at [117:47] the engagement you're looking at the [117:49] potential the rational thing actually is [117:52] just keep running this company because [117:54] I'll never discover something like it [117:55] again yep I completely agree with you [117:59] and Mark has said so much in so many [118:01] words to You and Me personally that he's [118:04] not optimizing for financial outcome [118:07] here he's optimizing for impact and I [118:09] think he likes running this company he [118:11] would want to go start a similar company [118:13] if he didn't have this one yes totally [118:16] also though to his skill as a strategist [118:18] even at this very young age yes Ben you [118:21] and I just made a comp telling argument [118:23] about why you could have believe [118:26] Facebook was going to top out with the [118:27] college Market at this point in time [118:30] Mark knew though and he had planted the [118:32] seeds that both Newsfeed and open [118:35] registration were going to be coming [118:37] later that year and he was definitely [118:40] making that calculus here too of like [118:42] well I don't know what the probability [118:43] necessarily is that those are going to [118:45] hit but Mark had the belief like a if [118:48] they do they're going to be really big [118:50] and B I think the prob is higher than [118:53] the people around me think it is yep [118:54] that's totally fair even if you weren't [118:56] going to bet on the growth the user [118:58] engagement was still currently great and [119:00] you had a lot of reason to believe that [119:02] it was going to get even greater and you [119:04] compare all the metrics with frster and [119:07] mypace as time went on got worse people [119:11] churned their daily active to monthly [119:14] active would just go down over time the [119:16] page loads times would take forever [119:18] especially with well actually with both [119:20] frster just was not architected properly [119:23] it was a software engineering computer [119:24] science issue Myspace was this weird [119:27] thing that was sort of born out of this [119:30] combined media conglomerate and they [119:32] just never really had excellent Tech [119:35] architectural talent and so as you sort [119:38] of looked at not the highle metrics of [119:40] how many users have ever signed up [119:41] because Myspace and frster I think we [119:43] still way ahead at this point if you [119:45] looked at like hm but how's this going [119:47] to play out if it sort of keeps [119:48] compounding and you looked at the deeper [119:50] metrics you kind of thought oh I'm [119:52] running the better company by a lot once [119:54] especially you know we got to figure out [119:56] the business but I am running the better [119:58] product by a lot yeah so okay to [120:01] figuring out the business coming out of [120:03] this episode with Yahoo They say no the [120:06] whole management team turns the whole [120:08] management team basically turns over the [120:10] next set of months Mark gets religion on [120:13] a couple things and actually we should [120:16] say not Dustin but a lot of the people [120:19] around the table who were expecting a [120:20] nice cash payout and aren't getting one [120:22] or are not pleased yep totally so what [120:26] are they going to do the first thing is [120:29] I think Mark had always operated the [120:31] company this way but now he gets real [120:33] religion of like we need to get more [120:35] revenue and we need to focus on becoming [120:38] truly cash flow positive so that we are [120:40] never in a position again where we would [120:42] even consider doing something like this [120:46] so in August of 2006 they do the first [120:50] partnership with Microsoft [120:52] it's funny that you know the Microsoft [120:54] Facebook relationship in some ways is [120:56] sort of a precursor to the Microsoft [120:59] open AI relationship today and how that [121:01] gets built here's what it is Microsoft [121:04] says we're going to take over selling [121:08] all of the display ad inventory [121:11] domestically within the US for Facebook [121:14] and we will give you a guaranteed CPM [121:16] that we can sell and then we are going [121:19] to use this to help bootstrap our online [121:22] services division which as we talked [121:24] about in our Microsoft series becomes [121:26] super super important not only for their [121:29] efforts launching Bing making that into [121:31] ultimately a successful business but [121:33] even more importantly for Azure that [121:36] comes out of it Microsoft knows they [121:37] need some scale to bootstrap up and get [121:41] started with both the ad business but [121:43] also just like the online services [121:45] division period yeah this is a ton of [121:47] inventory like you just think about the [121:49] amount of page views that are happening [121:50] here suddenly Microsoft's problem is not [121:53] how do we find enough inventory to sell [121:54] it's how do we go find enough [121:56] advertisers to actually fill all these [121:58] slots on Facebook yes and Microsoft have [122:01] been trying to do this they tried to do [122:03] this with Myspace a little bit earlier [122:05] and they lost that deal to Google so [122:08] they really need Facebook and on the [122:10] other hand this is a great deal for [122:11] Facebook because Facebook sucks at [122:13] selling ads and Facebook has no [122:15] targeting or anything so they really [122:16] shouldn't justify high cpms at this [122:18] point yeah Outsource it to a good third [122:20] party Network right [122:22] so this is like basically all of [122:24] Facebook's revenue for the next couple [122:26] years here so Facebook had made9 million [122:30] in Revenue in 2005 when they were [122:33] selling themselves this is been as you [122:35] were saying like hey they're starting to [122:37] turn on ads in 2005 it's starting to [122:39] work like $9 million that's amazing for [122:41] year two as a company got to cash flow [122:43] positive got to cash flow positive in [122:46] 2006 with this Microsoft partnership [122:49] that jumps to $48 million in Revenue oh [122:53] that's where the money SP is now we're [122:56] talking and then the next year in 2007 [122:58] again I think almost all of this is [123:00] Microsoft that goes to $153 million in [123:04] Revenue so like okay we are way out of [123:08] any League where we consider selling [123:10] ourselves here we can control our own [123:13] destiny isn't it wild thinking about you [123:16] thought it was crazy to turn down a [123:18] billion dollar offer or many people [123:20] thought you were crazy and then just two [123:22] years later that is only a 6X Revenue [123:26] multiple and you're tripling [123:28] year-over-year oh and by the way in the [123:31] second iteration of this Microsoft [123:33] partnership which we will talk about in [123:34] just a minute it included an investment [123:37] from Microsoft at a $15 billion [123:41] valuation oh hold your horses let's [123:43] finish this great 20062 2007 Arc and [123:46] then we'll get there yes point being [123:48] though Mark made the right decision for [123:50] all shareholders of Facebook to walk [123:52] away from the Yahoo deal I mean it's an [123:54] all-time high in walking away from every [123:56] deal ever he's made the right decision [123:57] yes but like within a year he made the [124:01] right decision yes oh yeah that's true [124:03] that this became obviously right fast [124:05] very fast very fast so how does this [124:07] happen okay Adam D'Angelo finally [124:10] graduates from Caltech joins Facebook [124:12] fulltime as CTO I love that Adams are [124:14] like second protagonist here I know I [124:15] know I don't know if you did that [124:17] intentionally but yeah well he was right [124:19] there I mean they were an exer together [124:21] like I don't think you can separate it [124:22] out yeah and you know I think here part [124:25] of the reason why Adam keeps coming into [124:28] the story is he is a really really great [124:32] technologist and even though he's young [124:34] same age as Mark when he now arrives [124:36] full-time as [124:37] CTO this is when Facebook starts to be [124:40] taking the steps to be building like [124:43] real technology infrastructure so before [124:46] Adam joins full-time the team was [124:49] shipping new code to the website like at [124:51] a high velocity relative to everybody [124:54] else out there you know it was [124:56] weekly-ish maybe every couple days you [124:58] know maybe even up to daily like well [125:00] they were a web company for the first [125:02] time in history you actually could all [125:03] these companies with big CL mean [125:05] Microsoft would ship every three years [125:06] with a service pack once a year it's a [125:09] whole new era where all you have to do [125:10] is you know upload some new PHP code to [125:13] the FTP server and then boom the [125:14] application behaves differently the next [125:16] time someone pulls down the page it's uh [125:19] the new technology era of interpreted [125:22] languages running on web servers in a [125:25] browser now that products are in [125:26] browsers meant you could architect your [125:29] company differently and ship differently [125:31] yes and when Adam becomes officially a [125:34] CTO he's like actually we should [125:36] probably just ship multiple times a day [125:38] we should just be shipping all the time [125:41] and also we should probably start [125:43] recruiting like a real top tier [125:47] engineering team because remember it's [125:50] still kind of nuclear win for startups [125:53] Great Tech Talent is [125:55] available so they start going and [125:58] recruiting like really really top tier [126:00] engineering hires and it's a pretty [126:03] compelling offer like a we've got real [126:06] revenue from this Microsoft partnership [126:08] were the best funded startup in the [126:11] valley people want to buy us for a [126:13] billion dollars we turned them down [126:15] that's all sort of the financial reasons [126:17] to take it for a lot of people too it [126:20] was like such a breath of fresh air I [126:23] think Facebook was potentially maybe the [126:26] first company first startup to have an [126:28] open Office plan like just everything [126:30] about how they ran the company was [126:33] different it was the prototypical [126:35] startup I mean to this day early [126:37] Facebook is still what most startup [126:39] culture is aspiring to be I think often [126:41] without knowing it I mean everything [126:43] from the sort of posters on the walls [126:45] with your mantras and your values to [126:48] almost having like an employment brand [126:50] that you really care about cultivating [126:52] to the idea of like we're having just as [126:54] much fun together socially as we are [126:57] working together I I think the modern [126:59] startup culture especially when you [127:01] factor in the shipping everyday open [127:04] Office y comater kind of trains [127:07] companies to become like Facebook was [127:10] 100% I mean there was Google out there [127:12] which had a lot of this element but it [127:14] was a very different thing it was very [127:16] academic very wonky these guys were [127:20] hackers and they shipped you know and [127:22] they were all in it together the other [127:23] thing that was basically true is they [127:25] were not really interested in recruiting [127:28] industry veterans especially on the [127:30] technical side and Mark says this in [127:32] early talks that he was just [127:34] prioritizing raw intelligence over [127:37] experience so one of those super [127:39] intelligent super talented young [127:41] Engineers with a high slope who joined [127:43] back in [127:44] fall5 was an engineer named Chris Cox [127:47] who joined from the Stanford AI lab and [127:50] Chris joins a team of other young smart [127:53] Engineers led by Ruchi svi and Andrew [127:57] BOS Bosworth working on Mark's [128:00] personalized newspaper product idea [128:03] ruchie was one of Facebook's very first [128:05] engineers and along with her husband to [128:08] DTI araral they run South Park Commons [128:10] now BOS was two years ahead of Mark at [128:14] Harvard and had actually been marks TA [128:16] in the intro to artificial intelligence [128:18] CS class that he took and then he later [128:21] joined the company which is actually not [128:23] how we ended up getting a job but they [128:25] did happen to cross paths that way [128:27] before yes BOS went to Microsoft I think [128:29] before joining Facebook y just for a few [128:32] years so the three of them start working [128:34] on Newsfeed and so we've talked about [128:37] things like photos and the fundamental [128:40] architecture of the site Newsfeed [128:42] required like a whole new level of [128:45] engineering prowess to get this thing to [128:47] work like you're not just going to code [128:49] this up in PHP like photos also required [128:52] real engineering but Newsfeed was like [128:55] pushing the stateoftheart of what was [128:58] possible in technology and on the [128:59] internet so the first question you have [129:02] to ask yourself if you're a good product [129:05] designer developer Capital allocator [129:07] someone in Mark's shoes saying should we [129:09] do this is do people want this if we [129:11] build this will it be valuable the [129:14] reason they knew it would be valuable is [129:16] because the company is data obsessed and [129:19] they watch the analytics like a hawk to [129:21] figure out what are people doing on our [129:22] site and they noticed this Behavior [129:24] where people were browsing to other [129:26] people's profiles just to look at them [129:29] and see if anything changed the user was [129:32] doing the heavy compute lifting rather [129:34] than having a personalized newspaper of [129:36] just bouncing around to a bunch of [129:37] people's profiles and saying anything [129:39] new here so there was an engineer who [129:41] did something really kind of hacky [129:43] because they didn't want to at first put [129:45] all the engineering resources into [129:46] building out something like Newsfeed [129:48] which is well when something changes for [129:50] some period of time we just highlight it [129:52] in yellow yes so it's easy for you as [129:54] you're bouncing around to different [129:55] profiles to just see oh hey this thing [129:57] changed and that totally worked they [130:00] watched the lift and that and they were [130:01] like oh it's a good feature people like [130:02] that so that sort of gave them the [130:04] confidence of we should find a way to [130:07] make it more obvious to you when new [130:10] updates happen when things change [130:12] technically David for people who aren't [130:14] in the sort of tech ecosystem why is it [130:17] so difficult to build something like a [130:19] Newsfeed cuz now we're all trained to [130:21] believe that like a feed is um you know [130:23] A Primitive that is available to you as [130:25] a developer when you're building a [130:26] product because feeds are everywhere [130:28] feeds are the core feature of most [130:30] products when you hit the homepage it's [130:32] some sort of feed that was not true at [130:34] the time that was not true in any [130:36] product I actually challenge you right [130:38] now think back to [130:40] 2005 what was a feed on the web what was [130:43] sort of an infinite scrolling they [130:45] didn't exist this was the first one [130:47] right this was the invention of the feed [130:50] like maybe you've got Reddit and dig and [130:54] you could sort of make an argument that [130:55] there feeds their [130:57] paginated rankings of stories so you [131:00] could see like oh what are the most [131:01] important stories but even that's a [131:04] pretty different fundamental thing I [131:06] don't think Reddit had launched yet dig [131:08] maybe exist but yes as you said oh yeah [131:11] Reddit was in the first YC batch which [131:13] was 05 so it wasn't really it was right [131:15] around the same time okay so how do you [131:17] make a feed well if you are an [131:20] algorithms developer the way you would [131:22] sort of think of it is okay well first I [131:25] need to pick a point in time and from [131:28] that point in time and let's call that [131:31] maybe the last time someone looked at [131:32] their feed I need to store that [131:34] timestamp and now I need to go look at [131:36] every single profile of someone that you [131:39] are connected to so this is n and [131:43] download or cache all of their recent [131:46] updates since that time period so I need [131:49] to like store that somewhere I need a [131:50] new place to store a copy of all of this [131:54] information that lives on someone's [131:56] profile or at least pointers to that on [131:59] everybody's profile and I need that to [132:02] happen for every user so now it's N [132:06] squared every single person on the [132:09] entire Facebook needs to have something [132:12] running in the background that is [132:13] looking at every other person on [132:15] Facebook since a particular time and [132:18] then that compute and storage all needs [132:20] to happen fast enough such that by the [132:23] time they want to go check the feed [132:24] again it's happened again and obviously [132:28] like now that happens in real time and I [132:29] think it was something like every 3 [132:31] hours there was like a new batch yeah it [132:34] was four times a day when it initially [132:37] launched there was like a 4X a day [132:38] refresh okay so every six hours yeah it [132:41] was a Cron job because they didn't have [132:42] enough memory to run it in real time [132:44] that's right they needed to happen on [132:46] separate boxes to run this process cash [132:49] the results and then when you load your [132:51] news feed go fetch them so this is like [132:53] a whole new application using the same [132:56] data that the company has to build in [132:59] order to make news feed happen and so [133:01] what you just described is all the [133:03] technical work just making a feed [133:06] possible on the back end this doesn't [133:08] contemplate any of the front end design [133:09] or engineering or incredible [133:11] permutations of how to display this data [133:14] when it does come back given a massive [133:16] combinatorial problem of how might this [133:18] data come back exactly then you need to [133:20] make the feed actually compelling like I [133:24] don't really care about you know [133:25] somebody who I'm tangentially connected [133:28] to what they had for lunch today but I [133:30] really care about like a photo of Ben at [133:33] the acquired Meetup right so now you're [133:35] telling me you want to rank order it by [133:37] something other than chronology make [133:39] another pass and figure out what I think [133:41] is going to be the most interesting to [133:42] you which on its own is an incredibly [133:44] difficult computer science problem what [133:46] is interesting to you what data should [133:48] we use to inform that decision right [133:51] totally who are your close friends who [133:53] do you care about the most you know what [133:55] general news do you care about the most [133:57] what types of stuff do you care about do [133:59] we have to put weights on every [134:01] relationship in this entire complex [134:02] friend graph between every single entity [134:04] and how close they are and then rerank [134:06] that very often eventually yeah yes so [134:11] the team spends the better part of 2006 [134:13] working on this by September it's ready [134:17] and so by September 5th 2006 just in [134:20] time for the new school year they launch [134:23] it and uh people noticed because they [134:26] launch it like to everybody right away [134:30] and this is a massive Paradigm Shift [134:33] yeah it's not opt in or anything this is [134:35] Facebook changing the game of social not [134:38] opt in everybody gets it right [134:41] away they get [134:44] 30,000 angry emails to support on the [134:47] first day from users who are really [134:51] upset about this 10% of the entire [134:55] Global user base signs up for a group a [134:59] Facebook group called students against [135:02] Facebook Newsfeed I'm pretty sure I was [135:04] part of this [135:05] group the irony here is they had only [135:07] just launched open groups across School [135:11] networks days or weeks before so like [135:14] they enable the tool of their own [135:16] vitriol here and also guess how people [135:19] are finding out about this group through [135:22] Newsfeed yes this is the great irony of [135:24] the whole situation there's a literal [135:27] panic in the streets there are people [135:29] protesting outside the office there is [135:31] somebody trying to use a crane to get [135:34] into the third story basically a TV [135:36] truck is trying to cover what's [135:38] happening in the pandemonium and at the [135:40] same time despite everyone telling them [135:43] I hate this thing it's the worst thing [135:45] ever if you look at the analytics people [135:48] love it people cannot get enough of [135:51] scrolling through the new news feed and [135:53] the reason they're really upset is [135:54] actually quite interesting they're [135:56] saying oh it's sharing this with my [135:58] whole network okay but you updated it on [136:01] your profile which has always been [136:04] public to your whole network and I think [136:05] Facebook for the first time kind of [136:07] stepped in it and realized oh even [136:10] though technically this data has always [136:13] you know we didn't change how public or [136:14] private it is it's just as accessible [136:17] people react really strongly when you [136:19] change the [136:21] ease of obtaining that information or [136:24] whether it feels like you are pushing [136:26] that information out versus someone is [136:28] sort of pulling by going to your profile [136:30] and viewing it went from pull to push [136:33] yes so it's the start of the new school [136:35] year we've just come off this sort of [136:37] tumultuous summer growth had slowed [136:41] walked away from the Yahoo deal there's [136:44] this Revolt in the streets against [136:47] Newsfeed which to Mark's mind along with [136:50] open red [136:51] which was supposed to come like 2 Days [136:54] Later of opening up Facebook to anybody [136:57] they delayed that because of the [136:58] newsfeed reaction but to Mark's mind the [137:01] two of these things are like the big [137:03] growth levers to reignite growth for [137:06] Facebook the board the management team [137:09] everybody's like all right we got to [137:10] roll news feedback people hate it this [137:13] is PR 101 we apologize we roll it back [137:19] and to add a kicker [137:21] Newsfeed actually significantly hurt [137:25] Facebook's Revenue because where do you [137:27] think all the page views were coming [137:29] from and all the refreshes oh loading a [137:32] new ad yep loading all the profile views [137:35] and clicking around so page views are [137:37] actually going to go down in this new [137:40] paradigm every reason is aligned to roll [137:43] this back so Richie and Chris and BOS [137:46] though are looking at all the data and [137:48] they're like holy crap engagement is [137:51] through the roof yeah people can't get [137:53] enough of this people say they hate it [137:55] but yeah it looks like they can't get [137:56] enough of it so Mark [137:59] decides that he is going to write a post [138:03] on Facebook about this and it's titled [138:05] calm down breathe we hear [138:12] you and he announces that they're going [138:14] to launch a set of controls for you to [138:17] control what of your activities get [138:19] published to Newsfeed and what don't [138:20] which by the way sounds like something [138:22] that Mark would say today I think this [138:24] 2006 era is like the last time and then [138:28] you would have had like a 17-year break [138:31] and now we're sort of getting that Mark [138:33] again yes you can't imagine Mark Circa [138:36] 2018 saying calm down breathe we hear [138:39] you yes so it's kind of a brilliant [138:43] strategy of like hey I acknowledge that [138:46] this was surprising we didn't handle the [138:49] roll out right I'm seeing that you all [138:52] actually like this in the data let me [138:55] give you some controls so that what [138:57] you're really worried about you have [138:59] some control over and then let's just [139:02] see what happens and within two weeks [139:05] it's like magic everybody just gets used [139:07] to it and that this is now the way [139:10] Facebook and social media [139:12] operates and engagement continues to [139:15] Skyrocket y but it's funny I did just [139:17] pull up we've been getting a lot of [139:18] feedback about Newsfeed we think they're [139:20] great products but we know many of you [139:21] are not immediate fans and have found [139:23] them overwhelming and cluttered I don't [139:25] think clutter was what people were [139:27] complaining about little bit of [139:28] redirection there yeah it's funny we [139:31] didn't take away any privacy options [139:32] your privacy options remain the same the [139:34] privacy rules haven't changed none of [139:36] your information is visible to anyone [139:38] who couldn't see it before the changes [139:39] blah blah blah it's like it is [139:41] interesting it's true and not relevant [139:44] that's not what people are mad about [139:46] people are pretty aware that this is the [139:48] same information well and I think to [139:50] sort of set a theme here that we're [139:52] going to come back to several times the [139:55] definition of what social media is is [139:59] actually very [140:00] fluid and it changes in the consumer's [140:06] mind I'd even say before this there [140:08] wasn't social media or at least Facebook [140:11] wasn't social media Facebook was a [140:12] social network but this was the first [140:14] time they introduced a media component a [140:16] thing you would read right inspired by a [140:19] newspaper right and I think when these [140:21] Paradigm shifts happen people get upset [140:24] because their expectations are being [140:26] violated so it actually doesn't really [140:28] matter what the Privacy is or isn't it's [140:30] the expectations right I'm glad you [140:32] planted the seed because this will come [140:34] back over and over again in their [140:35] history of people now feel differently [140:39] and the product needs to change with [140:41] those societal expectations in order for [140:43] people to not be upset about it yep [140:45] unfortunately for the moment I think the [140:47] lesson that Mark and the company took [140:48] from this experience was well we [140:52] actually know what's best and the user [140:54] base will just get used to it whenever [140:56] we make a change it just happened to be [140:59] that that was true with Newsfeed but [141:01] that wouldn't always be true y so [141:03] originally the two big initiatives [141:05] Newsfeed and open registration were [141:07] supposed to basically launch together [141:10] and they chose to do Newsfeed first [141:12] because it was going to be a a little [141:14] bit easier to roll out and be more [141:17] valuable to the core College audience [141:19] that was just St in back for the new [141:21] semester the fall semester at school so [141:23] they launched that first and then yeah [141:25] the original plan was open registration [141:27] Facebook is now open to anybody in the [141:28] world was going to come two days later [141:30] on September 7th they obviously put that [141:33] on the shelf for a long time two and a [141:37] half [141:39] weeks at the end of September open [141:41] registration launches anyone can sign up [141:44] for [141:45] Facebook this is another major major [141:48] change before all these networks were [141:50] siloed you had to be part of a verified [141:52] email address Network to join Facebook [141:54] now anyone can sign up this is the [141:57] beginning of uh oh my mom is on Facebook [142:00] exactly exactly exactly which ultimately [142:03] would be a problem because that's what [142:04] created room for Instagram for Snapchat [142:07] etc etc but that would actually be down [142:11] the line in a weird sort of perverse [142:14] way I think because of all the [142:16] controversy around [142:18] Newsfeed when open regge launch [142:21] there wasn't the same kind of [142:22] controversy people were already [142:24] desensitized or they had just been [142:26] through news feed and they're like all [142:28] right yeah whatever nobody even really [142:29] like noticed that much especially [142:32] because the product experience didn't [142:33] change for the already engaged users and [142:37] to the extent that you did feel that [142:39] privacy had changed you're now kind of [142:42] used to anything I put on Facebook gets [142:44] broadcasted so who cares if more randos [142:46] come in it's already getting broadcasted [142:49] yep yep and [142:51] now with Newsfeed there is this [142:53] mechanism that makes sure like hey even [142:56] as randos from your perspective join [142:59] Facebook you're still seeing the updates [143:02] of people you care about well I don't [143:05] think at first when it was this like [143:06] pure chronology yeah I guess that's a [143:08] good point as they make it more [143:10] algorithmic but this was still a period [143:12] in time where your friends were your [143:13] friends I mean Facebook was only 2 years [143:15] old and so for the vast majority of [143:18] users they joined in the last year so [143:20] they didn't have anybody who wasn't [143:22] really their friend as a friend now I [143:25] just have to treat Facebook posts as if [143:27] they're public because the group of [143:30] friends is aged but in that period of [143:33] time you could trust if something was [143:35] getting published to your friends that [143:36] like just going to your friends right [143:39] the age of the network itself was only [143:41] two years so your friendships hadn't [143:44] shifted that much well open regge [143:48] despite High School and workplaces not [143:52] working well open regge works really [143:54] well yeah I mean Facebook has a brand at [143:56] this point it's the best social network [143:58] it's the fastest growing social network [143:59] it's the one that all the college kids [144:01] will always be the cool people in any [144:04] society at any time that's the age group [144:06] of trend setting and they have conquered [144:09] that market high school kids want to be [144:11] like college students and people not in [144:13] college want to be like college students [144:15] yep so over the summer Facebook was [144:17] adding like 5 to 10,000 users a day by [144:23] late fall after launching open [144:24] registration they're now adding 70,000 [144:27] users a day wow by fall of 06 yeah wow [144:31] growth really picks up so this brings us [144:34] now to Spring of 07 and the final piece [144:38] of Mark's original product road map I'm [144:42] bored AKA platform and platform is this [144:47] kind of Forgotten thing right now I mean [144:49] like it's Farmville it's Zinga it's apps [144:52] on Facebook it's quizzes it's quizzes [144:56] yes that will come back up later that'll [144:58] come back [145:00] for a period of about four years from [145:05] 2007 to [145:07] 2011 this is the most important thing in [145:11] the company and this is what everybody [145:14] believes the core and future of the [145:16] company is Facebook has become a [145:18] platform this is the goal of a [145:19] technology company become a platform be [145:21] like Microsoft this is now Facebook is [145:24] like Microsoft we are a developer [145:26] platform and there was this incredible [145:29] tension at least it seemed like the tech [145:32] media wanted to play up this tension are [145:33] they an advertising company or a [145:35] platform company because they're super [145:37] different business models with very [145:38] different incentives and users kind of [145:39] need to know how to think about it and [145:41] it was kind of at the same time Cheryl [145:44] had just joined the company or was about [145:45] to join the company the next year in [145:47] 2008 so yeah not yet and so they really [145:50] had gotten serious she sort of LED them [145:52] to figure out hey let's survey all the [145:54] business models do some work on each of [145:55] them and figure out and commit to being [145:56] an ads company at this point in time [145:58] like they were showing ads but the whole [146:01] leadership team at least I'm convinced [146:05] believed we are on the way to becoming a [146:08] platform 100% so May 24th 2007 Facebook [146:14] holds its first developer conference to [146:16] launch their developer platform f8 which [146:19] again let's take a step back this whole [146:21] idea is kind of wild a developer [146:24] platform on the web the web is a [146:26] developer platform Facebook is now [146:29] becoming a platform on top of an open [146:32] developer platform but they have two [146:34] important things that as an application [146:35] developer you're interested in one user [146:38] attention you know the same way that [146:40] Microsoft had user attention with PCS [146:43] because they had the install base of [146:44] Windows and two a whole crap ton of data [146:48] about each person that you could then [146:50] and build into your application to make [146:52] it really rich and feel personalized and [146:55] it turned out that that was really [146:57] attractive to developers so at f8 they [147:00] announced that the Facebook user base is [147:02] now over 20 million and it is growing by [147:05] over 100,000 users a day so growth is [147:07] accelerating wait it's 20 million total [147:10] and growing by 100,000 a day yes I mean [147:13] open regge had just launched a couple [147:14] months before so like this thing is [147:17] really taking off oh my God every 10 [147:20] days they're adding another 20th of [147:22] their user base and they're growing by [147:24] five per what is that uh it's like 7% [147:26] every two weeks it's like three and a [147:28] half per a week yeah pretty good wow at [147:31] a scale of over 20 million already wow [147:34] it is the sixth highest traffic site in [147:37] the world and they're announcing to all [147:41] the developers in attendance the [147:43] Facebook API Facebook graph is now open [147:46] to you you can build apps and publish [147:49] and run them [147:50] natively on Facebook for users and part [147:55] of Mark's original product road map from [147:58] 2005 this is huge until this point in [148:01] time Facebook was about digitizing [148:05] everything that happened in your offline [148:06] life your real friends your real parties [148:09] you went to the real photos you took [148:12] there wasn't anything detached from your [148:16] offline reality to do on [148:18] Facebook now all of a sudden you can [148:20] play games on Facebook you can take [148:22] quizzes on Facebook you can use apps on [148:24] Facebook and you can do them all with [148:27] your friends it's pretty compelling I [148:29] remember playing was it scrabulous yeah [148:32] I think I was playing with my sister [148:34] with my grandma I mean is a great way to [148:36] like do things digitally with the people [148:39] that matter to you yep and I think it's [148:41] also an early foreshadowing of the job [148:43] that mobile and the smartphone would [148:45] really do in people's lives which is hey [148:48] there's actually like a lot of white [148:50] space where you're just bored during [148:53] your day I mean Mark's original title of [148:56] the feature of I'm bored was perfect [148:58] this is the cure for boredom it just so [149:01] happened that it was on a desktop and so [149:04] when you were bored at your desk or [149:05] bored at your home mobile then opened [149:08] that up to like hey I'm bored anywhere [149:10] right and it gave Facebook from a [149:13] business perspective it sort of gave [149:14] them another stakeholder hey developers [149:17] are someone who can keep people on the [149:20] platform longer so we can show them more [149:22] ads they might do their own advertising [149:24] so they might drive traffic to Facebook [149:26] that they could help grow the core [149:28] platform itself it gives us sort of more [149:31] lock in as people develop for us and as [149:35] users seek out applications on us [149:38] they're hoping for sort of that [149:39] Microsoft Playbook of platforms get [149:42] really epic lock in yep oh and by the [149:45] way developers on the platform will [149:47] probably have their own business models [149:49] where they'll make make revenue and [149:52] especially if we ultimately introduce [149:53] our own virtual currency we can probably [149:56] make revenue from their revenue too yeah [149:59] so the internal goal of the company was [150:03] to get 5,000 developers in the first [150:06] year who would be making social apps on [150:09] the platform how many they get they got [150:12] 5,000 in two [150:13] days developers went freaking nuts it [150:18] was a great keynote it was a great know [150:20] yes you know they went nuts because yeah [150:23] Ben as you say like oh distribution to a [150:26] lot of people and then like bringing [150:28] along your friends and data on them like [150:30] that's very powerful incentive as a [150:32] developer to go go make an app or a game [150:34] for the platform you're building a web [150:36] app that has an unbelievably Rich set of [150:39] data that you can hook into the thing [150:41] that I was I don't think I could have [150:42] put into words then but I can now is has [150:46] a successful scale durable platform ever [150:49] been built that wasn't an operating [150:53] system like it always felt weird to me [150:55] in that moment that Facebook thought [150:57] they could be a platform because I was [151:00] like well it's a website I mean it's a [151:01] web app and I have profile and I have [151:04] all this Rich information but it's not [151:05] like running on my [151:10] device well [151:12] Ben turns out you would have been right [151:15] if only I had the the words but what do [151:17] you do with that information what was I [151:18] going to be bearish on the company that [151:20] would have been a super wrong call right [151:22] well this is the Amazing Story that [151:24] we're telling here of like they keep [151:25] surviving and thriving despite not [151:29] having the operating system yeah [151:31] listeners right in if you can think of [151:33] one a successful scaled durable platform [151:37] where the platform is able to make money [151:39] and they're able to make a lot of money [151:41] for Developers for people on the [151:42] platform that is not an operating system [151:44] that runs on Hardware Hardware either [151:49] captively controlled Hardware like apple [151:51] or open Hardware like Android and [151:53] Microsoft and this gets into the [151:54] question of what is a platform because [151:56] let's say I make a web app that I run on [151:58] my own website I do my own marketing I [152:00] do my own monetization but I let users [152:03] authenticate Facebook so I can pull some [152:05] information out of their profile it's [152:06] like it's not really built on the [152:08] Facebook platform it's not like the core [152:11] apis that enable my application to run [152:15] are using Facebook's apis I'm using [152:17] Facebook's apis to grab some data the [152:19] core set that allows it to run is the [152:22] browser y it's almost like if you think [152:25] about the intermediation layers Facebook [152:27] was trying to build a platform on top of [152:29] a browser that was a platform on top of [152:30] an operating system that was a platform [152:32] and it the reason why we aren't all [152:34] using Rich Facebook apps all the time [152:36] today and think of it as the default [152:38] platform it was just too many layers of [152:39] abstraction away from the hardware to [152:41] win I think that is totally right but [152:44] for the moment thousands and thousands [152:46] of both Indie devs and Venture back [152:48] companies like Flock to the platform but [152:51] what this does tease out is [152:53] Facebook and now we can talk about this [152:56] 15 plus years later has kind of a weak [153:00] position for launching a platform since [153:02] they don't control the OS or the [153:03] hardware and so they have to make a [153:06] little bit more of an appealing sale to [153:08] a developer and that includes big [153:10] distribution for you but it also [153:12] includes a ton of access to data and I [153:15] remember being an early Facebook [153:16] developer and after user authenticates [153:18] looking at that Json I'm being like okay [153:20] so it's their whole profile oh it's [153:22] their whole friends list wow I get a lot [153:24] of stuff here and Facebook was [153:26] incentivized to do that because they [153:29] almost had to sell harder than other [153:31] platforms who controlled Hardware [153:33] historically would have had to yeah it's [153:35] a super great point now also the reality [153:38] was at the time I think developers were [153:41] getting plenty of value just out of the [153:43] distribution and that was coming from [153:46] Newsfeed if a user starts using your [153:49] Facebook app [153:50] and then the activity that they're doing [153:52] on your app is getting published to the [153:54] news feeds of all their friends oh yeah [153:56] you could build a whole Zinga on that [153:58] interaction Paradigm oh yes you can [154:01] there's this great quote from Josh Elman [154:03] in the Steven Levy book Josh says if [154:05] you're a developer and you can get [154:07] someone to bother 10 friends to get one [154:09] more user to join you're very happy [154:11] because you just got one more user [154:13] Facebook though has nine other people [154:16] who've just gotten bothered and he [154:18] doesn't say but the and are quite [154:20] unhappy about that and that's the [154:21] trade-off they have to weigh that's sort [154:23] of their in the type of platform they [154:25] are trying to build they have set up [154:27] that incentive set and they need to [154:28] figure out what to do with it yep so for [154:32] a year or two that is the state of play [154:35] Facebook brilliantly for them eventually [154:38] pulls back on organic Newsfeed [154:41] distribution for apps says Hey like we [154:44] got to stop this we got to take care of [154:45] the problem you don't need to know every [154:47] time there's a new mafia War move by so [154:49] so showing up in your news feed and that [154:52] does effectively kill a large percentage [154:54] of developers and apps on the platform [154:57] yep but it's the right long-term move [155:00] that was the right thing to do for [155:01] Facebook with a 20-year view absolutely [155:04] and it doesn't kill the platform though [155:08] it actually makes it even more valuable [155:10] to Facebook because for the developers [155:13] and apps who have gotten to scale they [155:16] kind of look at this and they say okay [155:19] my organic distribution just got [155:21] knapped but Facebook is still this [155:25] pretty powerful platform what if I just [155:28] buy ads how well would that work can I [155:31] generate positive Roi if I buy ads to [155:35] effectively accomplish the same thing [155:38] natively on the platform and drive usage [155:41] back to my app turns out that there was [155:43] a very profitable Arbitrage there still [155:45] for many years to come and for Facebook [155:48] they're like well hell yeah okay we're [155:52] monetizing the platform through user [155:54] engagement and ads that we're showing on [155:56] the right rail while people are using [155:59] Facebook apps we are also monetizing [156:04] through currency and commerce that is [156:06] happening in the apps where we're taking [156:08] a cut of that through [156:09] payments now we're going to monetize [156:12] platform in a third way which is the [156:15] biggest scale developers and app [156:19] distributor on the platform are going to [156:22] use our native advertising as their [156:24] primary form of [156:26] distribution I mean you can see why for [156:29] a couple years here Facebook is like [156:32] well this is it we've won you know we [156:34] are a platform this is it yep and then [156:39] mobile comes along oh mobile that is uh [156:44] quite the chapter in the Facebook story [156:47] but before we do that now is a great [156:48] time to tell about one of our favorite [156:50] companies the climate aligned AI [156:53] infrastructure company cruso yes cruso [156:56] is a vertically integrated Cloud [156:58] platform built specifically for AI [157:00] workloads that was recently named the [157:02] gold standard of AI Cloud providers by [157:05] Dylan Patel over at semi analysis and [157:09] something that's new and really cool [157:11] since we started working with cruso last [157:12] year they've opened up the raw metal in [157:15] their data centers to customers so let's [157:18] say you're a large Enterprise who wants [157:19] wants to run your own infrastructure and [157:21] not use a cloud layer you can actually [157:23] now do that with cruso directly as a [157:25] data center customer which several of [157:27] the largest companies in the world are [157:29] now doing yep they've totally reimagined [157:31] traditional data center architecture to [157:33] support the huge Power Cooling and [157:35] compute density needs of AI yep that's [157:38] important because power demand in gpus [157:40] is increasing dramatically which means [157:43] the traditional data center design and [157:44] Engineering of the hyperscalers is no [157:47] longer optimal so kuso's infrastructure [157:49] built from the ground up for gpus with [157:51] elements like highdensity racks direct [157:54] liquid to chip cooling that enables them [157:56] to support the most demanding AI [157:58] workloads that traditional data centers [158:00] just can't handle and at the same time [158:03] as GPU cluster sizes continue to [158:05] increase there's an ever increasing [158:07] demand for energy cruso has 15 gaw in [158:10] its development pipeline which is an [158:12] astronomical amount of power their abene [158:15] Texas facility alone has over 1.2 gaw [158:18] planned which will make it one of the [158:20] largest clusters in the world yep and [158:23] it's also not just about the amount of [158:25] capacity in the pipeline it's about how [158:27] fast it can come online kuso's team has [158:30] Decades of experience constructing and [158:31] operating data centers which enables [158:33] them to bring these new data centers [158:35] online super fast as many of you already [158:37] know cruso sources the energy for these [158:40] data centers in the most efficient and [158:41] climate aligned way in the entire [158:43] industry through clean lowcost and [158:45] abundant energy that otherwise goes to [158:47] waste for example in oil Fields where [158:50] natural gas is flared in congested parts [158:53] of the grid where Renewable Power is [158:54] curtailed or other areas where energy is [158:56] stranded which actually accomplishing [158:59] that is a crazy hard thing kuso's energy [159:02] first approach means they can build data [159:04] centers in some of the most challenging [159:06] locations on Earth bringing Computing to [159:09] the energy rather than the other way [159:11] around yep the net of all of this is [159:13] that cuso can provide nuclear levels of [159:15] power for far less cost than other [159:18] providers and low or in some cases [159:21] actually negative emissions and that's [159:23] super important because the biggest [159:24] bottleneck to AI progress is actually [159:27] energy it's not compute yep so anyway [159:29] they're just a great company we're super [159:30] proud to work with them and to be [159:32] investors so to learn more about cruso [159:34] you can go to [159:35] cruso acquired that's crus so. acquired [159:40] or click the link in the show notes and [159:41] just tell them that Ben and David sent [159:43] you thanks cruo all right so David we're [159:46] here in the era of mobile right this is [159:48] January 2007 Steve Jobs just announced [159:51] the iPhone the whole world changed [159:54] right well actually [159:56] ironically no or at least not for quite [160:00] a while yes January 2007 Steve Jobs [160:03] announces the iPhone people forget [160:05] though it took 5 years for mobile and [160:07] smartphone adoption to really ramp up [160:10] and like actually change the world there [160:11] was no SDK in that first iPhone and iOS [160:14] that didn't come till 2008 it was [160:16] $700 Android was a Blackberry clone at [160:19] the time not an iPhone clone the Droid [160:22] does campaign wouldn't be another year [160:23] or two yep yeah I think two years I [160:26] think that was 2009 if I remember right [160:28] yeah so yeah I mean hell Facebook [160:30] platform didn't even launch until May [160:32] 2007 so after the iPhone announcement [160:35] yep so for the moment platform is [160:39] rocking and rolling like we were just [160:41] talking about Facebook is becoming the [160:42] next Microsoft which brings us to the [160:45] second Microsoft partnership in October [160:47] of 2007 so just 5 months after f8 and [160:51] platform launch Facebook now has like [160:54] all of the leverage Dan Rose has some [160:57] great stories on Twitter about how this [161:00] negotiation goes down oh yeah you spent [161:01] time with Dan didn't you prepping for [161:03] this yep and I chatted with Dan about [161:04] this this is great so Microsoft perhaps [161:08] obvious now in retrospect to everybody [161:10] really just wants to buy the company but [161:12] Mark's of course not going to do that [161:14] after the whole Yahoo experience [161:16] Etc Microsoft is also happy though to [161:18] just keep partnership going and expand [161:20] it because one it is really helping them [161:23] spin up their online services division [161:25] and get really good at online ad sales [161:27] and AD [161:28] serving and two they know they made a [161:31] huge error in missing search and letting [161:33] Google get big and social sure seems [161:36] like it's going to be the next search [161:38] sized platform so they're happy to just [161:40] be in bed with Facebook in a way that [161:42] they never will be with Google as long [161:44] as they can keep it going I mean for [161:46] them I think the thing that shocked them [161:48] about Google is oh my God the business [161:52] model of the web is advertising search [161:55] and browsers and everything are [161:56] monetized by advertising and that's [161:58] completely orthogonal to our traditional [162:00] license-based business model social sure [162:03] looks like where all the page views are [162:05] going and is going to monetize exactly [162:07] the same way as search so actually the [162:09] thing that's important to them is [162:11] locking in those page views to the [162:14] extent that they can participate in this [162:16] market it kind of has to be either we [162:18] own you or we are the long-term ad [162:20] provider which they didn't end up being [162:23] or we're a big Equity holder yep so [162:26] enter this second partnership [162:28] negotiation October 2007 Microsoft comes [162:32] down to Palo Alto Facebook tells them [162:35] that hey Google is actually really [162:37] interested in taking this partnership [162:39] over Larry and Sergey have been banging [162:42] down the door and we've been trying to [162:44] hold them off but they're coming to meet [162:46] with us tomorrow to talk about switching [162:49] our Facebook ad serving partnership over [162:51] to [162:53] Google and this has been great you know [162:55] we love working with yall we want you [162:56] are our preference our preferred partner [162:58] we want to keep it going with you but we [163:00] need to get the deal done [163:03] tonight so they lock themselves all in a [163:06] conference room they start negotiating [163:08] it's getting late 10 o'clock goes by 11 [163:10] o' the Microsoft team you know by [163:13] Facebook standards is all old guys like [163:15] they're getting [163:16] tired God and then you know old guys [163:19] like you know 40 like so old and then at [163:24] midnight this is amazing you can't make [163:26] this up Facebook probably planned this [163:29] all of a sudden this really loud like [163:32] house music starts blaring like with [163:34] like heavy Basse in the office and the [163:36] office is hopping like everybody's there [163:39] the whole company's there it's midnight [163:41] the Microsoft guys is like what's going [163:44] on and uh the Facebook you know SES like [163:47] oh yeah uh we have a hackathon scheduled [163:51] tonight and actually Javier olivan had [163:56] just started at Facebook who's now coo [163:58] who is now the COO of meta had just [164:01] started at Facebook as an engineer and [164:04] uh he had organized this hackathon to [164:06] Begin work on [164:08] internationalizing the site so tonight [164:11] is the beginning of internationalization [164:14] which is a critical critical part of [164:17] Facebook's next chapter of growth yes [164:19] and it is beginning at the same time [164:22] that this Microsoft partnership is [164:23] happening in a conference room and [164:25] didn't this Microsoft deal end up being [164:27] about International yes yes it did so [164:31] they keep going at one point during the [164:33] night one of the Microsoft guys [164:35] according to Dan looks at him and says [164:37] wow this is awesome this is just like [164:40] the old days of [164:42] Microsoft so they finished the [164:44] negotiation at 6:00 a.m. they announced [164:48] the new partnership that very same day [164:50] as promised with the thread of Google [164:53] Microsoft becomes the worldwide [164:56] exclusive thirdparty advertising Network [164:59] for all display ads on Facebook so it [165:01] was domestic in the US before now it's [165:03] International too and domestic [165:06] everywhere around the world Facebook of [165:08] course can still also sell their own [165:10] inventory themselves but anything that [165:12] they don't want to sell themselves or [165:14] that is excess inventory Microsoft gets [165:16] exclusive access to be the only ad [165:18] Network [165:19] where you can access it and the big [165:22] kicker Microsoft is going to invest 240 [165:27] million in Facebook at a [165:29] $15 [165:31] billion valuation this is like 13 or 14 [165:35] months after the failed Yahoo deal for [165:38] $1 billion that is the thing that no one [165:40] talks about here I'm so glad you brought [165:41] this up because before they started this [165:43] negotiation the attempt was to buy [165:45] Facebook the highest offer floated we [165:48] talked about the in the Microsoft [165:49] episode it was a complicated set of deal [165:51] terms that basically nett it out to a [165:52] $24 billion offer and that was less than [165:56] two years after the1 billion doll offer [165:59] that everyone talks about with Yahoo oh [166:02] remember the time where Mark Zuckerberg [166:03] turned down a billion dollars less than [166:05] two years later he turned down 24 [166:07] billion and then they were really good [166:08] for the money it's [166:10] Microsoft but turned it around into an [166:12] investment at 15 billion yeah it's crazy [166:15] so the great thing is like all great [166:18] part [166:19] Partnerships everyone makes out [166:22] wonderfully here so except maybe the [166:24] venture capitalist in Silicon Valley [166:26] because this just like breaks everyone's [166:28] brain it was the high watermark for [166:31] Facebook's valuation for a while because [166:32] then the great financial crisis would [166:34] happen after this valuations would all [166:36] reset and then Facebook's next deal [166:38] would get done I think With Yuri mner at [166:40] 10 billion1 billion valuation With Yuri [166:44] yep and Yuri would actually also buy [166:46] common stock through a secondary tender [166:48] off offer in that so his dollar cost [166:51] average would get down to more like six [166:52] so he got a oh wow screaming deal but [166:55] you know for the moment like there are [166:56] no comps to this ever do com bubble you [166:59] name it like never has been a comp like [167:02] this for an investment in a private [167:04] technology start up the closest comp [167:06] that I can even sort of think of was [167:09] Google's market cap at IPO was 23 [167:13] billion so like higher but not that much [167:15] higher and Google at IPO was a $2 [167:19] billion Revenue run rate spitting off [167:21] hundreds of millions of dollars in free [167:23] cash flow annually Facebook at this [167:26] point in time is basically break even [167:28] and doing 150 million in Revenue most of [167:30] which is Microsoft [167:34] itself all that said Microsoft this is a [167:37] home freaking run like we've been saying [167:40] all along helps them spin up Bing doing [167:43] all the ad serving for this getting in [167:45] like real online services and then B you [167:48] know Microsoft is is not a hedge fund as [167:49] we talked about on that episode but that [167:52] $240 million investment by the time they [167:55] start selling it down years after [167:57] Facebook is public ends up being worth 8 [168:01] billion wow and then maybe the most [168:04] important benefit to Microsoft out of [168:06] all of this is like Microsoft and [168:08] Facebook have always had a good [168:10] relationship unlike Microsoft has with [168:12] Google and apple I remember Facebook [168:15] events like when you look at the page it [168:18] was a Bing Map not a Google map that [168:20] always like felt emblematic of the [168:22] partnership to me yeah every single [168:24] place it could be a Microsoft product it [168:26] was yes and it wasn't a Google product [168:28] yep okay so interestingly here we are [168:31] 2007 Facebook is still saying we don't [168:35] need to be in control of the core [168:37] Revenue creator for us like we think [168:39] platform is the future Microsoft you're [168:42] our preferred partner to handle making [168:44] the money yes and I think you know [168:47] there's two things here [168:49] yes I think that is true specifically at [168:52] this moment in time but this is where [168:54] you know again Mark is Just Master [168:56] strategist you know kind of like calling [168:58] it all the way back to civilization in [169:00] the 4X strategy game he's got multiple [169:02] bets he's placing on the chessboard one [169:05] as you said is like platform like hey [169:08] platform is rocking and rolling we think [169:09] we're building the next great technology [169:12] company and the next great business [169:13] model just on [169:15] platform also yeah we're a social media [169:20] company that involves a lot of page [169:21] views and engagement and like the right [169:23] way to monetize that is advertising so [169:25] yes we've just continued this big [169:28] Microsoft partnership but we have the [169:30] right to sell our own inventory and we [169:32] should probably start building that [169:34] muscle too the other thing that's [169:36] happening here is Mark still kind of [169:39] thought ads were gross at this point in [169:42] history so the things he cared about [169:44] were product and engineering and design [169:46] and he kind of wanted to Outsource [169:48] everything else or at least have someone [169:51] else at the company think about it and [169:53] having Microsoft take care of the ads [169:55] was I think in his mind kind of a [169:57] win-win that way the commodity stuff can [169:59] happen elsewhere and I can focus on the [170:01] thing that really matters and the [170:03] company really did not have the Insight [170:05] yet oh we can do ads different and [170:08] better than anyone has ever done them [170:11] before well I'm smiling as you're saying [170:13] that because um in practice for this [170:17] moment yeah in reality they were trying [170:21] to build their own native Facebook [170:25] advertising unit so to speak it was just [170:30] Beacon yeah Beacon is one of the most [170:32] epic failures in Facebook's whole [170:35] history yeah okay so Beacon well first [170:39] off you were totally right and I think [170:41] you were foreshadowing Cheryl Samberg [170:43] there of like Mark wants somebody within [170:45] the company to just you know run and [170:48] manage this thing and build it great and [170:49] world class before Cheryl do you know [170:54] which Facebook executive LED Beacon I do [170:58] not chamath [171:00] poopaa no way I he always talks about [171:04] the growth team I never knew he was in [171:05] charge of Beacon first this was Jam's [171:09] big thing before growth and actually the [171:11] epic failure here of Beacon leads [171:13] directly to growth as we shall see in a [171:15] sec now what was Beacon [171:19] so like you said Mark is kind of [171:22] allergic to traditional [171:24] advertising but he's like well we have [171:26] this [171:27] incredible social activity happening on [171:30] Facebook and we know through Newsfeed [171:32] that people engage with and they and [171:33] they love it what if there were a way [171:36] for a native ad baked into the very [171:39] fabric of the product itself that Brands [171:43] could sort of control and monetize well [171:46] if we gave BRS a way to essentially [171:50] boost how people are already engaging [171:53] with them that might be the way to do [171:57] this and what this ends up being in [172:00] practice is Brands publishing your [172:03] friends Ecommerce activity into your [172:06] newsfeed now to be fair to Mark and chth [172:12] and the company we didn't know yet what [172:15] the right advertising unit on Facebook [172:17] was going to be and this is probably as [172:20] good an idea as any because like the [172:22] core thing that you do on Facebook is [172:24] you engage with your friends and you [172:25] engage with their stories on Newsfeed [172:28] there's kind of no evidence really that [172:30] people would want to do anything else [172:32] but that and so you're trying to kind of [172:34] think of like how do we shoehorn [172:35] advertising in Brands into this I [172:39] actually I don't buy it at all the [172:42] obvious thing is show an image that [172:45] people can click on and take them to a [172:47] website it's display ads and like I [172:51] don't understand why they needed to try [172:54] to way overthink it and say our ad [172:56] format has to be something that no one's [172:59] ever thought of before like just offer [173:01] advertisers to do the thing that they [173:02] know how to [173:04] buy well I was trying to like think of [173:06] how to be the most charitable possible [173:09] to the company and how they can come up [173:11] with it but yeah I have next in my notes [173:13] here to be clear this was a truly truly [173:18] terrible head up your rear end idea [173:22] these are actually two different things [173:24] I think there was Beacon which was [173:27] JavaScript that an Advertiser could [173:29] embed on their website on an e-commerce [173:31] providers website that would do exactly [173:32] what you're talking about publish into [173:34] the news feed purchases that people were [173:36] making and like this was at the point [173:39] where you wanted to publish all sorts of [173:42] interesting different things on social [173:43] media I don't think we knew for sure [173:45] that purchases wasn't going to be one of [173:47] them like people still keep their venmo [173:48] history public you don't know what [173:49] people are going to do the second thing [173:52] that I think is in your description [173:53] there a little bit is the idea of social [173:55] ads yep that Brands could take [173:59] interactions that happen on pages and [174:03] boost them but they couldn't just take [174:05] out a regular old ad yep so it was this [174:08] weird thing where like you could only [174:10] advertise as a brand to people where [174:14] someone in their Network had interacted [174:16] with your brand page [174:19] yeah pretty convoluted I get that it's [174:21] like this super natively social thing [174:24] but again it did feel like they started [174:26] in this place where like they let let's [174:28] get way too clever for our own good [174:29] first and then work backward to the most [174:31] basic ad unit yep totally agree that [174:35] said like as soon as it's live it [174:37] becomes clear like one nobody wants this [174:40] either of these things yep users don't [174:43] want to users are confused why they're [174:45] seeing I mean certainly there's instant [174:47] blowback against they're publishing my [174:49] purchases someone I think had a [174:52] engagement spoiled by a diamond ring [174:55] getting broadcast yes oh yeah I was [174:57] going to say I mean this is number two [174:59] like the Privacy implications here [175:00] actually are like horrific yes like [175:03] spoiling engagements like that's the [175:04] wholesome horrific privacy things like [175:07] you can imagine the nonh wholesome [175:10] publishing stuff that you're buying [175:11] versions of this Y and then of course [175:14] advertisers are confused and people [175:16] don't really understand what's an ad [175:18] what's it's not an ad they just burned [175:19] credibility Everywhere by launching both [175:21] of these things yep totally and Mark and [175:25] the company is coming off the news feed [175:27] experience where they're like yeah if we [175:29] just give it enough time users will get [175:31] used to it so they let it run for a [175:33] couple weeks they run the newsfeed [175:35] Playbook of like yeah yeah yeah we hear [175:37] you you know but you'll get used to it [175:39] here's some more controls oh the social [175:42] ads was years and years but the [175:44] purchases you're right that was just I [175:45] think a couple weeks yeah for Beacon [175:47] ultimately after a few weeks they do [175:50] acqu and they say like okay we will turn [175:53] Beacon off we'll make it opt in as [175:55] opposed to opt out and you can have a [175:58] way to like completely turn off all [176:01] Beacon tracking permanently from your [176:03] profile a lot of people do that Beacon [176:06] activity basically drops to zero and [176:08] then yeah two years later it's like [176:10] completely killed off and this sets the [176:12] stage for Cheryl joining the company [176:15] because I think you know Mark's takea [176:17] away from this is like [176:19] okay maybe I don't understand you know [176:22] advertising as well as I thought I did [176:24] like I thought I didn't like it but I [176:26] thought I could be clever and engineer [176:28] this new thing and like actually let's [176:30] bring in somebody who like really really [176:32] does understand this and Cheryl was the [176:36] perfect person coming from Google Cheryl [176:38] joined Google in 2001 right as they were [176:41] figuring out AdWords I mean that's like [176:43] three years after the company was [176:44] founded totally well and and the company [176:46] didn't figure out AdWords for a couple C [176:48] years so it was like that was like the [176:49] key moment and then Cheryl built and ran [176:52] the whole self-served business at Google [176:55] which was the most important and the [176:56] most technology like enabled part of [176:58] what they were doing she had built the [177:01] greatest digital advertising system in [177:04] the world and Mark was like oh I'm [177:08] trying to build the greatest digital [177:09] advertising system in the world who [177:10] could I ever get to do that with me and [177:13] wouldn't it be great if that person was [177:15] also a good like manager and leader who [177:17] could teach me how to manage and Lead [177:20] yes yes and I think that really is the [177:24] quote unquote deal they make and they [177:26] say and you know Cheryl agrees too it's [177:27] not like hey Mark said like you're going [177:29] to take this side of the company and I'm [177:30] going to take that side of the company [177:32] she always reported to him like the [177:34] whole company was his company but Mark [177:37] is really focused on product engineering [177:39] and importantly platform Cheryl takes [177:41] ads Sherl takes operations Etc and when [177:45] she comes on board like you referenced [177:47] this earlier been there's basically no [177:51] ad targeting that's happening natively [177:54] on Facebook like there probably is [177:56] targeting that's happening through the [177:58] third-party Microsoft advertising [178:00] network but none of that is like [178:02] technology that Facebook's building or [178:04] value that's accr to Facebook and so the [178:07] first thing Mark and Cheryl did when she [178:09] joined was like okay Beacon was such a [178:13] disaster we've got this great [178:15] partnership with Microsoft like let's [178:16] just do a full exploration should we be [178:18] in the ads business at all yeah she [178:21] definitely famously kind of led the [178:24] exploration of what business are we [178:25] actually in here yep and you know [178:29] ultimately they decide yes we are in a [178:32] media business thus we should be in the [178:35] advertising business and we should do it [178:37] right and we should make it work and [178:39] like ultimately that is going to be [178:41] building a really really really great [178:43] targeting engine which is the thing that [178:45] Facebook is like uniquely capable of [178:47] doing yep so though back to here where [178:50] we are in 2008 two you know seemingly [178:54] unrelated but about to get very related [178:56] things are going on one after the beacon [178:59] Fiasco when Cheryl joined chamath kind [179:02] of needs a new job at the [179:04] company and you know like total credit [179:07] to Mark to Cheryl to Facebook as a [179:09] company I think at most places like [179:11] jamath would have been fired immediately [179:13] for what happened with Beacon but the [179:16] culture at Facebook is like nope we move [179:18] we break things and like we're always on [179:20] offense and this didn't work and let's [179:23] find the right thing for you to do and [179:26] thank goodness because that brings us to [179:27] number two right around this time once [179:31] again user growth slows at Facebook yeah [179:35] so it's funny that I don't feel like in [179:37] the public eye we really knew these [179:40] things but as you kind of dig into the [179:41] company history you're like oh there are [179:43] these moments in time where growth [179:44] really did stall out and you know chth [179:47] has said on stage since then oh I [179:49] totally thought that growth was done and [179:52] there's a few things to note about the [179:54] growth team one is it was effectively [179:56] the first growth team yes I think it was [179:59] the first growth team yeah everyone who [180:01] wants to get into growth now I don't [180:03] think really fully grasps that is a [180:05] brand new discipline invented in 2008 [180:08] and the way they sort of defined it at [180:10] Facebook was there's marketing there's [180:13] product but marketing and product both [180:16] touch customers in different ways and so [180:19] the left hand needs to talk to the right [180:21] hand so for example you should not have [180:23] a marketing team that is sending out [180:25] emails through an email marketing system [180:27] and a product team that is sending out [180:29] notifications to users through the [180:31] product with no notion of how they're [180:33] speaking to each other you should a [180:36] unify those efforts through one team and [180:38] B that team should live within product [180:41] or at least that team should be tightly [180:43] coupled to product with the general [180:46] belief or um reason for being that your [180:50] product is the best lever to grow no [180:54] amount of marketing you could ever throw [180:55] at something that is not integrated into [180:57] your product will be as effective as [181:00] your product doing a good job with [181:03] features hitting the right users with [181:05] the right message and the right value at [181:07] the right time in a native way to the [181:10] core feature set of the product and so [181:13] you've got shth you've got Alex Schultz [181:16] you've got Naomi glight and Javier [181:18] olivan and so you've got this core early [181:21] team it's four people and then it [181:22] expands to kind of six or seven folks [181:25] that are formed really on this agreement [181:28] between chamath and Mark of we're going [181:31] to have this dedicated growth team and [181:33] our mission is going to be to grow [181:35] Facebook using Facebook itself not [181:37] through extrinsic measures yep and this [181:41] was like an [181:43] incredible insight and was so perfect [181:46] for Facebook like I I think this growth [181:49] Insight of growth through product can [181:50] apply to any company but it especially [181:53] applies to a virally growing social [181:55] network like Facebook yes there's this [181:58] interesting things that make it tick the [182:00] first of which is this has to be the [182:03] most data obsessed team in the company [182:05] every team should be data obsessed blah [182:07] blah blah like the growth team is really [182:08] the one who sort of discovers oh here [182:12] are the obvious places where users are [182:14] deriving value here are the obvious [182:16] places where users are getting confused [182:18] here are the obvious opportunities to [182:20] find new users here are the obvious [182:21] opportunities to reduce friction [182:23] analytics is the answer for that and I [182:26] think that there's a lot of really [182:27] interesting stories especially around [182:29] internationalization of the growth team [182:33] and partnering with other product teams [182:35] around the company to say what is an [182:38] engineering and product approach to [182:40] something that traditionally has been [182:42] done other ways like the way most people [182:45] would translate their product is by [182:49] hiring [182:50] translation servicing firms or by hiring [182:53] a dedicated person to go through string [182:55] by string and edit well we're Facebook [182:58] we have all these users what if we just [183:00] you know when we're launching in Spain [183:03] surface different words to Spanish [183:06] speakers and say okay is this the right [183:08] word for Spanish or not hey can you [183:09] translate this and crowdsource it and [183:11] double check it with everyone and that [183:13] way you can not just translate five or [183:16] 10 of the top languages in the world you [183:17] can be in every language all the time up [183:20] to date using Facebook itself cuz oh by [183:23] the way when you translate the product [183:26] way more people can use it and so [183:28] translation itself internationalization [183:30] is a growth lever and we have product [183:32] ways to do the translation with our [183:34] users yep and that's what actually [183:36] started even before the official growth [183:38] team with with jav in that first [183:39] hackathon yes but that was the seed of [183:42] this idea of like oh actually the way to [183:44] get past this growth wall as people [183:47] refer to it in retrospect that we're [183:49] hitting is use the product itself and [183:51] internationalization there's even more [183:53] to it than what you said one when you [183:57] have the users themselves who are [184:00] translating a product as you're seeding [184:02] it in these new [184:04] markets they're so much more bought in [184:07] and like feel a sense of ownership over [184:09] it versus like oh I'm just air dropping [184:12] this you know translated American [184:13] product onto your country here so it's [184:16] not just that it's a a better [184:18] translation it's that like you're [184:20] actually seeding those first Power users [184:22] who are going to be deeply engaged and [184:24] feel ownership over the product yeah and [184:28] then two the Nuance in like the local [184:32] markets that might even be in the same [184:34] language I mean Spanish is the perfect [184:36] example Spanish is spoken in so many [184:38] places around the world but like the [184:40] local nuances and needs in Spain are [184:43] like super different than Mexico or [184:46] Argentina or Latin America ET [184:48] that's exactly right and I think my [184:50] point with all this is a growth is a [184:52] pretty new discipline in our industry [184:55] two growth is not marketing it's very [184:58] tempting you see us especially with [184:59] incumbents like big Fortune 500 [185:01] companies who have someone whose title [185:02] is growth and then you ask them what [185:03] they do and you're like oh no that's not [185:05] growth that's marketing you don't [185:06] actually modify the product at all so [185:09] that's a different thing and it's [185:11] important to realize like not only did [185:12] Facebook kind of invent the discipline [185:15] they are perhaps still the best at it [185:18] they really eat sleep and breathe the [185:19] idea that growth comes from product y it [185:22] was that core initial Team all of them [185:25] of which Javi jav's now the COO Alex [185:28] Alex is now the CMO of meta Naomi Naomi [185:31] is now meta's longest tenured employee [185:33] besides Mark and runs a ton of stuff at [185:35] the company and chamath who deserves so [185:37] much credit for starting the team [185:40] pulling it together protecting it [185:42] advocating for it within the rest of [185:44] product you in the way that only chamath [185:47] can man it really is crazy with the [185:49] Facebook diaspora we're going to keep [185:51] seeing this as we go along there's other [185:52] names where people will be like whoa I [185:53] didn't realize they were at Facebook but [185:55] many of the names that have come up so [185:56] far yeah the diaspora is pretty talented [185:59] yep so there were a million things that [186:01] growth did over the years within you [186:04] know the Blue app and then ultimately [186:06] within all the apps at Facebook growth [186:08] today is a core Center of Excellence [186:10] discipline that spreads across all the [186:12] products at the company at meta but the [186:14] other big thing in the early days [186:15] besides internationalization is people [186:18] you may know which crazy like the [186:20] Carousel of people you may know on [186:22] Facebook is still a core part of the [186:24] Blue app to this day dude I just got [186:26] lost in it did you know that that's what [186:28] I was looking at right now no I was [186:31] sanity checking I was like oh is that [186:33] person part of the original growth team [186:34] and I opened Facebook the news feed [186:36] distracted me I scrolled down there were [186:38] stories at the top I scrolled down one [186:40] post below that first post was an ad [186:42] below that is people you may know and I [186:44] am like three panels over in people you [186:46] may know cuz I like oh my gosh I'm not [186:49] friends with them we're doing a podcast [186:51] Together live and I just had that [186:54] experience totally people you may know [186:57] become such an important lever [187:03] and I love it I love [187:05] it okay so people you may know become [187:08] such an important lover yes for a whole [187:11] bunch of reasons the growth team [187:13] discovers you were talking about [187:14] analytics and data and really [187:15] understanding what's happening [187:18] the key thing that they discover for New [187:19] Year is joining Facebook about whether [187:22] they become an Engaged active you know [187:25] evangelizing user or not is how many [187:28] friends they get in the first like I [187:30] think it's 24 hours or 48 hours or like [187:33] short period that they're joining the [187:35] network and there's like a certain [187:36] number of active friends that you need [187:40] oh this is funny by the way I watched [187:42] two different Talks by people on the [187:43] early growth team and one of them cited [187:46] 10 friends in 14 days the other cited [187:48] seven friends in 10 days of you know how [187:51] do you create the Magic Moment Alex [187:53] Schultz who's now the CMO gives this [187:55] great talk gosh a decade ago at startup [187:57] school with Sam Altman and uh he makes [187:59] the point look it's kind of a linear [188:01] thing yes you want as many friends as [188:03] possible in the least amount of time [188:04] it's not like oh magically at 10 friends [188:06] in 14 days it's like super different [188:08] than nine friends in 14 days but you [188:10] just said a threshold somewhere and you [188:12] set the threshold and then you're like [188:13] okay if we can deliver this delightful [188:16] experience where now people have a rich [188:18] news feed and the people they care the [188:20] most about in the world to interact with [188:21] they're going to retain yep totally the [188:24] other thing is it marries perfectly with [188:27] all of the investment that is going into [188:29] the algorithmization of Newsfeed because [188:34] choosing who to show you in that [188:37] Carousel of people you may know is like [188:40] all the difference in the world you know [188:42] show you seven randos and like you're [188:45] going to churn show you your seven best [188:47] friends and you're going to friend them [188:48] all and become an Engaged user I'm so [188:51] excited for I guarantee you at least one [188:54] of the people that I just friended is [188:55] going to be listening to this podcast [188:57] and realize what just happened because I [188:59] haven't friended anyone on Facebook in [189:01] years and and I just sent out like three [189:04] or four friend request they just [189:05] re-engaged you as a user I love it yes [189:08] they just have to New Growth tactic have [189:09] people do podcasts about the company and [189:13] then I think people caught on to that [189:15] one a while ago yes but yeah like that's [189:17] a super hard data problem of like cold [189:21] new user coming into the network how are [189:24] you going to accurately predict who [189:26] their best friends are who are already [189:28] on Facebook well if you're on the web [189:31] and this is the late 2000s you ask them [189:34] to authenticate their web-based mail [189:36] service of choice and then find out who [189:39] they've been emailing and then use that [189:41] to figure out who their friends are [189:43] magic magic oh you know there's other [189:45] less nefarious ways you know see who has [189:48] sent them links in the past who are also [189:51] logged in Facebook users there's all [189:52] sorts of stuff you can do but yes so by [189:55] the end of 2008 thanks to the growth [189:57] team growth is like really going again [190:01] they go from like just barely getting to [190:04] 100 million users in August of 2008 to [190:07] almost 150 million by the end of the [190:09] year so 50% growth in the last four [190:12] months of the year and then 2009 is just [190:15] like lights out they grow [190:19] 250% in 2009 to 350 million users [190:24] globally internationalization is like [190:26] really firing on all cylinders here [190:29] that's right they went from 150 million [190:31] users to 350 million no in like one year [190:35] isn't that wild crazy the other core [190:38] piece of the stat whenever you're [190:39] looking at growth is to look at [190:40] engagement engagement in 2008 was also [190:43] in the doldrums 50% of monthly actives [190:46] were actives and from what I can tell [190:49] that basically was just an artifact of [190:52] as the company got bigger every marginal [190:54] user they added was sort of less engaged [190:57] than the early core users you know when [190:59] they went from colleges to high schools [191:00] to open registration it it just was [191:03] going to have slightly less product [191:04] Market fit but growth team is focused [191:07] not just on growing new users actually [191:10] an even better lever every long-term [191:12] goal is retaining your existing users [191:14] and the best lever for retention is [191:15] engagement and so that was where a huge [191:17] amount of their energy went so this is [191:19] really interesting 50% of monthly [191:22] actives were daily actives in ' 08 since [191:25] then they have basically improved that [191:27] metric every year there's been a little [191:29] bit of variance but it is now almost 70% [191:32] today been a maniacal focus on how do we [191:36] make as many monthly actives daily [191:38] actives as possible year over year over [191:41] year over year that's wild and that [191:43] figure is just in the Blue app in [191:45] Facebook that is just in the Blue app [191:47] app yep exactly wow well on the [191:50] engagement front in [191:52] 2009 the big thing that happens there is [191:55] like button you know where I'm going [191:56] with this like button Y and I know you [191:59] know who invented the like [192:02] button cuz it wasn't [192:04] Facebook uh I don't know friend feed oh [192:09] no way Brett Taylor and Paul buite [192:13] invented the like button how did that [192:15] work cuz friend feed would just Federate [192:17] stories from other social networks into [192:20] one aggregated feed yep I don't recall [192:24] exactly did it only store that [192:26] information on friend feed and not [192:27] propagate it back I suspect that's [192:29] probably it like they probably felt like [192:31] they needed some sort of like hey native [192:33] mechanic on friend feed man friend feed [192:36] was so awesome it's funny in this era [192:39] where I'm now checking multiple feeds [192:40] every day Twitter threads Instagram all [192:45] the messaging services to catch up on my [192:47] messages occasionally the Blue app I [192:49] kind of need friend feed again that was [192:51] like the most amazing product I know I [192:53] know uh the world is so bulcan ised [192:56] it'll never happen but yeah I mean talk [192:58] about another amazing part of the [193:00] Facebook diaspora Facebook would later [193:03] acquir friend feed Brett Taylor would [193:05] become CTO when Adam Dantel left to go [193:08] start Kora and then when Facebook [193:11] acquired friend feed Brett became CTO [193:13] and then famously Brett becomes the [193:16] co-president resident of Salesforce by [193:18] way of acquisition and then eventually [193:21] board share at Twitter when the whole X [193:24] thing was going on and now he's on the [193:26] board at open AI yep board chair I think [193:28] at open AI crazy and he's got a new [193:30] startup of his own so I'm telling you [193:32] the talent that moved through that place [193:34] totally well and Paul buite the other [193:37] friendfeed founder invented Gmail [193:39] created Gmail within Google yeah so he [193:41] ended up at Facebook for a while then [193:43] after I forgot about Paul too that's [193:45] unbelievable yeah he must not have [193:46] stayed very long if at all yeah cuz he [193:48] went to YC I think after that okay so [193:50] friendfeed invents the like button [193:54] Facebook I believe was going to call it [193:57] the awesome button and then at some [194:00] point right before launching I think [194:02] Mark personally made the call it's just [194:04] too weird let's go with [194:06] like yes the story as I understand it is [194:09] that Facebook had been considering [194:12] something like this for a long time I [194:14] mean the benefits are obvious right like [194:18] if you have a lightweight way to engage [194:21] and feed more data about vote positive [194:25] votes on posts into the newsfeed [194:27] algorithm like that's going to really [194:29] turbocharge things on the flip side the [194:32] concern was that it would actually [194:33] decrease meaningful engagement there [194:35] would be less comments exactly and so [194:37] ultimately they started running some [194:38] tests with like small user groups and [194:41] specific geographies and they found that [194:43] it actually often boosted engagement of [194:45] comments fascinating that's why you test [194:47] things you never know exactly so yes [194:51] they launch the like button and that [194:53] becomes huge for overall engagement but [194:56] also more importantly for just improving [194:59] the algorithm which is now starting to [195:01] kick in in Newsfeed and making sure [195:02] you're seeing stories that you really [195:04] care about and the like button then also [195:06] got used for Pages it became this sort [195:10] of form of capital among brands of how [195:12] many Facebook likes does your company [195:16] have which for a while actually meant [195:18] you could [195:19] organically get messages out to them [195:22] every time you posted it was as if a [195:23] company was a person and you just show [195:25] up in the feed before Facebook [195:26] ultimately was like H those are [195:28] advertisements you're going to pay us [195:29] for those yes you're exactly on the [195:32] right track let's flip back now to [195:35] Cheryl and building the advertising [195:37] business right around the same time [195:39] we've decided like okay yes we're going [195:41] to be in the advertising business and to [195:44] do that right to do that natively to do [195:47] that in a way that generates defensible [195:49] sustainable value for Facebook that [195:51] means targeting and the like button is [195:54] the perfect gift to this because all of [195:57] a sudden there's this really lightweight [195:59] way to engage with things to Signal your [196:02] preferences about what you like yep and [196:06] the like button ended up being the [196:08] perfect Trojan Horse to move Facebook [196:12] platform off of Facebook so there was a [196:15] reason for [196:17] thirdparty websites to embed Facebook's [196:20] JavaScript in their Pages because of [196:23] course you'd want to be able to like an [196:25] article or like a brand all up on their [196:28] website you know how many Facebook likes [196:30] do you have if that's what matters we [196:32] want people click and like on our [196:33] website too not just over on Facebook [196:35] and so suddenly now every company on [196:38] Earth has some Facebook and has a reason [196:40] to embed some Facebook JavaScript right [196:42] there on their page which my goodness [196:44] that's going to serve as a great signal [196:47] back to the advertising algorithm where [196:49] we can drop cookies and we can see who [196:51] is moving around the web in what ways so [196:54] it's perfect for platform but it's also [196:56] perfect for feeding data into their [196:59] advertising system now that they have [197:01] something that Brands and thirdparty [197:04] websites are incentivized to drop right [197:06] on their page yes ABS freaking lutely so [197:10] we talked about growth reigniting at the [197:13] end of 2008 into [197:15] 2009 the business is also on fire in [197:19] 2009 in a good way so platform still [197:22] rocket and rolling Microsoft display ad [197:25] partnership still growing strong as [197:27] growth is happening now under Cheryl you [197:30] finally got real native highly targeted [197:34] ads coming online in Facebook in 2008 [197:38] the company had done 280 million in [197:40] Revenue up from 153 the year before in [197:44] 2007 2009 they do almost 800 million in [197:49] Revenue he's crushing it [197:52] 2010 an even 2 billion in Revenue [197:57] they're now very very cash flow positive [198:00] yes reaching the scale that Google was [198:03] at their IPO 2010 they hit 600 million [198:07] users 2011 850 million users 2011 [198:12] Revenue grows almost another 100% to 3.7 [198:16] billion [198:17] it's like all right time to start [198:19] thinking about taking this company [198:20] public it's also time to start thinking [198:22] about what is our biggest existential [198:24] threat it's so funny that we're talking [198:26] about and the existential threats are [198:29] over we've won we're we we have platform [198:32] going well we have this advertising [198:34] business going well we're getting data [198:36] from the open web because we have like [198:39] buttons everywhere everything is going [198:42] our way we finally have it dialed and [198:45] none of that matters in a mobile world [198:47] as people shift from spending time on [198:50] the web to other apps that open web data [198:53] used as signal goes away all of your [198:56] ability to take payments by the way [198:58] Facebook had launched payments at some [199:01] point they started requiring apps to use [199:04] their virtual currency Facebook credits [199:07] Facebook credits that's right all of [199:09] these amazing pillars that they've built [199:11] were for the open web and created the [199:13] most incredible business known demand [199:16] and Mobile's account completely [199:17] different thing yep that seed from 5 [199:21] years ago that has been growing a mobile [199:23] is finally going to turn into a big big [199:26] big problem for the company all right so [199:29] let's just get into it I thought it [199:31] would be fun I listed out all of the [199:34] ways that I could think of that mobile [199:37] was a big problem for Facebook one and I [199:40] think this is kind of like the Uber [199:42] problem that everything else is [199:43] Downstream of IOS and Android are closed [199:48] proprietary ecosystems not controlled by [199:51] Facebook yeah you can argue Android is [199:53] open source and all that but for all [199:54] intents and purposes Google controls it [199:57] and they control the Play Store put [199:59] another way the web is the only open [200:02] platform in history yes and Facebook was [200:07] born on the web I mean how crazy is it [200:09] that they could build Facebook on [200:11] entirely free Technologies at the [200:13] beginning and then get distribution just [200:16] by people sharing URLs around browsers [200:18] are interchangeable operating systems [200:20] are interchangeable it works on any [200:22] device that anyone wants yeah if you [200:25] live on the web you have infinite [200:27] degrees of freedom and flexibility to [200:28] control your own destiny yeah and [200:30] Facebook had kind of built this you know [200:32] wal Garden on the web and it was pretty [200:35] nice yep and wal Gardens are great as [200:37] long as the foundation under you can't [200:39] shift yes build your wall Garden [200:41] directly on the earth not on someone [200:43] else's [200:44] Foundation but okay spefic specifically [200:47] why is IOS and Android being closed [200:49] proprietary ecosystem such a problem [200:51] that means my number two reason platform [200:55] which is you know 50% or more of the [200:58] value of Facebook in their minds at this [201:00] point in [201:01] time no longer works at all like full [201:06] stop flip a switch does not work will [201:09] never work at all ever again on these [201:11] platforms and why is that the whole [201:14] premise of platform is you can run apps [201:17] inside of Facebook on mobile you will [201:20] never be able to run an app inside [201:22] another app just not going to happen [201:24] Apple's never going to let it happen and [201:26] Google's never going to let it happen [201:27] and that's primarily because of that [201:29] thing around you can't launch an [201:31] alternative app store right yeah I mean [201:33] it's the core part of their business [201:34] model are the app stores not to mention [201:37] if you are paying for things on their [201:39] phone you have to use their payment [201:41] system so Facebook credits ain't going [201:43] to work because you can't charge 30% on [201:45] top of 30% yeah and there's technical [201:48] aspects to this too right like code runs [201:50] in native code on these devices as [201:53] opposed to open HTML on the web right [201:56] you get told what your development [201:57] environment is this is the language [201:59] these are the Frameworks these are the [202:00] apis you're allowed to call yep so okay [202:03] greater than 50% of you know the revenue [202:06] and the sort of like theoretical future [202:08] value inherent theoretical value and [202:11] strategy of the business [202:13] kneecapped it gets worse Facebook and [202:16] Mark and Cheryl have been hard at work [202:19] building the other core pillar of the [202:22] business this new super sexy targeted [202:26] ads business well guess what all of the [202:29] ads that are running on Facebook at this [202:31] point in time are running on the [202:33] right-hand column on the [202:35] website there is no right-hand column on [202:38] a mobile app there's only room for a [202:41] feed which to this point in time has [202:43] never had advertising in it and in fact [202:45] there's like a cultural allergy to the [202:49] idea of polluting the beautiful pristine [202:52] organic Newsfeed with an ad it require [202:54] big cultural change this is why you [202:56] can't make up this stuff in these [202:57] stories that we tell Facebook has I kid [203:01] you not no way to make money at all on [203:06] mobile here is an actual honest to God [203:09] statement that is in Facebook's S1 that [203:12] they filed before the IPO in 2012 I am [203:15] reading as a quote we had more than 425 [203:19] million monthly active users who used [203:21] Facebook mobile products in December [203:23] 2011 we anticipate that the rate of [203:26] growth in mobile users will continue to [203:28] exceed the growth rate of our overall [203:30] Maus for the foreseeable future and our [203:33] users could decide to increasingly [203:35] access our products primarily through [203:36] mobile devices which let's be clear [203:38] comparing those growth rates another way [203:40] to simplify that and say it is our user [203:42] base is Shifting to mobile from desktop [203:45] another way to simplify that is our [203:47] desktop business is going to zero yes [203:50] now the kicker this is a statement in [203:52] the S1 we do not currently directly [203:55] generate any meaningful revenue from the [203:57] use of Facebook mobile products and our [203:59] ability to do so successfully is [204:02] unproven accordingly buy our stock yeah [204:06] accordingly if users continue to [204:09] increasingly access Facebook mobile [204:11] products as a substitute for Access [204:13] through personal computers and if we are [204:16] unable to successfully Implement [204:18] monetization strategies for our mobile [204:20] users our revenue and financial results [204:22] may be negatively impacted yeah no [204:25] [Laughter] [204:28] kidding oh my God uh I'm still not done [204:31] that's only reason number three okay [204:33] reason number [204:35] four another benefit of the open web is [204:39] that you can push code whenever you want [204:41] you can make changes rapidly you can [204:43] move fast you can break things and then [204:45] you can fix them [204:46] rapidly this is how Facebook has always [204:49] operated on mobile apps you can push [204:53] code when Apple and Google tell you that [204:55] you can push code and at this point in [204:57] time there was approximately a two week [205:00] delay between pushing an update and it [205:02] being reviewed and accepted not only is [205:05] there a twoe delay during that two weeks [205:07] Apple and Google are reviewing and if [205:09] they see anything that they don't like [205:10] for whatever reason they are telling you [205:12] no you cannot push that code or your [205:14] company could be in a fight with that [205:16] company and they could decide H I don't [205:18] think you should push any updates for a [205:19] while until we resolve our fight yep [205:22] okay so that's number four and then that [205:24] brings me to number five reason that I [205:26] could think of of why this is really bad [205:28] for Facebook the competitive field is [205:31] about to get a total reset on the web [205:36] Facebook is dominant Network effects are [205:38] super real everything we talked about in [205:41] this whole first part of the episode of [205:42] they can suck up all of the new social [205:45] functionality that they are anyone else [205:46] invents they can put it into the main [205:49] Facebook app photos chat platform [205:52] Newsfeed Etc on mobile though Facebook [205:56] is just one icon of many on the home [205:58] screen exactly and the wisdom at the [206:00] time was that mobile apps should be [206:02] narrow in their functionality and you [206:04] did not expect a single app to bundle in [206:06] a ton of different use cases it's super [206:09] easy for users to be like H okay I get [206:12] this utility from Facebook but like well [206:15] I don't know here's this in thing like [206:17] that's kind of cool I get great photo [206:18] utility from that which also lives on my [206:20] home screen right there and I'll still [206:22] use Facebook for seeing when my friends [206:24] get married but like Photo is cool and [206:27] messaging you know this WhatsApp thing [206:28] is super cool like maybe I'll just use [206:30] that for [206:31] messaging it's super easy to use one app [206:33] for one thing and other apps for others [206:36] and you can see why Facebook adopted [206:38] that uh sort of early 2010's [206:40] constellation of apps strategy for a [206:42] while they had slingshot poke messenger [206:46] p paper rooms riff camera the belief by [206:50] a lot of companies for the direction [206:52] mobile was going was there's going to be [206:54] specialized apps each for their own tiny [206:56] little purpose and that's not great if a [206:58] lot of your value is we bundle a lot of [207:00] stuff in to create the most user user [207:02] engagement to all feed into each other [207:04] for people to use all these different [207:05] components of our application yep [207:08] totally okay those are my short list of [207:11] five reasons of why the shift to mobile [207:13] is dangerous to Facebook anything else [207:15] you would Dad no so they're going public [207:18] right into this for the first time since [207:21] our business is founded we Face a real [207:24] existential threat completely out of our [207:26] control that is going to make the next [207:28] few years look really bad let's go [207:30] public yeah well let's talk about that [207:32] IPO yes but first it is time to talk [207:35] about one of our favorite companies [207:38] stats Sig a phrase that many of you will [207:40] know from Facebook's early days is move [207:43] fast and break things but despite [207:45] instilling this in Facebook's [207:46] engineering culture Facebook didn't [207:48] actually break very often and it [207:50] essentially never goes down now how well [207:54] Facebook invested hundreds of thousands [207:56] of engineering hours in a set of [207:57] internal tools these tools let any [208:00] engineer set up new metrics ship new [208:02] features and measure performance in real [208:04] time this meant that anyone could just [208:06] ship a new feature but they always had [208:08] metrics to use as guard rails and they [208:10] could always roll back the feature if [208:12] anything broke they took this so far [208:14] that they have every new engineer ship a [208:16] feature on their first day at the [208:18] company wild you might wish your team [208:21] could build products like Facebook right [208:23] ship fast make data inform decisions [208:25] iterate rapidly but uh you don't work at [208:27] Facebook and you don't have those tools [208:29] so you're stuck right well not anymore [208:32] thanks to stat Sig stat Sig was founded [208:35] by an xmeta team three years ago with a [208:37] mission to make these same tools that [208:39] Facebook has available to any company [208:41] could there be a better sponsor for this [208:43] episode David this is insane I know it's [208:45] perfect it's perfect perfect today [208:47] they've done it combining tools like [208:49] feature Flags product analytics [208:51] experimentation and observability into [208:53] one connected platform that runs off one [208:55] set of data and infrastructure no more [208:58] stringing together expensive Point [208:59] Solutions and internal builds and no [209:01] more using technology vendors built five [209:03] or 10 years ago when the needs of modern [209:05] product teams have grown by Leaps and [209:07] Bounds and they've also gotten some [209:10] crazy traction over there at Stats many [209:12] of the world's leading tech companies [209:14] rely on them including open AI Microsoft [209:17] notion anthropic figma plus thousands of [209:20] early stage startups in fact their scale [209:22] has gotten so crazy that they process [209:24] over a trillion events per day for any [209:27] Engineers listening they have a great [209:29] blog post about how they do this and [209:31] best of all stat Sig is pretty [209:33] affordable they have an insanely [209:35] generous free tier for small companies a [209:37] startup program with 1 billion free [209:40] events which is $50,000 in value and [209:43] significant discounts for Enterprise [209:44] customers to get started just go to [209:46] stat.com Acquired and remember to tell [209:49] them that Ben and David sent you thanks [209:51] stat all right so David here we are [209:54] we're going public despite uh everything [209:57] you just listed that is wrong with the [209:59] future of mobile and this company oh man [210:02] which seems like a not very smart idea [210:05] right and uh you know look one thing [210:07] about Mark Zuckerberg you could say lots [210:10] of things about him but one thing I [210:11] don't think anybody could say is that [210:12] he's not smart so what gives why are [210:16] they doing this well back then once a [210:20] company crossed 500 shareholders the SEC [210:24] actually mandated by law that you had to [210:26] start reporting quarterly like a public [210:29] company whether or not you were traded [210:31] on an exchange whether or not you were [210:33] traded on an exchange and at the end of [210:35] 2011 Facebook crossed 500 shareholders [210:39] so for Facebook this means that they [210:41] were going to have to report like a [210:43] public company anyway and in practice [210:46] the common wisdom was you have to go [210:49] public as part of this anyway because [210:52] otherwise you're going to report and [210:53] then you know you're going to lose all [210:55] the momentum and excitement about having [210:57] an IPO and doing an S1 and having a big [211:00] Road Show etc etc so you might as well [211:02] just do the IPO as long as you're going [211:04] to have to do this and I think Mark and [211:07] Cheryl and the team like really debated [211:09] like well should we skip the IPO but [211:12] like well shoot then we're going to be [211:14] reporting as a public company and [211:16] heading into all of these headwinds [211:18] without having actually gone public and [211:19] raised the money that we might need to [211:22] be able to make these Investments so [211:23] they're like nope we got to just do it [211:26] and go public yes it is a odd time to go [211:30] public given everything with mobile and [211:32] of course they're being forced into it [211:34] there actually are a couple of kind of [211:36] Tailwinds that they have that are [211:38] probably worth touching on here the [211:40] first of which is they had just beat [211:43] goog plus that's right Facebook treated [211:46] this like a total existential threat we [211:49] laugh about go+ now it's a butt of jokes [211:51] but that's because Facebook was so [211:53] effective in competing against it I'm [211:56] not saying the product itself was [211:57] amazing and it would have been fine [211:59] without Facebook but like Google did not [212:00] end up executing that well on that [212:02] product it actually is a really good [212:04] point though Google+ may or may not have [212:06] been a good product and may or may not [212:08] have succeeded but the Facebook [212:09] competitive response was like [212:11] unbelievable unbelievable I mean this is [212:12] like the growth team plus Facebook's [212:14] engineering culture at it's finest in [212:17] defeating this whether or not they [212:19] executed the product well Google was the [212:23] big web tech company at the time and [212:25] they put a ton of resources and a lot of [212:27] their best people on it and they had [212:29] Gmail and they had Gmail right so it's [212:32] interesting to look at this because [212:33] Google basically is Facebook's biggest [212:38] business model comp they show ads on the [212:41] web and they monetize that really well [212:44] this theoretically could have been in [212:46] their wheelhouse this is just more real [212:47] estate to show ads on the web and [212:49] they've already got all these people [212:50] with user accounts you can see why [212:53] Facebook took this really seriously we' [212:54] talked a lot about Facebook's technical [212:56] prowess well here is an example of where [212:59] it really kicks in when you need it to [213:01] to be a key competitive Advantage so [213:04] they structured themselves in a way that [213:06] encouraged flexibility of Engineers and [213:09] they really broke from the tide of [213:11] microservices in this era they had one [213:13] monolithic code base that everyone [213:15] worked out of and you might say well [213:17] that's stupid why would they that's not [213:19] the way the world was going but what [213:21] they basically did was they wanted to [213:23] encourage portability of teams if you're [213:26] an engineer you get hired into the [213:27] company not onto a team you learn the [213:29] company's codebase you learn the [213:30] company's conventions and you can easily [213:32] sort of move around after that you do [213:34] have to deal with the trade-off then of [213:36] you have this big monolithic codebase [213:37] with gigabytes of PHP code for thousands [213:39] of Engineers what do you do about that [213:41] well then they just had their cake and [213:42] eat it too they would go and have the [213:45] infrastructure team figure out how to [213:46] deal with that problem rather than [213:48] saying oh everyone can just work in [213:50] their little silos and so that meant [213:52] that Engineers could kind of be quickly [213:55] reorg they could have this companywide [213:57] lockdown to fight goog plus and they did [213:59] all sorts of things they launched video [214:01] calls to compete with Hangouts they [214:03] launched something to compete with [214:04] Google Circles but either way like they [214:06] could really quickly reallocate [214:08] resources and people who sort of knew [214:10] how to work together to defeat what [214:13] could have really been an existential [214:15] threat from their biggest similar [214:17] company oh man Mark famously in a [214:20] internal speech around this uh you know [214:23] referenced uh what was I think KO the [214:24] Elder yes from uh ancient Rome of [214:27] Carthage must be destroyed yep there is [214:30] a second way in which things had settled [214:32] down and now was a good time to go [214:35] public and that's around privacy [214:37] Facebook had just been I mean playing [214:40] way too fast and loose with user privacy [214:42] for years and it had finally caught up [214:44] with them by 2011 and just to jog your [214:46] memories I'm sure people may remember a [214:47] lot of these some of these included even [214:50] though users could restrict the audience [214:52] of their posts with a setting this [214:54] apparently didn't apply to apps which [214:57] could access these posts regardless of [214:59] how you restricted the audience and for [215:01] a time this even included when your [215:03] friends installed an app too you didn't [215:05] even have to Grant the permission [215:07] yourselves similarly they made friend [215:10] lists public without user consent at one [215:12] point Facebook could decide without user [215:14] consent to change private fields of your [215:17] profile to be public and this wasn't [215:19] always messaged as clearly as it could [215:21] have been to users so to remedy this [215:23] they had just signed what is called a [215:25] consent decree with the FTC the US [215:28] Federal Trade Commission in 2011 and [215:31] they promised to make a bunch of changes [215:33] regarding user privacy issues going [215:35] forward and so all this was behind them [215:37] now and interestingly David an FTC [215:40] consent decree is the same thing that [215:43] Microsoft signed yeah hark back to the [215:45] Microsoft episode yes the FTC consent [215:47] decree with Microsoft was the [215:49] predecessor to the big doj suit in this [215:51] case the FTC consent decree is the [215:53] predecessors to what eventually would [215:55] become the Cambridge analytica [215:56] settlement so here in 2011 the way that [215:59] they settle this is there's a bunch of [216:01] Provisions that with Facebook promising [216:03] they'll be tighter on making sure user [216:05] data is treated in a very particular way [216:08] they're subject to audit every year for [216:09] 2 years for the next 20 years you know [216:11] there's there's all these things that [216:12] they sort of agree to but you know once [216:15] you a consent decree it's like okay [216:16] we're through it we don't have this [216:18] thing hanging over our heads we can go [216:19] public and say oh yeah that's in the [216:20] past we've taken care of it that's [216:22] another reason to go public right now [216:24] yep it's such a good point there [216:27] actually was this little window here at [216:30] the beginning of [216:31] 2012 where desktop Revenue hadn't yet [216:36] been impaired so badly that they could [216:39] still show growth they had these couple [216:42] little wins they could point to Y and [216:44] say like all right not great but like if [216:47] we're gonna do it we got to do it now [216:49] yep you could almost look at it like a [216:51] little bit of a win of everyone knows us [216:55] as the company that is uh a little bit [216:57] dodgy on privacy practices like at least [217:00] the public perception is this that [217:01] they're constantly changing the terms of [217:04] the game shall we say yes to the [217:06] company's Advantage when it was [217:08] confusing or misleading to users and now [217:10] you could say look not only do we sign [217:12] that we have these five product [217:13] initiatives in place where we're just [217:15] buttoned up now I actually think that's [217:17] pretty true I think they became a [217:19] company that had rigorous privacy [217:22] practices because of some government [217:24] regulation when they otherwise may not [217:26] have I mean if you look at the early [217:28] days of what you could do as a developer [217:30] on platform it was pretty wild west and [217:33] so I'm not saying it's fully because of [217:35] the consent decree but they could [217:37] definitely tell a story around look we [217:39] made some mistakes we got some things [217:41] wrong in privacy and going forward we're [217:43] in good shape we're all good yeah well [217:45] we'll come back to that in a little bit [217:47] but for the moment the IPO February 1st [217:51] 2012 they filed the S1 they enter the [217:53] traditional quiet period before the IPO [217:56] where legally according to the SEC once [217:59] you have filed for an IPO all material [218:02] information about the company needs to [218:04] be included in the S1 registration [218:06] statement you can't go give interviews [218:08] you can't go talk to the Press you can't [218:09] say anything meaningful about the [218:10] company besides what is fully publicly [218:13] disclosed to any potential investor in [218:16] the S1 statement and then a week later [218:20] on April 9th [218:22] 2012 during the quiet period Facebook [218:26] acquires Instagram for a billion [218:31] dollars time's now baby the time is now [218:34] and they can't say anything about it [218:36] besides what they file in the amended S1 [218:38] because they're in the quiet period [218:40] amazing the company's previous largest [218:42] acquisition was $70 [218:44] million obviously this is a way [218:46] different thing now sitting here today [218:49] you might be really tempted to say oh [218:51] well there's the fix for mobile they [218:54] just bought Instagram that solved all [218:55] the [218:56] problems except not because Instagram [219:00] didn't have any revenue on mobile [219:02] either it's like uh you know you're [219:05] tying one anchor to another here well [219:07] not to mention Instagram had 27 million [219:09] users Facebook had 900 million users [219:12] this was potentially a problem for [219:15] future or Facebook but this was not [219:17] currently a problem totally I was just [219:20] about to say that too like yeah also not [219:24] currently a fix to Facebook's major [219:26] problem of the shift to mobile and I [219:28] mean my God talk about the hutba of Mark [219:32] to do this you're already in a really [219:35] challenging period for the company a [219:37] really challenging IPO process you just [219:39] entered the quiet period and you decide [219:42] to spend a billion dollars on something [219:44] with 27 million us and no revenue and [219:46] you're not going to be able to talk [219:47] about it except what you file with the [219:49] SEC wow you got a really really really [219:54] freaking believe in the future potential [219:57] of your company and the combined [219:58] companies of Facebook and Instagram here [220:00] to make this move not to mention on top [220:03] of it all we're not going to go into it [220:04] here because we had a whole episode on [220:06] Facebook and Instagram but this was done [220:08] over two days over the course of a [220:10] weekend Mark didn't do a whole lot of [220:12] socializing this before pulling the [220:14] trigger he just knew knew it was the [220:15] right thing to do and did it yep and [220:18] that ended up working out pretty well in [220:20] the long run in the short run yet [220:23] another question mark to add to the [220:24] already very large pile around this [220:26] company finally long road show process [220:30] by the week of May 14th they're ready to [220:32] price the IPO pricing is set for [220:35] Thursday night May 17th [220:38] 2012 on Tuesday May 15th General Motors [220:45] GM who by the way was not a particularly [220:47] large Facebook Advertiser announces [220:50] publicly to the world that they are [220:51] pulling all their spend off of the [220:53] Facebook platform because it's not [220:55] effective not really what you want to [220:58] have happen two nights before you price [220:59] your IPO was that motivated by someone [221:02] that feels like a hit feels like it must [221:04] have been I've never heard one way or [221:05] the other but like a why would you [221:07] announce that publicly B why would you [221:09] announce it publicly right then like [221:11] there's for sure more to that story [221:13] nonetheless Faceook goes ahead Thursday [221:16] night they priced the IPO at $38 a share [221:20] at the very top end of the range after [221:22] the road show again the huta here they [221:25] rais $16 billion at a [221:28] $14 billion market cap making it I think [221:33] the third largest IPO of all time at [221:35] that point behind visa and ironically [221:38] General Motors coming out of uh the [221:41] bankruptcy during 2008 which both of [221:43] those are mature companies that are [221:44] going public under weird circumstances [221:46] those are not comps totally yes by far [221:48] the largest IPO of a like new startup [221:52] company in history on Friday morning [221:55] trading [221:56] starts the NASDAQ is overloaded with so [222:00] many orders for buying and selling [222:02] Facebook that trading in Facebook stock [222:04] doesn't actually open until 11:30 [222:07] a.m. which is you know two plus hours [222:09] after the open of the markets got to be [222:11] terrifying if you're a big Facebook [222:12] shareholder yeah that's pretty [222:14] terrifying initially it looks like oh [222:16] that delay might have been good because [222:18] the price shoots up to $45 a share from [222:21] 38 but then it crashes back down it is [222:26] one of the most volatile trading days in [222:28] any stock I think in history on any [222:31] exchange Facebook shares end the day at [222:36] $382 3 up 23 from the IPO price but only [222:41] because all of the underrating banks in [222:42] the IPO stepped in to buy shares and [222:45] support the price which was totally [222:47] crashing throughout the day thank God [222:50] they did it on a Friday so everybody had [222:51] the weekend to breathe once trading [222:54] starts again the following week the [222:55] price falls on nine of the next 13 [222:58] trading days and ends up by the end of [223:01] May down 25% from the initial IPO price [223:06] of [223:07] $38 kind of in freef fall at this point [223:10] the stock continues going lower and [223:12] lower and lower ultimately bottoms out [223:15] on September 4th 2012 at [223:20] $17.68 down [223:22] 53.5% from the IPO price so market cap [223:26] cut in you know over half in what's that [223:31] 5 months since the IPO it wouldn't end [223:34] up reaching IPO price again until August [223:36] 2013 anyone who bought at the IPO was [223:39] underwater for 16 months underwater for [223:41] 16 months and 5 months later had lost [223:44] half their money on Facebook so you have [223:46] to have nerds of Steel to even just keep [223:48] holding I'm sure anybody who got back to [223:50] even at 16 months was delighted totally [223:53] you've also got the employee lockup [223:55] coming up uh around November so like now [223:57] you're staring down the gun of shoot [224:00] I've already lost half of my value as a [224:02] public company and pretty soon all the [224:04] employees in the company are going to be [224:06] able to sell it that's going to put more [224:07] selling pressure on the [224:09] stock holy crap this is pretty darn [224:13] bad yes that is horrible I mean losing [224:17] 50% of your value as any company public [224:20] or private but especially a public [224:21] company is a disaster of Epic [224:24] Proportions having that happen [224:27] immediately following your IPO for most [224:30] companies there's no way to come back [224:31] from that you're entering a death spiral [224:33] you're going to end up being acquired or [224:35] you know you're just going to get driven [224:37] to zero basically yep and just to be [224:40] super clear desktop Revenue wasn't going [224:42] to go to zero immediately but Wall [224:45] Street investors are really worried that [224:47] like hey the world is transitioning to [224:49] mobile eventually growth will start [224:51] slowing and it will start declining [224:53] which it does fairly quickly here and so [224:56] like if you can't show me that you're [224:57] going to replace that with mobile [224:59] Revenue why should I value your stock at [225:02] anything yep and so here is where [225:05] founder control matters this is David [225:08] it's interesting you're explaining the [225:09] mechanics of a death spiral that would [225:11] presume there's a board of directors who [225:13] feels a strong desire to do whatever is [225:15] best in the interest of the shareholder [225:18] and they might think that it's these [225:19] short-term things but if you have a CEO [225:23] who has a 20-year View and believes deep [225:25] deep in his heart that The Best Is Yet [225:27] To Come and that this is just a [225:29] temporary problem and that person [225:32] controls the company and owns the [225:34] majority of voting shares you can [225:36] withstand things like this you just [225:39] better be right yes so uh little fun [225:43] anecdote for everyone David and I among [225:45] the 20 other people we talked to to prep [225:47] for this episode one of them was Cheryl [225:49] Sandberg and we were asking Cheryl in [225:51] particular how did you start the [225:54] monetization effort on mobile they were [225:56] in the third column there's no third [225:57] column on mobile what did you do and her [226:00] comment to us was oh we just stopped [226:02] caring about the right side ads on [226:04] desktop and we took every engineering [226:06] resource we could off of that and we put [226:08] it on mobile and we knew we were going [226:10] to miss the current quarter I think they [226:12] missed a lot of quarters right after [226:14] their IPO but this was us trading the [226:16] present for the future and all we cared [226:18] about was our future and she said this a [226:21] great quote this is up there with one of [226:23] the best quotes on a chir of all time [226:25] she was sitting there with Mark late at [226:27] night and when they sort of arrived at [226:30] this plan of we are going to foro a lot [226:33] of desktop Revenue to basically bet it [226:35] all on figuring out mobile Revenue she [226:37] said well Mark nobody can fire you and [226:40] only you can fire me so if you're in I'm [226:43] in and we buckled our seat belt and we [226:45] said here we go and it's amazing I mean [226:48] that can only happen in a Founder [226:49] controlled company and it really did [226:52] force them to figure out mobile [226:54] advertising if they really are saying [226:56] like this is the new first class product [226:58] this is where we're going to point all [227:00] advertisers toward if they're wrong on [227:03] that they're wrong across the whole [227:05] board because if ads suck on mobile [227:08] since it's the only thing and it's [227:09] filling up your whole screen user [227:11] engagement is also going to nose dive [227:13] and so it's basically a bet all moment [227:15] where the ads are front and center so [227:19] they must be good it is essential for [227:21] the mobile product and thus your company [227:23] for them to be good users attention will [227:26] be pointed at these ads like no other [227:29] ads you have ever run before and [227:31] actually the flip side of this is it [227:33] ended up being the best thing ever for [227:35] the company because the ads are front [227:38] and center the value per ad is actually [227:41] higher so they ended up creating a much [227:44] much more valuable ad unit than they [227:45] ever had on desktop purely because of [227:47] this incredible attention on them when [227:50] you're scrolling in feed on mobile [227:51] necessity is the absolute mother of [227:53] invention yes totally you have no choice [227:56] but to do your best work and boy did it [227:59] work out fabulously yeah Ben nobody can [228:02] fire you well I guess you can't fire me [228:05] either but either way if you're in I'm [228:08] in the Corp structure acquir is still a [228:10] little that's unclear but yes totally so [228:14] when we say [228:15] taking your best most trusted Engineers [228:17] taking them off whatever they're doing [228:18] and putting them on ads there's actually [228:20] an amazing story around this so Mark [228:24] goes to someone within the company who [228:27] he knows can just ship and get stuff [228:30] done and that is BOS Andrew Bosworth his [228:34] old TA from Harvard BOS at this point is [228:37] running the core profile and timeline [228:41] within Facebook and Mark goes and talks [228:44] to him and says is BOS I think you [228:47] should come work on ads to which BOS [228:50] replies okay I mean like I get we're in [228:52] this situation but you know Mark like I [228:55] can't do this for too long because [228:57] remember I'm getting married this fall [228:59] and I've got this six-month sabatical [229:02] plan for my honeymoon and my wife and I [229:05] we're going to travel the world it's all [229:07] like booked and paid for we're going to [229:09] be out of the [229:10] country and Mark is like yes well we've [229:13] got a few months until then and I think [229:15] you should come over and work on [229:17] ads I was like okay okay I will do that [229:20] I will come up with something here [229:21] before the wedding so he teams up with [229:23] Will Cathcart from the newsfeed team [229:25] from Chris Cox's team and they go get to [229:28] work by the way will is the head of [229:30] what's app today and boss of course is [229:31] the CTO of meta and head of reality Labs [229:34] Y and the first thing they do during [229:36] these couple months is like all right we [229:38] just need like a Band-Aid to stop the [229:40] bleeding and at least produce some [229:43] mobile Revenue and what they decide to [229:45] do is like hey we do have on mobile the [229:48] Carousel of people you may like what if [229:50] we just slot some Brands into that [229:52] Carousel and say hey in addition to [229:55] people you may like here here are some [229:57] brand Pages you may like h a reasonable [230:00] first step and look they knew that this [230:02] wasn't going to be the long-term [230:03] solution and solve all the problems but [230:06] it did start to produce some mobile [230:08] Revenue that they could report and this [230:11] was super important because as I alluded [230:13] to the lockup on employees and insiders [230:15] selling their shares expired in November [230:18] they had to get something out that was [230:19] going to prop the stock price up a [230:22] little bit before naturally employees [230:24] are going to start selling here and [230:26] that's going to put downward pressure on [230:27] the stock price yep also bummer if you [230:30] are an employee selling six months after [230:32] that IPO hopefully as many people held [230:34] as they could to get through it yes [230:36] hopefully and I think most people who [230:39] were at Facebook at this time really did [230:41] believe in the future of the company at [230:44] the same time you have life [230:46] circumstances you might need to buy a [230:48] house you might need to do things for [230:50] your family Etc it's a tough spot so [230:54] fall rolls around they've launched [230:56] sponsored Pages you may like BOS gets [230:59] married they're still grinding on the [231:02] big project which is called project [231:04] whale shark and that is getting real [231:07] honest to God native ads in the news [231:09] feed which frankly they should have done [231:11] years ago on desktop but like now now [231:14] the gun is to their head there is [231:16] nothing in the mobile app except [231:18] Newsfeed we got to put ads in Newsfeed [231:21] so finally at the end of the year they [231:24] launch it BOS is like okay mission [231:27] accomplished I can now leave on my [231:29] sabatical we can go on our honeymoon but [231:32] there's just one problem shipping this [231:34] thing is only like 20% of the battle now [231:37] you have to go sell it to [231:40] advertisers which this is a whole new ad [231:43] Paradigm you got to like educate brands [231:45] on why this is going to work for them [231:48] especially when your overall brand as a [231:50] company and effectiveness of your [231:52] advertising is kind of in the dumps [231:53] right now and it's an iterative process [231:56] where you're going to hear a bunch of [231:57] feedback from Brands and you're to say [231:58] oh we got to modify the ad products [232:00] which still continues to this day they [232:03] roll out and modify ad products there [232:05] not like a uh all right we now have ads [232:07] in the newsfeed we're good the job is [232:09] never done here so of course [232:11] quarterbacking all this is going to fall [232:13] to Cheryl and her team but again as you [232:17] said Ben this is an iterative process [232:18] you can't divorce product from the [232:21] feedback from [232:22] advertisers and the big advertisers are [232:25] sure as hell going to expect someone [232:26] from product who is leading this thing [232:28] to show up and tell them about it yep so [232:30] BOS kind of needs to come along to these [232:32] meetings so on December 18th BOS [232:36] remembers the exact day Mark calls him [232:38] up and is like bz great job working on [232:41] ads hey I really still need you to come [232:45] back and work on ads and BOS is like no [232:47] way man I did the past six months for [232:49] you I just got married I'm going on my [232:51] honeymoon I'm out Mark's like okay okay [232:54] okay the next day calls him back [232:56] December 19th and he's like BOS I just [232:59] want to say really really great job [233:01] again and you know the company's in a [233:04] really tough spot and like I really [233:06] think that you should come keep running [233:09] ads [233:11] here and uh B is like okay all right I [233:15] will reduce my sabatical from six months [233:18] to six weeks I'll take the holidays the [233:20] beginning of the year I'm going to [233:22] cancel the last four months of my trip I [233:24] will do six weeks with my wife and then [233:26] we'll fly her friends out to the last [233:28] two weeks with her and uh then I'm gonna [233:31] go fly around the world with Cheryl on a [233:33] very different kind of trip [233:35] here so that's what they do and for most [233:39] of 2013 BOS and Cheryl are out on the [233:41] road they're pitching advertisers [233:43] they're explo exping sponsored stories [233:45] in the news feed on mobile why this is [233:48] so [233:48] important not only do advertisers need [233:51] education about this new Facebook ad [233:54] unit they also need education about [233:56] mobile advertising in general they'd [233:59] only just gotten used to digital [234:00] advertising on desktop totally do you [234:02] remember those Mary Meer Decks that used [234:04] to go around where they would show the [234:06] mobile advertising basically the [234:08] shortcoming look at all this attention [234:09] that has shifted to mobile and yet the [234:11] monetization just isn't there and that [234:13] was a story every year for like a decade [234:15] yep totally and it was the same story [234:17] with desktop on the web for a decade [234:19] before that yep so what's interesting is [234:23] like as they're getting going here of [234:25] course advertisers don't come in right [234:27] away but also the ones that do come in [234:30] early are kind of as you would imagine [234:32] not the best advertisers and so the [234:36] whole effort starts to have this real [234:38] problem where the ads that are now [234:40] showing up in Newsfeed are like pretty [234:42] freaking crappy ads and Mark keeps [234:45] getting more and more pissed because [234:46] like obviously this is really important [234:48] but he's like BOS Cheryl like why the [234:51] hell is all this crap showing up in my [234:53] feed like if this keeps happening I get [234:55] that we have to save the business but [234:56] like I care more about the user [234:59] experience and like I'm gonna dial this [235:01] back this is the most interesting thing [235:03] at this point in history Mark is putting [235:05] pressure on BOS and Cheryl to show fewer [235:08] ads yes and BOS is the one with the [235:11] contrarian view who comes to Mark and [235:13] says we're thinking about this all wrong [235:15] we actually need way more ads not just a [235:18] little more ads huge amounts more ads [235:22] because a great ad is on par with [235:27] content and if you have tons and tons of [235:30] ads then we can do way better targeting [235:32] like you have this beautiful liquid [235:34] Marketplace of hundreds of thousands of [235:36] advertisers billions of users and at any [235:40] given time somebody can see the optimal [235:42] ad for them and get recommended amazing [235:44] product that perfectly fits their needs [235:46] in that moment with messaging that's [235:48] perfectly tailored for them great [235:51] advertising can be great but you need a [235:54] really really liquid Marketplace and you [235:56] need fantastic algorithms which you can [235:59] only have if you train them on tons and [236:01] tons of data so like we actually need [236:04] 10,000 times more ads than we have right [236:06] now not less it's almost like the only [236:09] way out is through we're going down this [236:11] path we must be extremely successful in [236:14] order to be successful at all this was [236:17] such a contrarian View and also right [236:21] and I think only could have come from [236:25] someone like BOS being focused on this [236:28] problem because as a product person and [236:33] an [236:34] engineer it's the only way that you're [236:36] going to have that Insight that like oh [236:39] this is the same problem we had when we [236:42] launched Newsfeed the way to fix [236:44] Newsfeed and to make it really really [236:46] awesome is more Newsfeed you know is [236:49] more data right like the only way out is [236:51] through yeah if somebody had purely come [236:54] I think from the advertising World they [236:56] would have been like well you know [236:58] that's just the nature of advertising [237:00] right it's attacks on the users and [237:02] that's the way monetization of media [237:03] works but like BOS was like no no no [237:06] actually this is a relevance problem and [237:08] we can make this awesome and in order to [237:10] make it awesome we need a lot more of it [237:12] yep and basically like that's what [237:14] happened I mean Flash Forward 10 years I [237:16] open Instagram to get great product [237:18] recommendations totally It's actually [237:21] kind of wild I believe this is true [237:23] there are markets in the world today [237:25] where I think meta has run this test [237:27] where if you turn off ads in the you [237:31] know meta family of apps products [237:33] engagement in those products actually [237:36] goes down which is just like freaking [237:38] wild you would think like oh wow now I [237:40] get to use meta products without [237:41] advertising no tax on me like I'm going [237:43] to engage more like no it goes down [237:45] that's the definition of ads as content [237:47] if that's actually the case totally wild [237:50] so by the fourth quarter of [237:52] 2013 which honestly in the scheme of [237:54] things is pretty freaking fast the whole [237:58] thing finally flips Facebook announces [238:01] that mobile advertising is now 53% of [238:04] Revenue and total ad revenue for the [238:06] company grew 76% year-over-year in the [238:10] fourth quarter which of course is the [238:11] most important quarter of the year which [238:13] means I did all the math on that that [238:16] means that all of desktop Revenue all of [238:19] it so the old platform business the old [238:22] ad business the Microsoft third party ad [238:25] serving all of that actually [238:28] shrank in the fourth quarter of 2013 so [238:31] that's the counterfactual here if [238:33] Facebook had not gotten mobile [238:35] monetization right that's what would [238:37] have happened is Q4 [238:39] 2013 all of the growth momentum that it [238:42] had would have flipped and it would have [238:44] started the downward spiral and probably [238:47] not that early if I guess that's right [238:50] cuz all the resources got pulled off of [238:51] it exactly but that's what would have [238:53] happened eventually to it eventually yep [238:55] yep so man it's like this wild 18-month [238:59] Journey but it totally saved the company [239:03] because now here they are they've built [239:06] this unique [239:09] pioneering and totally defensible native [239:13] mobile ad unit in the feed they Define [239:16] the whole generation of you know [239:17] monetization here it's the best ad unit [239:20] in history it's an ad that completely [239:22] fills the screen on your device and that [239:25] users are okay with so this is the most [239:28] insanely captivating engaging ad unit [239:32] that you could possibly ask for and it [239:34] came out of necessity it's wild the fact [239:36] that they thought their backs were [239:37] against the wall they were totally [239:38] screwed and instead actually it's the [239:40] thing that monetizes better and is [239:42] better for advertisers than any other [239:44] advertising product they've ever had and [239:46] you know I know we keep saying this and [239:47] it sounds maybe sort of um silly but [239:51] like really better for users too before [239:54] this Paradigm that they invent here of [239:57] you know algorithmically driven ads in [239:59] the feed the biggest [240:02] monetizing thing on mobile was an [240:05] interstitial like popup freaking display [240:08] ad that took over your screen right like [240:10] what an awful freaking experience for a [240:12] user right right around this time too [240:15] the other thing that's happening is kind [240:17] of glossed over this for time but [240:18] Facebook's initial attempt at a mobile [240:21] app was to try to work around all the [240:25] constraints of the mobile app ecosystem [240:27] and they thought well that's nice that [240:29] the app stores are going to try and box [240:30] Us in but we'll just ship our mobile web [240:33] views inside of a thin little app [240:36] wrapper and that way we can deploy [240:38] multiple times a day you know we're [240:40] Facebook this is what we do it's part of [240:41] our culture HTML 5 wooo provided a [240:44] horrible user experience I mean the [240:46] engagement on the app was low time spent [240:48] was low it was a risk to start selling [240:51] these ads because people don't want to [240:54] spend any time in the app even without [240:55] ads let alone when you start layering [240:57] these in and so they're finally starting [241:00] to sort of pull out of this Tail Spin [241:01] they basically spent a year completely [241:04] rewriting all their mobile clients to be [241:06] these rich beautiful native experiences [241:09] this is a thing that Facebook has always [241:10] been good at whenever they decide to do [241:12] something they go and recruit like like [241:14] the actual best talent in the world to [241:16] do it the group of people that they [241:18] pulled together from Joe huitt forward [241:20] to write their original IOS app it's [241:21] just like some of the best iOS [241:23] developers and designers in the world [241:25] they hired Mike Mattis and the push poop [241:27] press team when they acquired them [241:29] that's right that's right that became I [241:31] think it was creative Labs is that right [241:33] Facebook creative Labs but you know a [241:35] huge amount of that Talent worked on [241:37] their their mobile apps and so well they [241:40] had the wrong strategy at first once [241:41] they got religion around native they [241:44] really created something probably one of [241:46] the best apps ever on mobile totally [241:51] It's funny I'm laughing you know the [241:53] version of history that we just told is [241:56] maybe slightly too charitable to [241:58] Facebook and to mark because yes it's [242:01] absolutely true they did the really hard [242:04] thing they took the pain they invested [242:06] in the future they built you know [242:09] everything that would become Facebook [242:11] and meta today out of it but the first [242:14] reaction was to bet on Hope as a [242:16] strategy and breur your head in the sand [242:17] and say HTML 5 is going to save us I [242:20] don't know that it was hope as a [242:21] strategy I think it was more like if [242:24] this can work it's going to solve a lot [242:25] of our problems and I don't think they [242:28] correctly estimated how wide the user [242:33] experience Chasm is between web apps and [242:35] Native apps and I think they had to like [242:37] have an app in Market where users were [242:39] actually using it to realize oh man the [242:42] state-of-the-art in native that the [242:45] platform vendors have developed IOS and [242:47] Android is really really good and they [242:50] have brought very little of that to the [242:52] web experiences partially because of [242:54] Standards bodies but also partially [242:56] because it's not really in their [242:57] interest to why would they make web apps [242:59] great when they can force everything [243:01] through a an App Store that they have [243:03] more control over so okay after the Q4 [243:06] 2013 earnings announcement the stock [243:09] pops way way way up to like a 175 [243:13] billion ish market cap so up 75% from [243:17] IPO you know what is that year and a [243:19] half earlier and from there it's [243:21] basically if you look at the market cap [243:23] chart kind of a straight line Journey [243:26] from 175 up to half a trillion over the [243:29] next couple years they eventually bring [243:32] this whole native feed monetization [243:36] engine over to Instagram and that works [243:39] obviously incredibly well there probably [243:42] better than even on the Blue app and [243:44] that in particular the I mean I [243:46] everyone's GNA laugh when we say this [243:47] word the legitimate Synergy between [243:50] going to an Advertiser and saying you [243:52] can use this dashboard to get placements [243:54] on Facebook and on Instagram is massive [243:59] both of those products monetize better [244:01] than they ever could without that single [244:03] channel that the advertiser only has to [244:05] go through and use One dashboard to [244:07] place on both products if you flash all [244:09] the way forward to today the Lion Share [244:12] of meta's Revenue comes from the ads [244:14] that run on Facebook and Instagram their [244:17] whole business today can just be [244:18] summarized as that totally there's [244:21] another element for a couple years [244:23] that's a big part of the Native mobile [244:25] Revenue story and that is app install [244:27] ads and actually is worth calling out [244:29] because it plants the seeds of this much [244:32] larger struggle that's going to come in [244:34] a little bit but a very large percentage [244:37] of the inventory that they were selling [244:40] and especially the revenue of these [244:42] native mobile ads in the beginning was [244:44] actually for app install ads that game [244:47] developers and other folks were buying [244:49] on Facebook so it was kind of like our [244:51] old platform business all over again [244:53] Facebook is actually the best place to [244:55] discover apps on mobile yeah it's pretty [244:58] interesting realization where the Big [245:01] Brand advertisers may not move to this [245:04] new type of AD format right away but the [245:06] people that are going to be really [245:07] hungry to move to that ad format are a [245:10] game developer who makes a mobile app [245:12] and wants to Market their mobile app to [245:15] people who are a on that platform and B [245:17] in a Leanback experience where they're [245:19] open for some entertainment and when you [245:21] are scrolling through a feed of your [245:23] friends and Brands and you are open to [245:26] oh hey look at this a game where I could [245:29] click one button and then boom install a [245:32] game and play it is there a better [245:35] moment and channel to reach someone for [245:38] an app no I mean even if you're apple [245:40] apple doesn't have a better way to do [245:42] this people don't search the App Store [245:44] for apps that often so you'd have to [245:47] show them like a popup ad or something [245:49] Facebook just has this opportunity where [245:50] you're in an experience where you're [245:52] open to some new form of entertainment [245:55] and they have the ability to place a [245:57] button there with rigorous targeting and [245:59] incredible ad [246:00] Salesforce Facebook was almost built to [246:03] be the monetizable front end to the App [246:06] Store oh boy all right put a p in there [246:10] am I leading the witness too much [246:12] David put a pin in that we are going to [246:17] talk about Apple and Facebook in just a [246:19] little bit here now I said a minute ago [246:21] that the journey to 500 billion in [246:24] market cap from here is you know a [246:25] straight line if you look at the market [246:28] cap chart that's [246:29] true internally within the company you [246:32] know I think almost certainly not there [246:34] are a couple big challenges that come up [246:36] along the way during this period you've [246:38] got snap and you've got WhatsApp both of [246:41] which are like pretty [246:44] big deals y we've covered both of those [246:47] pretty fully on separate episodes that [246:49] you can go listen to but I think [246:51] ultimately what they represent and [246:53] obviously snap was more a competitor and [246:56] WhatsApp turned into an acquisition and [246:58] is now a big part of meta and part of [247:00] the company but I think together what [247:01] they represent was this next fundamental [247:05] shift in the consumer social lscape [247:08] which Mark totally picked up on and [247:11] ultimately announced as the core of the [247:13] company's product strategy in [247:15] 2019 which was the default social [247:19] behavior was moving from Town Square [247:22] which was the old Facebook Paradigm of [247:25] like Hey we're all in this together this [247:27] is our College network this is our [247:29] friends hey Network or public here's [247:31] what I think yep and in many ways [247:33] Instagram was also Town Square totally [247:36] the Instagram team noticed over time [247:39] that even before snap started eating [247:40] their lunch that engagement would [247:42] decrease the you stayed an Instagram [247:44] user cuz you over time being done with [247:47] these permanent posts and you sort of [247:49] saved it for the big announcements in [247:50] your life that there wasn't a natural [247:53] way to just effortlessly share because [247:56] when these platforms all started [247:57] everyone was in sharing with the Town [247:59] Square the whole time and everyone was [248:00] getting a little bit more clammy about [248:03] that as time went on yeah as the stage [248:06] got bigger in the Town Square it's like [248:07] well I'm not going to take the stage [248:09] unless I like really have a great [248:11] performance that I'm going to give and [248:13] so every piece of data and Metric that [248:15] they had you're right they sort of [248:16] realized oh the world is shifting from [248:19] Town Square to to living room is the way [248:21] that Mark put it that social was [248:23] shifting from Town Square to living room [248:26] right and by living room he means small [248:28] private groups of super close [248:29] relationships yep and snap and WhatsApp [248:32] were both really interesting data points [248:34] about this WhatsApp obviously with [248:36] private messaging snap also in its own [248:39] way there was the original Snapchat use [248:42] case of disappearing onetoone photo [248:45] messaging but stories really is where it [248:47] became clear hey this is widespread and [248:49] this is a big problem that is going to [248:52] threaten Instagram Facebook the whole [248:55] default motion of like sharing Beyond [248:59] just a onetoone messaging platform I [249:01] want that to be much more private and [249:03] locked down too if you own an engagement [249:05] platform and someone figures out a new [249:08] mechanic to make them much more willing [249:10] to free form share and your platform [249:13] seems to encourage them to stay back be [249:16] quiet only post once in a while maybe [249:18] lurk it's not good content creation on [249:21] the platform going down is really really [249:23] bad if you are an engagement company you [249:25] know snap represents the idea that [249:27] people are sharing way more if it's [249:30] ephemeral and WhatsApp represented the [249:32] idea that people are shifting the places [249:36] they communicate from more public to [249:38] more private and from larger groups to [249:40] smaller groups both of which are [249:42] concerning if you are a Facebook Blue [249:44] app okay so now here's an interesting [249:47] question though how would Facebook know [249:51] that attention is really shifting to [249:53] these apps because remember we're now in [249:55] the mobile Paradigm not on the open web [249:57] where services like comcore and Alexa [250:00] and whatnot exist and you can see [250:02] traffic and engagement competitively on [250:05] other apps in the mobile ecosystem it's [250:07] all siloed and locked down no just Apple [250:10] and Google really have that information [250:13] yep but somehow Facebook was able to get [250:15] pretty good insights on what was [250:17] happening in the landscape well to have [250:20] that you would either need to have some [250:21] kind of SDK that gets bundled into apps [250:23] like an analytics provider or you would [250:26] need like a VPN where like the traffic [250:28] was going through it so you could see [250:30] the traffic well in 2013 Facebook [250:33] acquired a small Israeli company called [250:36] onavo and onavo was exactly what you [250:39] just said Ben a VPN actually I think it [250:42] in itially started as like a data [250:44] compression tool so if you were worried [250:47] about your data plan which was really [250:49] important in many parts of the world you [250:51] could install onavo on your device and [250:53] it would compress the data and then [250:55] serve it to your apps eventually that [250:57] morphed into a VPN tool the net of that [250:59] is that they could actually see traffic [251:01] and usage data going into specific apps [251:04] on mobile phones in their Network yeah I [251:07] think they looked at it as we need some [251:09] way to level the playing field if we're [251:11] competing against Apple and Google in [251:12] different ways they have this data [251:14] because they own the platforms we kind [251:16] of need to be able to see those same [251:17] Trends on the other hand there is [251:20] another way to view this if you're [251:21] looking for examples where Facebook you [251:24] know may have considered their own [251:25] interests over being forthright with [251:28] users over how their data is used well [251:30] this could be another big example users [251:33] who were using onavo didn't download it [251:36] with the intent of sharing you know [251:37] their app usage data with Facebook right [251:40] they were just looking to compress their [251:42] data or for a VPN service either way [251:45] onavo ended up being like a hugely [251:48] important acquisition for Facebook cuz [251:50] it really tipped them off to WhatsApp [251:52] and snap too yeah so obviously they [251:55] fought those two companies in very [251:56] different ways one thing they learned [251:59] from fighting Twitter over the years is [252:02] that there are these social mechanics or [252:06] perhaps an interaction Paradigm might be [252:08] the right way to talk about it a post or [252:11] a like or a retweet or or a disappearing [252:13] photo message and the thing that kind of [252:17] matters is owning a valuable Network the [252:22] idea that people are going to come and [252:25] give you their attention and you own the [252:27] place that they connect with other [252:29] people that they authentically [252:31] verifiably know that's the scarce [252:33] commodity in order to kind of win the [252:35] engagement game and the mechanics are [252:38] actually kind of fungible and they're [252:41] totally a means to an end so if you [252:43] discover some mechanic and you build [252:45] this whole multi- hundred million user [252:47] Network platform ad platform based on it [252:51] that's great as long as no one comes up [252:53] with a better mechanic than you and then [252:55] goes and rebuilds the network somewhere [252:57] else so one thing they learned with [252:58] Twitter was [253:00] hm they seem to be growing really fast [253:03] with this status update thing we need to [253:06] look a lot more like status updates and [253:08] that worked pretty well there were a lot [253:10] of people that basically never switched [253:12] to Twitter because they thought oh I can [253:14] just use Facebook for this yep and I [253:16] already have my network in here I [253:17] already have my friends I already have [253:18] my connections totally and it also has [253:21] photos so like Twitter's this weird [253:22] esoteric text sharing thing I'm not [253:24] really all about that I'm just going to [253:26] keep using Facebook oh Ben Twitter did [253:29] have photos for a while it was called [253:31] Instagram well yeah that API got turned [253:34] off fast yep so as they're looking [253:37] around at snap hey someone has [253:39] discovered this new stories mechanic my [253:42] goodness [253:43] that is suddenly obvious that that is [253:45] what the future is it now feels old to [253:48] do anything else it's kind of like when [253:50] you got a Rea iPhone for the first time [253:52] and you're like holding your non-retina [253:54] iPhone and you're like this is instantly [253:55] a piece of crap like I'm not ever going [253:58] to touch this discusting thing again I [254:00] think when someone invents a new [254:02] interaction Paradigm it's one of these [254:04] things where like you have to adopt it [254:06] because otherwise people are just going [254:08] to flee and of course your business [254:09] depends on you adopting it because you [254:11] can't let someone use this new [254:13] discovered mechanic that's perfectly [254:16] timed for this moment in history with [254:18] these set of cultural acceptances and [254:20] this new set of Technologies to go [254:23] rebuild the network somewhere else and [254:25] so I think the thing that they sort of [254:27] discover is either through buying or [254:30] through copying a mechanic we need to [254:33] protect our Network by bringing these [254:37] interactions into our family of apps [254:40] either by as Ben Thompson would put it [254:42] the audacity of copying well or of [254:44] course by buying them yes copying well [254:46] or buying well yes so listeners of [254:49] course we don't actually ever know what [254:51] anybody's intent is or what they're [254:53] thinking when they decide to buy a [254:55] company or something like that this is [254:57] just sort of David and I guessing at [254:58] strategy from the outside we do have [255:00] thanks to a court case an actual email [255:04] from Mark Zuckerberg on February 28th oh [255:07] yeah this is so good 2012 to their then [255:11] CFO talking about [255:13] at the time they're sort of discussing [255:15] the Instagram acquisition but laying out [255:18] the idea behind an acquisition strategy [255:21] the basic plan would be to buy these [255:23] companies and leave their products [255:24] running while over time incorporating [255:27] the social dynamics they've invented [255:29] into our core products one thing that [255:31] may make neutralizing a potential [255:33] competitor more reasonable here is that [255:35] there are network effects around social [255:37] products and a finite number of [255:39] different social mechanics to invent [255:41] once someone wins at a specific mechanic [255:44] it's difficult for others to supplant [255:46] them without doing something different [255:48] it's possible someone beats Instagram by [255:49] building something that is better to the [255:51] point where they get network migration [255:53] but this is harder as long as Instagram [255:55] keeps running as a product pause which [255:58] is pretty interesting that is the [255:59] argument of why to keep Instagram [256:01] separate and running as its own product [256:03] is because Instagram's already [256:05] discovered this fascinating new mechanic [256:08] around publishing one image at a time [256:10] with these beautiful filters if anybody [256:12] else tries to come after them they're [256:13] already ahead so actually the best thing [256:16] to do is own Instagram and let it keeps [256:18] doing its thing anyway resuming [256:21] integrating their products with ours to [256:23] improve the service is also a factor but [256:25] in reality we already know these company [256:27] social mechanics and we will integrate [256:29] them over the next 12 to 24 months [256:31] anyway the integration plan involves [256:33] building their mechanics into our [256:35] products rather than directly [256:37] integrating their products if that makes [256:38] sense by a combination of these two [256:40] things neutralizing a potential [256:42] competitor integrating their products [256:44] with ours to improve the service one way [256:46] of looking at this is that what we're [256:48] really buying is time even if some new [256:51] competitor Springs up buying Instagram [256:53] path for square Etc now will give us a [256:57] year or more to integrate their Dynamics [256:59] before anyone can get close to their [257:01] scale again within that time if we [257:03] incorporate the social mechanics they [257:04] are using those new products won't get [257:07] much traction because we will already [257:09] have their mechanics deployed at scale [257:11] it is [257:13] godamn brilliant [257:14] David here's my commentary based on all [257:18] this it is so smart to basically say [257:21] well if we buy them we basically get two [257:25] strategies that we get to execute at the [257:27] same time one we just leave them alone [257:31] and let it keep succeeding it's actually [257:32] not a current Potential Threat we don't [257:35] know if it will turn into a Potential [257:36] Threat hell Instagram wasn't making any [257:39] money yet WhatsApp didn't have a feed it [257:41] was just a way people connected it's not [257:43] like just because they had a network [257:45] they were going to turn it into a feed [257:47] so there's this idea that it's not a [257:49] competitor Now by buying it we basically [257:51] get this option on if it becomes a [257:54] competitor if we completely leave them [257:56] alone and let them decide what to do but [257:58] then there's also look we will integrate [258:01] those mechanics into our core product [258:02] Facebook the Blue app that already has [258:04] well-built out Network effects and by [258:07] owning this thing that could become a [258:09] competitor there isn't white space in [258:11] the middle anyone just like us is also [258:14] going to take time and they're not going [258:16] to do it as good as the original and [258:18] they don't have our Network therefore [258:20] whether the winner turns out to be the [258:23] original product or us incorporating the [258:26] mechanic Into the Blue app we've won [258:27] either way and probably what's going to [258:29] happen is both yes so brilliant and that [258:32] is what happened with Instagram WhatsApp [258:34] is actually kind of different it never [258:36] really turned into a competitor it just [258:38] serves a completely different use case [258:40] and is also owned by meta and then with [258:43] SNAP they tried to buy snap over and [258:45] over again and it didn't work and so [258:48] they basically figured out how to bring [258:52] those mechanics into Facebook's core [258:54] products with stories in a way that snap [258:57] kept doing their thing but there was [258:59] really no reason to leave being an [259:01] Instagram user because you already had [259:03] that functionality with your own network [259:04] anyway and here's what's really [259:06] interesting about the stories quote [259:08] unquote story in Instagram and then [259:10] ultimately coming to the Blue app [259:13] stories as initially launched by snap in [259:18] the Snapchat app was a separate tab with [259:22] a list of your friends who had published [259:25] stories it was literally like a table [259:27] like a list oh I don't remember that [259:29] really yeah yeah yeah no it wasn't you [259:31] know the Paradigm we all think of today [259:32] as like the heads in the row at the top [259:35] of your app Instagram did that so when [259:38] Instagram copied quote unquote stories [259:41] and launched stories they launched it as [259:44] the algorithmically ordered row of heads [259:48] at the top of your app and here's where [259:52] the internal advantages that Facebook [259:55] the family of apps had they already had [259:58] this worldclass ranking algorithmic [260:01] expertise within the company that was [260:03] being applied to Newsfeed that was being [260:04] applied to ads that was already starting [260:08] to incorporate early ML and early AI [260:11] start the beginning a fair Facebook AI [260:14] research lab had already happened here [260:16] this was coming into the product well [260:18] now when you're launching a new feature [260:20] like this in this case stories you can [260:21] say oh great is this going to be made [260:23] better by our algorithmic and AI [260:25] expertise of course it is let's do that [260:28] right that's interesting basically even [260:31] if you copy the feature wholesale it's [260:33] actually even better not just because [260:35] you have the network but because your [260:37] algorithm Tech is very Advanced so you [260:38] can make sure it's the most possible [260:40] engaging version of that particular [260:42] inter action design yes so when [260:44] Instagram in those early days when it [260:46] was launching stories they were [260:49] optimizing the ranking of heads at the [260:51] top of the app by How likely you are to [260:53] tap on that person's head of course of [260:56] course they're doing that I do want to [260:58] jump forward to Tik Tok yes let's do it [261:01] because it's a continuation of this [261:02] story in a very different way in the [261:04] mid-20s here there's a bunch of other [261:06] stuff that we got to get to but Tik Tok [261:08] is particularly interesting because it [261:11] kind of blinds ited Facebook yes if you [261:15] believe that the asset that matters is [261:18] the authenticated real name network of [261:22] people you actually know or people you [261:24] care about following you kind of think [261:27] oh I always have a lead as long as I can [261:28] incorporate someone's mechanic what if [261:31] engagement is possible on an app that [261:35] has nothing to do with your social [261:36] network that's terrifying because this [261:40] big asset that you've built this durable [261:42] competitive advantage of we already have [261:44] all the people so why would you want to [261:45] go anywhere where your friends aren't [261:47] the magic of Tik Tok was AI suddenly in [261:51] the late 201s became sufficiently [261:53] Advanced that it could show you the [261:55] optimal thing created by anyone in the [261:57] world most of the time you don't [261:59] actually know that person so suddenly [262:02] the only competitive advantage that [262:03] Facebook the company has is people have [262:06] a habit of tapping on their apps when [262:08] they're bored but a h that's a pretty [262:11] thin competitive advantage that's really [262:13] easy to break habits are changeable [262:16] especially on a mobile phone where [262:18] another app is just an icon away totally [262:21] and so if someone is a great competitor [262:23] like Tik Tok who is incredibly [262:25] well-funded very good at strategically [262:28] buying ads has created their own growth [262:31] function that is they're just a very [262:33] different animal and they're a very well [262:35] executing machine and you know they obey [262:38] a completely different set of laws rules [262:41] regulation norms being a Chinese company [262:44] that created a competitor for Facebook [262:46] that is more significant than anything [262:48] they've ever faced this wasn't like oh [262:50] there might be an existential threat [262:51] from this in the future this was like oh [262:54] crap as soon as people form a habit [262:57] around oh I just open this black app [262:59] with the white little music note on it [263:02] we have no more competitive advantages [263:04] to throw at that problem so we must as [263:07] fast as possible make something like [263:09] reals just to stem the bleeding and then [263:12] we can figure out what to do from there [263:14] and hopefully get to a market stalemate [263:15] with them yeah which is basically what [263:18] has happened but interestingly Tik Tock [263:22] and bite dance it's parent company as [263:25] you point out Ben was so completely [263:28] orthogonal to everything that Facebook [263:31] had learned over the past 15 years [263:35] because yes number one most importantly [263:37] it's not social media it's media there's [263:40] no social and your graph doesn't matter [263:44] anymore that's a great point right it's [263:46] digital media it's user created media [263:48] it's Mobile Media but it's really not [263:50] social media yep doesn't matter who your [263:52] friends are on Tik Tok what matters is [263:55] in the first 10 videos you look at is [263:57] their AI pretty good at picking up the [263:58] stuff that you care about and the stuff [264:00] that you don't it's not can you find a [264:02] whole bunch of other people you know yep [264:05] and I really want to keep hitting on [264:06] this Insight again that Mark correctly [264:08] identified of social shifting from the [264:11] town Square to the living room this is a [264:14] second order effect of that shift that [264:17] the company didn't see coming because [264:19] once you shift social from the Town [264:21] Square to the living room it now becomes [264:25] possible to divorce media from social [264:28] you're already getting your social now [264:30] in private in your digital living room [264:32] the Town Square can become something [264:34] that is completely not social right it's [264:37] turned into this pretty interesting dual [264:38] pronged approach that they now have [264:40] which is you open Instagram and you have [264:43] reals so the job to be done by Tik Tok [264:46] is now done by a thing that you already [264:48] have habit around it's bundled into the [264:49] place where you actually consume your [264:51] social media great that's step one step [264:55] two is this is where it's nice to have [264:57] WhatsApp this is where it's nice to have [264:59] messenger for these living room style [265:01] conversations with just a few of your [265:03] friends the new user behavior is I see a [265:07] thing that got millions of views I can [265:09] think of one or five people to share [265:11] that with and then I have a private [265:13] conversation about this public object in [265:16] private I will share that on WhatsApp or [265:18] I will share that on iMessage or I will [265:20] share that on you know whatever my my [265:22] actual friend living room network of [265:24] choices and from where social networking [265:27] started back in 2004 and ending up now [265:31] in this reasonably complex interaction [265:33] Paradigm of I get AI served videos from [265:38] people I don't know tailored to me and [265:40] then I privately share those in multiple [265:43] groups with those who I love and care [265:45] about that is a huge Evolution it's [265:47] crazy that we use the term social [265:49] network or social media for these two [265:52] things that are like light years apart [265:54] totally the social and the media is now [265:57] completely divorced and that is because [265:59] of the shift from Town Square to living [266:01] room yeah we really need to evolve our [266:03] language around all this stuff okay so [266:05] that I think is dimension one that Tik [266:09] Tok is completely orthogonal to Facebook [266:11] dimension two is actually [266:15] organizationally so Tik Tok as we [266:17] chronicled on our episode about it a [266:19] couple years ago came out of bite dance [266:22] well it was an acquisition of musically [266:23] by bite dance and then bite Dance turned [266:24] it into Tik Tok bite dance had totia [266:27] before having Tik Tok bite dance is an [266:30] AI company bite dance is not a product [266:33] company and the core product that bite [266:37] dance makes is its AI along with all [266:41] sorts of other stuff Tik Tock is like [266:43] the product Vector one of many and the [266:46] biggest by which they deploy their [266:48] actual product which is [266:51] AI this is very very different from [266:54] Facebook which has always been a product [266:56] company and so like organizationally the [266:58] way bite dance is set up is there are [267:00] these centers of excellence within the [267:02] company where the algorithms the AI that [267:04] is controlled by a centralized team [267:08] within the company at least that's my [267:10] perception of how they're different [267:12] organizationally from how Facebook used [267:15] to [267:16] be Facebook and Facebook leadership was [267:19] obviously watching this very very [267:21] closely and if you look at how meta is [267:23] organized now it looks a hell of a lot [267:26] more like that so meta today is [267:29] organized there are centers of [267:32] excellence across the company same way [267:34] that I was just talking about bite dance [267:36] there's Fair that's the fundamental AI [267:38] research team there's infrastructure [267:40] shared across the whole company deployed [267:42] to all the products and apps there's [267:45] revenue and monetization shared across [267:47] the whole company deployed to all the [267:48] products and apps there's growth shared [267:50] across the whole company deployed across [267:52] the various products and apps there's [267:54] integrity and safety which is a like big [267:58] big big advantage that Facebook has [268:00] they've reorganized how the whole [268:03] company looks there's still all the [268:06] individual products in the family of [268:07] apps those all have their own heads and [268:10] they all roll up to Chris Cox who is [268:11] head of product for the whole company [268:13] and AI too actually reports to him but [268:16] all the infrastructure all the revenue [268:19] all the growth all the integrity and [268:21] safety are decoupled from those actual [268:23] products now and live in these centers [268:25] of excellence across the company [268:27] fascinating and I think there's a lot of [268:29] bite dance influence in that makes sense [268:31] it's the most formidable competitor [268:33] they've ever faced and they really have [268:35] no way to neutralize it yep all right so [268:38] all this talk of Tik Tock them being an [268:41] AI first company clearly meta today is [268:45] an AI obsessed company how did that [268:47] start well that is a great question Ben [268:50] but first we've got a brand new partner [268:52] to announce that we are very excited to [268:54] share with you all Huntress Huntress is [268:57] one of the fastest growing and most [269:00] loved cyber security companies today [269:02] it's purpose-built for small to midsize [269:04] businesses and provides Enterprise grade [269:06] security with the technology services [269:09] and expertise needed to protect them yes [269:11] they provide a revolutionary approach to [269:14] managed cyber security that isn't only [269:16] about tech it is about real people [269:18] providing real defense around the clock [269:21] so how does it work well you probably [269:23] already know this but it has become [269:25] pretty trivial for an entrylevel hacker [269:27] to buy access and data about compromised [269:30] businesses this means cyber criminal [269:32] activity towards small and medium [269:34] businesses is at an all-time high so [269:36] listeners huness created a full managed [269:39] security platform for their customers to [269:41] guard from these threats this includes [269:44] endpoint detection and response identity [269:46] threat detection and Response Security [269:48] awareness training and a revolutionary [269:50] security information and event [269:52] management product that just launched [269:54] essentially it is the full Suite of [269:56] great software that you need to secure [269:58] your business plus 247 monitoring by an [270:00] elite team of human threat hunters in a [270:04] security operations center to stop [270:06] attacks that software only solutions can [270:08] sometimes Miss Huntress is democratizing [270:12] cyber security by taking security [270:13] techniques that were historically only [270:15] available to large Enterprises and [270:17] bringing them to businesses with as few [270:19] as 10 100 or a thousand employees at [270:22] price points that just make sense for [270:23] them in fact there are over [270:27] 125,000 businesses Now using Huntress [270:30] and they Rave about it from the hilltops [270:32] they were voted the industry leader in [270:34] endpoint detection and response eight [270:36] times in a row and the industry leader [270:40] in manage detection and response by [270:42] customers in the G2 rankings for the [270:44] second time this summer so if you want [270:46] cuttingedge cyber Security Solutions [270:48] backed by 247 expert humans who monitor [270:52] investigate and respond to threats with [270:54] unmatched Precision head on over to [270:56] hess.com acquired or click the link in [270:59] the show notes and just make sure to [271:01] tell them that Ben and David sent you [271:04] okay so David meta today clearly an AI [271:09] obsessed company how did that start yeah [271:12] yeah so I mentioned this organization [271:14] Fair a couple times I think it's fair to [271:17] go back and tell the actual origin of [271:19] that cuz it's really interesting heyo [271:21] it's fundamental AI research originally [271:24] it was Facebook AI research and now it's [271:27] fundamental AI research given the [271:29] company is now meta so back in like 2013 [271:33] timeline when the company was finally [271:36] fixing Mobile coming out of the [271:38] hellscape of 2012 of going through that [271:41] transition during the IPO and the shift [271:43] to mobile Mark and the company's VP of [271:46] engineering Mike shreer or shrep as he's [271:49] known we're trying to figure out like [271:51] okay we've finally got a little moment [271:53] to Breathe Here the tiger isn't chasing [271:56] us at the moment can we take a step back [271:58] and try and figure out like when is the [272:00] tiger going to jump out next and try and [272:02] get us yeah or all these crazy things [272:04] are happening because we've been having [272:06] to play defense as user attention shifts [272:09] with this new paradigm of mobile how how [272:12] do we not let that happen next time like [272:14] how do we play a more active role in [272:16] whatever the future of technology is so [272:18] we're not getting whiplashed around and [272:21] we can kind of control our own destiny a [272:22] little more yeah basically not actually [272:25] owning the platform sucks we've realized [272:28] this now how do we make sure this never [272:30] happens again yes and that spawns some [272:33] of the early work on AR and VR come back [272:35] to that in a sec and even this early in [272:39] 2013 on AI and alexnet had just happened [272:44] right at this time and for whole history [272:46] of alexnet go check out our Nvidia [272:48] Series where we talk all about that and [272:50] the importance to Ai and ML and Facebook [272:53] actually had a small ml team already at [272:58] this point they were mostly doing [272:59] computer vision the early machine [273:02] learning folks at Facebook were mostly [273:04] looking at can we do automatic image [273:06] tagging to reduce the you remember how [273:09] cumbersome it used to be when you'd [273:10] upload a big batch of photos then you [273:11] had to go tag every single one and then [273:13] somehow magically I don't remember the [273:15] exact year it was like all pre- [273:17] suggested and you were like whoo this is [273:20] pretty cool that was what the original [273:21] ml Folks at Facebook were doing yep and [273:25] Alex net and those big AI ml [273:28] breakthroughs in that era were all [273:29] around image recognition and so Facebook [273:32] was kind of already attuned to this like [273:34] oh okay there's like some pretty big [273:36] technological stuff happening maybe this [273:38] is going to be like the next platform [273:41] Vector of the future we should get [273:43] serious about this AI stuff yep so [273:46] summer 2013 Mark and shrip go out to [273:49] recruit Yan laon and convince him to [273:51] come start an AI research lab at [273:54] Facebook and Yan was then and is now one [273:57] of the top AI luminaries in the field [274:00] top academics out there and Yan says [274:03] like well I'm flattered that you would [274:05] want me to come work here that's [274:07] interesting but like I live in New York [274:09] I teach at NYU I don't want to leave I [274:11] don't want to quit teaching at NYU and [274:13] I'm pretty darn committed to open source [274:16] and Publishing all the work that we do [274:19] and Mark and Sher like one sounds good [274:22] you can stay in New York two okay you [274:25] can stay at NYU to three yeah we love [274:27] open source we were built on the lamp [274:29] stack great publish everything we [274:31] haven't talked about this at all at this [274:33] very long episode so far but Facebook [274:36] had been doing a ton of Open Source [274:38] publishing in basically everything they [274:40] were doing they were open sourcing [274:42] things like Cassandra in their back end [274:44] they were open- sourcing frontend [274:46] Frameworks both on mobile and on web [274:49] react react native yep exactly they'd [274:52] been open sourcing a lot of the language [274:54] modifications they were making uh I [274:56] don't think open compute had started yet [274:59] but that was right around this time [275:00] where they were starting to open source [275:01] their data center designs data center [275:04] architecture yep Facebook definitely has [275:06] a very particular open- Source strategy [275:09] that we're going to talk about in [275:10] analysis but needless to say they have [275:12] been Advocates of Open Source since the [275:16] very beginning y so Yan is like wow well [275:20] okay when you put it that way sure and [275:22] what's really important about how they [275:23] do this is this is not Microsoft [275:26] research this is not blue sky yeah let's [275:29] go focus on anything this is a highly [275:32] focused AI research group we are going [275:35] to research Ai and I think that has kind [275:39] of made all the difference it is super [275:41] super clear exactly how that is going to [275:44] translate into helping Facebook and [275:46] Facebook's products this was pretty I [275:49] don't want to say contrarian that would [275:51] have been wrong but it's one of a dozen [275:53] specializations of computing that you [275:55] could have sort of glimpsed into for the [275:57] future yep and it was one of those [276:00] earned secrets that Facebook had and [276:04] Google had at this time you know 2013 14 [276:08] 15 of like oh we can very very [276:11] profitably apply this work to our [276:13] current products in social media feed [276:16] recommenders yes so it's very funny that [276:19] the rest of the world is having their AI [276:21] moment now CU meta had theirs in [276:23] 20134 before language models like long [276:26] before llms Facebook realized they could [276:29] profitably spend billions of dollars on [276:32] AI systems to recommend a what post you [276:35] should see next in your feed and B what [276:37] ad out of the entire inventory of ads we [276:40] could show you what is the best one for [276:41] you at this moment in time for that [276:44] Advertiser spend and today there's a lot [276:47] of companies who are spending on AI in [276:50] hopes that the use case materializes in [276:52] the future Facebook's is like wildly [276:55] proven and has been for a decade and is [276:57] incredibly profitable for them so that's [276:59] just like something to keep in mind [277:01] whenever you're looking at Facebook [277:02] talking about AI today yeah there's a [277:05] lot of future looking stuff they're [277:06] talking about but it is already super at [277:09] scale and a great business and has been [277:11] for a long long time and they've just [277:12] been you like they've been quiet about [277:14] it but no one cared until now it's one [277:18] of the fundamental advantages that [277:20] Facebook has today relative to all other [277:23] AI companies is that they are guaranteed [277:26] profitable Roi on all of their AI spend [277:30] all of their research all their [277:32] infrastructure all their gpus they have [277:36] highly highly profitable ways to put it [277:38] to work today right it is a nice to have [277:41] for meta if the next form factor of [277:43] computing is an AI assistant like meta [277:45] AI but it's not essential that's you [277:48] know one of many things that they could [277:50] do profitably with AI technology yep the [277:54] other thing you kind of got to remember [277:55] is 2012 Facebook their engineering brand [277:59] still wasn't topnotch like everyone knew [278:03] that they were a upand cominging great [278:07] startup that went public but they [278:10] weren't necessarily rep able in the [278:12] computer science Community as these [278:14] people are Microsoft level researchers [278:18] to advance the State ofth art or Google [278:20] level in the way that Google always was [278:22] from the beginning yes and I think Mark [278:24] and shre had to make this point to Yan [278:27] and then when Yan joined it kind of made [278:29] the point to the rest of the industry [278:31] that this is not something that's going [278:33] to like Wax and Wayne with our financial [278:34] results were in for a decade on this [278:38] particular area of research it's kind of [278:40] amazing it really only took like two [278:42] years to bear fruit for them before they [278:43] had profitably deployed AI systems in [278:46] production CU most of the time these [278:47] research projects take much much longer [278:49] it's funny actually thinking through [278:51] this now I think there are three [278:53] dimensions to the payoffs from this [278:56] investment in Ai and fair back in 2013 [279:00] one is the near-term and immediately and [279:03] ongoing profitable paybacks to [279:05] incorporating AI in existing Facebook [279:08] products and social media feed [279:09] recommenders highly highly profitable it [279:12] just makes every ad that's displayed [279:14] more likely to make money and it makes [279:16] every feed post that is displayed more [279:19] likely to get engagement yes 100% okay [279:22] then there's two which is this like hey [279:26] AI might be the next Computing platform [279:28] there's a lot of interesting stuff going [279:30] on this allows us to have a piece on the [279:32] chessboard and be a player in the next [279:35] platform transition should it be to Ai [279:37] and this is just like what meta did with [279:40] threads with WhatApp you know with [279:42] threads they've said we're not [279:42] monetizing it right now we're going to [279:44] see if it becomes a big close to billion [279:46] user platform and if it does then great [279:48] we'll figure out the right monetization [279:50] strategy just like we did with WhatsApp [279:51] it may or may not be ads specifically [279:54] but we build things that get engagement [279:56] and then we later figure out how to make [279:57] money on them yep and then three they [279:59] had no way of seeing this at the time [280:02] but fair and AI actually becomes the way [280:05] that they can catch up and re architect [280:08] to compete with Tik Tok via reals yeah [280:11] you know no Ai No reals that's such a [280:13] good point if not for fair Tik Tok could [280:17] have disrupted Facebook totally I mean [280:21] reals would not have come out as fast as [280:23] it did right Facebook would not have had [280:24] that near-term weapon to go huh we need [280:28] something that kind of stops our users [280:31] from leaking out and jumping over to [280:33] this other app you know for a use case [280:37] that we basically can't match I mean it [280:40] is mat [280:42] how the reals and Tik Tok algorithms [280:44] work and now YouTube with shorts that [280:46] very quickly they do figure out how to [280:48] put things in front of you that are [280:50] incredibly engaging just for you I mean [280:52] they call it the for you page and that [280:54] have no social connection to you yes [280:56] exactly you're right the decision looks [280:58] more preent than ever given that totally [281:01] okay so that's fair in AI let's jump [281:04] forward now to 2016 2017 2018 and I [281:08] think for our purposes here today [281:11] it's interesting to view the 2016 [281:14] election Cambridge analytica all the [281:17] Fallout that comes out of that which [281:19] much much much ink has been spilled [281:21] through this lens of the consumer social [281:24] transition from Town Square to living [281:26] room because I think it actually [281:29] explains a lot o all right I'm curious [281:31] to hear what you're thinking here but to [281:33] catch us up here's kind of what happened [281:36] until 2015 is Facebook could kind of do [281:40] no wrong other than all the data privacy [281:42] issues that they had sort of worked [281:44] through with the FTC in 2011 but they [281:47] are now increasingly finding themselves [281:49] in very controversial situations that [281:52] they are ill-prepared for as sort of all [281:55] of humanity is now uh joining the [281:57] network first of all content moderation [282:00] is becoming an issue oh yeah for sure [282:03] Facebook is trying to figure out their [282:04] role in this are they a neutral platform [282:07] do they need to kind of Poli what is [282:10] said on their platform on top of that [282:12] actually defining what is allowable [282:15] speech on their platform is becoming [282:17] trickier than ever and it is hard to [282:18] create one set of rules and abide by [282:21] them globally and Facebook was just [282:23] caught flat-footed in a big way by the [282:25] fire hose of questionable content that [282:28] people would post ultimately they end up [282:30] throwing huge amounts of people at this [282:32] hiring tens of thousands of contractors [282:34] to deal with content moderation building [282:36] out really sophisticated kind of [282:38] policies and programs and escalation and [282:41] review processes they have an Ever [282:44] escalating set of posts videos and live [282:46] streams shared around the platform [282:48] that's right CU it's not just text [282:50] anymore it's obviously photos but like [282:52] videos now too reviewing realtime video [282:55] absolutely and as you would imagine they [282:57] are not perfect at this so they are [282:58] creating headlines left and right about [283:00] objectionable things being shared on the [283:02] network in Building Technology that [283:05] enables everyone around the world to [283:07] communicate with each other and organize [283:09] they also have the flip side of [283:11] anyone can share anything and if it's [283:13] engaging enough there's a ton of [283:15] eyeballs that see it so misinformation [283:18] is starting to become a big deal too [283:20] ultimately this comes to a head in the [283:22] 2016 to 2018 period after they have [283:25] Whitted away a lot of the Public's [283:27] Goodwill and we're just going to zoom in [283:29] on this moment for the company the 2016 [283:32] election happens and a lot of people are [283:34] looking around for an answer to wait a [283:37] minute yeah this isn't the outcome that [283:40] I expected [283:42] what happened exactly there's a lot of [283:44] people who feel this couldn't have [283:45] happened on its own legitimately you [283:47] know somebody did something and one [283:51] thing that absolutely did happen is [283:53] there was a firm called Cambridge [283:55] analytica that sold their services to [283:58] the Trump campaign and had a methodology [284:02] to create psychographic profiles that [284:04] they believed would work that were [284:06] derived from a Facebook quiz application [284:09] this ends up being an absolute Honeypot [284:12] of a story for anyone who is fired up [284:14] about a big change to the country that [284:16] they are not excited about yep and so [284:19] there were basically four stories [284:22] concurrently happening of how Facebook [284:26] was involved in the election one there [284:29] was Russian interference yep two people [284:32] are spinning up fake news sites not [284:35] necessarily for politics but actually [284:38] for the ad dollars generated on them [284:41] three Cambridge analytica four the Trump [284:45] campaign may have been really good at [284:47] using Facebook's digital marketing tools [284:50] yes I think that is like undeniably true [284:52] yeah so there's been years of [284:53] Investigation into each it seems like [284:55] that last one is actually the biggest [284:56] lever yes Russians did spin up fake [284:59] pages and accounts with pretty modest [285:01] spend and people absolutely spun up fake [285:04] news sites and generated fake viral [285:07] stories for politics and also profit [285:09] Cambridge analytica yes it was a thing [285:13] but they ended up with actually pretty [285:16] crappy derivative data from a quiz not [285:19] the Treasure Trove of raw Facebook data [285:22] on you and your friends that everyone [285:23] feared but David your point really what [285:25] happened is the Trump 2016 campaign just [285:28] actually took meta platform seriously [285:30] and got really really good at using the [285:31] tools yeah quite ironic in that Obama [285:35] totally doubled down on Facebook as a [285:37] strategy and used it to great effect to [285:39] win the8 2012 elections yes so as you [285:43] can imagine I'm preparing for this [285:45] episode and I'm like okay I just heard [285:47] Mark Zuckerberg say on another podcast [285:50] by the way great interview with Alex [285:51] Heath at The Verge after meta connect [285:54] about this he dropped this line people [285:56] thought that all this data had been [285:58] taken and it had been used in this [286:00] campaign and it turned out the data [286:02] wasn't and the data wasn't even [286:04] accessible to the developer so David I'm [286:07] listening to that and I stop because I'm [286:09] thinking well that's not what anybody [286:11] thinks because there was a ton of [286:13] reporting on the fact that they did so [286:14] what actually happened here so here are [286:17] the chain of events back in the 2010s [286:19] Facebook built the platform that we [286:20] talked about that was incredibly [286:22] permissive in the early days when you [286:23] offed your profile against an app you [286:25] gave it all of your profile data and in [286:27] those earliest days it could actually [286:29] see your whole friend list too Facebook [286:31] eventually made their API much more [286:32] closed off but it took a couple years to [286:34] do that now there was a terms of service [286:37] that required app developers to delete [286:39] any data they had after a period of time [286:41] and it'll use it for very specific [286:42] purposes but of course Facebook couldn't [286:44] actually guarantee that people were [286:46] complying with the terms of service it [286:47] was just you know you were a breaching [286:49] contract so at some point a developer [286:51] made an app that was a quiz users could [286:54] opt into taking that quiz the quiz asked [286:56] you questions and by combining the [286:58] answers to those questions with the [287:00] profile data you authenticated from [287:02] Facebook that app developer then tried [287:04] to label you with a pretty basic [287:06] psychographic profile and then [287:08] ultimately that derivative data the lab [287:11] of personality characteristics that were [287:13] derived from Facebook data and your quiz [287:15] answers is what Cambridge analytica had [287:19] as best I could tell so I read a good [287:21] amount of the report that the UK [287:22] government put out about this the [287:24] conclusion was that the quiz was taken [287:26] by 320,000 people since those people had [287:29] access to friends the total surface area [287:31] was the public profile data of 87 [287:33] million Facebook users they also found [287:36] that Facebook had requested that [287:37] Cambridge analytica delete all the data [287:39] both the actual Facebook data and [287:41] anything derived from the Facebook data [287:43] back in 2015 before the election even [287:45] started so it was yes true an [287:48] application got access to Facebook data [287:50] on a lot of users is also true that it [287:52] violated terms of service to use that [287:54] data for other purposes and they did not [287:56] comply with a delete request or they may [287:59] not have complied it's kind of difficult [288:00] to [288:01] reconstruct on top of all this it [288:03] supposes that Cambridge analytica's [288:05] method of taking these quiz answers and [288:07] translating them into something that [288:08] could impact voting worked like it was [288:11] effective and that's pretty thin right [288:15] that's kind of the most thin thing of [288:16] all of it right so here's what the UK [288:18] report concluded and this is in [288:20] government speak so it's all very hedged [288:22] while the models showed some success in [288:24] correctly predicting attributes on [288:26] individuals whose data was used in the [288:27] training of the model the real world [288:29] accuracy of these predictions when used [288:31] on individuals whose data had not been [288:34] used in the generating of the models was [288:36] likely much lower through our analysis [288:39] of internal company communic ations the [288:41] investigation identified there was a [288:43] degree of skepticism within Cambridge [288:46] analytica as to the accuracy or the [288:48] reliability of the processing being [288:50] undertaken there appeared to be concern [288:52] internally about the external messaging [288:54] when set against the reality of their [288:56] processing so at the end of the day the [288:59] Cambridge analytica thing in particular [289:01] was kind of a nothing Burger but that is [289:04] not the story that gets told the UK [289:06] regulator was given full access to this [289:08] and ended up being quite skeptical that [289:10] the methodology even worked in the first [289:13] place so they've got this like kind of [289:15] crappy derivative data not complying [289:17] with terms of service requests to delete [289:19] it and the method may not actually work [289:23] now this doesn't exonerate Facebook in [289:25] any way from what could have happened [289:27] they had an incredible Treasure Trove of [289:29] data that apps could slurp up in the [289:31] early 2010s but the actual Narrative of [289:34] Facebook data that Cambridge analytica [289:36] had impacted election results is like [289:39] wildly oversold [289:41] yep all that aside as we were waiting [289:44] through all of this and trying to make [289:45] sense of it and think about how we're [289:46] going to talk about this on acquired [289:48] there is such a enormous enormous [289:52] disconnect between reading what actually [289:55] happened and then how people felt and [289:58] what the like lived reality of the brand [290:02] impact on Facebook was which that was [290:05] huge yeah nothing could have been Huger [290:09] to me the story here is there sure was a [290:12] lot of ill will and discontent with the [290:15] Facebook brand that people were willing [290:18] to dive in head first on it's almost [290:21] like Facebook isn't guilty of whatever [290:23] the heck people think happened with [290:25] Cambridge analytica all these years [290:27] later but something is very very wrong [290:29] that like hundreds of millions of people [290:31] believe that yeah and it's not just [290:33] wrong with the system they definitely in [290:36] behaviors along the way manag to earn [290:39] people's distrust I think that's the [290:41] issue so let's bring it back to my [290:44] supposition here which is that I [290:47] actually think this paradigm shift that [290:49] Mark identified from Town Square to [290:51] living room was like way more right and [290:54] way more powerful than even he realized [290:58] and this is one of the second order [291:00] effects of it the quiz that was the app [291:02] on Facebook that happened way back in [291:05] the early platform days totally solidly [291:08] in the Town Square days and then [291:11] the election happened in 2016 when [291:13] things are already starting to [291:15] transition WhatsApp exists snap exists [291:19] snap has launched stories Instagram has [291:21] now copied snap and launched stories in [291:23] that very same year two more years go by [291:26] the camridge analytica news doesn't [291:28] actually hit until [291:30] 2018 by 2018 we are solidly in the [291:33] living room era so all of the supposed [291:37] real or imagined impacts of this [291:40] Cambridge analytica thing had their [291:42] Roots back in the Town Square era we're [291:45] now here in the living room era and that [291:47] is what people's expectations are of [291:50] Facebook of social media of everything [291:53] that they are sharing in their lives and [291:56] I think this is to my mind at least [291:58] where the disconnect is people's [292:01] expectations shifted but now all of a [292:04] sudden all this stuff from the past [292:05] becomes really really relevant to the [292:07] present and when you look back at the [292:09] past through this l of what the present [292:11] is you're like well this is really effed [292:13] up like how the hell did the company do [292:15] this right and you know the answer was [292:18] they were trying desperately to grow [292:20] they were trying to will a platform into [292:21] existence of course the cultural norms [292:24] were a little bit more permissive but [292:25] like the cultural norms were really [292:26] never permissive to the point to say [292:29] sure you should download and store a big [292:31] Json dump of my whole profile [292:33] information and any pages that I like [292:36] and anybody whose names I'm connected [292:38] with like that always was probably a [292:41] little bit of an overstep but then once [292:43] developers were able to do that and then [292:45] violate terms of service and store it [292:46] forever then it just kind of becomes [292:48] this like ticking Time Bomb that's out [292:49] there exactly it's like this weird [292:52] temporal and cultural like mismatch of [292:55] it all Landing in 2018 when the world is [292:58] a very very different place in part you [293:01] know you can see why Facebook massively [293:03] prefers the ad-based business model of [293:06] we never even expose any information to [293:08] you Advertiser you target and then we [293:11] just deliver the ads you never get to [293:13] know anything about who the users are on [293:16] our platform we don't sell data like [293:18] that is a true fact whenever they stand [293:21] up there and say we don't sell data they [293:22] don't they sell the opportunity to [293:24] Target users but in fact if they did [293:26] sell data it would be a bad thing [293:28] because someone that could then build a [293:29] competitor like the asset that they have [293:32] is actually their data that they choose [293:34] not to sell for business reasons y [293:36] totally then there's this whole other [293:38] layer too here of like well why why did [293:41] they do this in the first place why did [293:43] developers have access to so much data [293:45] they were building platform they were [293:46] building an entirely different business [293:48] with a different strategy now they are [293:51] an advertising and media based business [293:53] we're almost talking about two totally [293:55] different companies here as I think [293:56] we've told in the story yeah I do think [293:59] when I search my heart of hearts though [294:01] when I was developing apps in the 2010 [294:04] period and would let people use Facebook [294:06] connect or whatever the current branding [294:08] of accessing Facebook accounts was at [294:10] the time I was always like whoa this is [294:13] a lot of data that comes with someone [294:15] clicking that button well I think you [294:17] made the point right of like Facebook [294:19] had to kind of sell the platform to [294:21] developers totally did they didn't have [294:24] the underlying hardware and technology [294:26] and operating system to offer so there [294:28] had to be a really big sweetener yep I [294:32] just can't believe it though after all [294:34] these years and all this writing and all [294:37] these headlines if Cambridge analytica [294:39] were more [294:40] competent they probably could have had [294:43] more impact but as it stands I just [294:46] can't believe there's not a big story on [294:48] the front page of major newspapers with [294:50] big headlines saying actually Cambridge [294:53] analytica didn't really have that much [294:54] in the first place their methodology [294:56] appears to have not worked and mayaula [294:58] sorry for getting yall all riled up if [295:01] you don't want to like Facebook sure but [295:03] go pick a different reason okay so what [295:05] happened with all this it is worth [295:07] knowing to finish the story earlier from [295:09] the consent decree in 2011 that said [295:11] Facebook is going to be carefully [295:13] monitored for privacy concerns as you [295:15] can imagine the Cambridge analytica news [295:18] coming out put a gigantic bullseye on [295:21] Facebook and the FTC said aha you guys [295:23] signed a consent decree a mere seven [295:26] years ago what the hell this is exactly [295:30] what we're talking about and Facebook [295:33] just decides you know what what's the [295:35] big dollar sign that check that we can [295:36] write here because there it seems like [295:38] we just want this to be done we want [295:40] this to be buried we're a different [295:42] company now we kind of want to put that [295:43] to bed the whole settlement here was not [295:46] specifically about Cambridge analytica [295:47] there were a bunch of other things that [295:49] are privacy related but the number is $5 [295:53] billion and David there is one thing [295:56] that it comes with and that is a [295:59] 20year window that they are monitoring [296:03] Facebook and that is exactly matching to [296:06] the words that Mark gave us on stage [296:08] that Cambridge analytica the way they [296:09] handled was a 20year mistake yep now I [296:14] think by the 20-year mistake he meant [296:16] much more of like the damage to [296:18] Facebook's brand yes that came out of [296:21] this yes and specifically the FTC [296:23] settlement yep it is funny I was [296:26] thinking this is a very um Facebook [296:28] appropriate comment if we had subtitles [296:30] for this episode this one would be it's [296:34] complicated yes so great it really is so [296:37] great it really is it really is and what [296:38] it comes down to is I think there is an [296:42] ethos at Facebook that at many steps [296:47] along the way there was an opportunity [296:49] to grow to win to compete and What [296:55] mattered was winning and then when the [296:57] dust settles you can kind of look around [296:58] and say okay what was the impact of that [297:00] that is one way to run a company the [297:03] other way is to be really careful and [297:04] ask permission Facebook fell into the [297:07] first camp and they probably wouldn't be [297:10] here today or certainly at least [297:12] wouldn't be a global player with 4 [297:13] billion users today if they hadn't [297:15] fallen into the first camp but by [297:17] falling into that first Camp you have [297:19] stuff that comes up where you're like ah [297:22] crap I wish we hadn't done that and this [297:23] is one of them right the time bombs [297:25] totally okay so what actually happens [297:29] Facebook becomes a lot less valuable [297:31] after all of this comes out the company [297:33] announces in their July 2018 earnings [297:36] call that they're going to be more [297:37] focused on user privacy and that Revenue [297:40] may slow as they make this intense Focus [297:43] they dropped 19% in one day o they wiped [297:47] out 119 billion dollars in market value [297:49] which was the largest single day loss [297:51] for any company in history at that time [297:55] at that time save at that time yeah to [297:59] be exceeded in the future so this whole [298:03] privacy issue the $5 billion settlement [298:06] all this stuff that just happens with [298:07] meta it has big ripples for for the [298:10] whole tech industry there's a phrase [298:14] that is known on the lips of every [298:16] American somewhere in the neighborhood [298:19] of Cambridge analytica election [298:21] interference fake news I hate social [298:23] media Facebook did this YouTube did that [298:26] blah blah and this is a persistent drum [298:29] beat that is underneath the whole [298:31] National discourse there is an [298:33] opportunity to be the anti faceb here if [298:37] you want to be to be the company that [298:38] says we are so unbelievably incredibly [298:41] about privacy even more than we ever [298:43] have been before David who does that [298:46] that is Tim Cooks Apple yes so Apple a [298:52] company that I love and adore and was [298:54] actually wearing an apple t-shirt [298:56] earlier today uh and had to take it off [298:58] before recording because I felt like it [298:59] was weird to be wearing my Apple t-shirt [299:01] when recording the Facebook episode [299:03] apple is a company that has always been [299:04] incredibly privacy conscious they both [299:08] use that as something they believe [299:11] deeply in their soul and design into [299:12] their products and is amazing for all of [299:14] us who use their products to get that [299:16] privacy and have that trust they also to [299:19] use Ben Thompson's Parlin use it as a [299:21] strategy credit there are areas in which [299:23] it really behooves them because they [299:25] don't need to do server side stuff they [299:27] don't need to do advertising because [299:29] they make a lot of advertising [299:30] indirectly from Google and they don't [299:31] need to do a lot of other things so they [299:34] can tout hey we're unbelievably secure [299:37] with your data we take privacy [299:39] incredibly seriously no one takes it [299:40] more seriously than us they start really [299:42] beating this drum and by 2021 they [299:45] decide you know what in iOS [299:48] 14.6 we are going to launch a new policy [299:51] called app tracking transparency yep at [299:55] and what that means is a few years ago [300:00] we mandated that anybody who is tracking [300:04] you across apps start using something [300:07] called the [300:08] idfa the ID identifier for advertisers [300:12] now in the past Apple had let you just [300:15] actually reach in and grab the device's [300:16] unique identifier which was pretty cool [300:18] because it as a developer it was [300:20] unrelated to advertising think of it [300:22] almost like a serial number of this [300:23] device and you could use that for things [300:26] like hey is this the same user across [300:28] multiple applications I've got my SDK [300:30] and my code running in multiple [300:32] developers applications so I can do [300:33] interesting things like say hey this [300:35] person both takes runs with straa and [300:37] they also use Google Maps and and you [300:40] know you can just kind of gather data [300:41] across apps kind of similar to how [300:43] Facebook was gathering across the web [300:45] with like buttons everywhere or with [300:47] Facebook connect everywhere so they [300:49] could build this holistic profile of [300:50] things you do off of Facebook yep so [300:54] Apple stops letting you use the device [300:55] identifier they force you into using [300:57] this idfa and then with iOS 14.6 in 2021 [301:02] they say hey if you're using idfa part [301:05] of that API is that now whenever someone [301:07] launches an app for the first time it's [301:10] going to ask them in this really [301:12] aggressive language are you okay with [301:14] getting tracked or do you want to ask [301:17] this app not to track what do you think [301:19] people are going to click [301:22] yeah there's a default answer to that [301:24] question yes so what actually ends up [301:27] happening most people click ask app not [301:29] to track a whole lot of Facebook's [301:33] targeting basically falls apart they no [301:36] longer have a picture of you outside of [301:39] app that they actually own and a lot of [301:42] the reason why advertisers can get so [301:45] good at targeting is because of this [301:46] holistic picture that is built for you [301:48] across your phone David this is the [301:50] example kind of manifest in practice of [301:52] the thing you were talking about all the [301:53] way at the launch of mobile of what's [301:55] one of the reasons why Facebook's [301:57] beautifully constructed business model [301:59] doesn't work in the mobile wal Garden [302:01] ecosystem it's because the operating [302:03] system maker can make a change like this [302:05] that just kind of affects core [302:08] functionality that you were relying on [302:09] and now you don't have access to that [302:10] data so you can't run as effective of an [302:12] advertising service totally and a minute [302:15] ago when you asked the question of who's [302:17] banging the Privacy drum and I said it's [302:19] Tim Cook's apple one of the reasons I [302:21] said that is that actually I didn't know [302:25] this till doing the research Mark and [302:28] Steve Jobs had a great relationship [302:31] which makes this story all that much [302:33] more interesting Mark is very stevik [302:36] very very stevik and they actually met [302:39] for the first time in 2007 right as all [302:43] of this is coming together and the first [302:45] meeting was that Steve actually reached [302:47] out to Mark and wanted to meet him and [302:48] build a relationship which I don't think [302:50] Steve did that with too many people and [302:52] then so on the back of that the next [302:55] year in 2008 when the iPhone SDK was [302:58] launching and people were building apps [303:00] Mark and Facebook actually go to Apple [303:04] and meet with Steve and they pitch Steve [303:06] on exactly what you were just talking [303:08] about Ben like hey how about you let us [303:11] bring our platform that we've just built [303:13] that has all these developers into iOS [303:18] into your platform really oh to make [303:20] Facebook platform part of the iOS [303:22] Developer whoa yes which you know from [303:26] Facebook's part is kind of like a hail [303:27] marry like yeah they're probably not [303:28] going to say yes here but this is really [303:30] big risk to us we might as well try but [303:33] they did have enough of a relationship [303:34] where Facebook and Twitter both were [303:36] privileged citizens on an early version [303:39] of the iPhone OS like in the setting [303:41] screen even before you installed any [303:43] apps there was like a Facebook and a [303:45] Twitter settings for I guess like native [303:48] integration between the operating system [303:49] and those networks and social networks [303:52] yes totally so Steve supposedly the [303:55] legend goes you know kind of Mark and [303:57] the Facebook team make this whole pitch [303:59] and at the end of it Steve just looks at [304:00] Mark and it's you know like almost a [304:03] little sheepish and it's just like thank [304:06] you we're just not going to let somebody [304:08] else build a platform on top top of our [304:10] platform and then he sort of pushes back [304:12] in his chair and it's like clear that's [304:13] the end and he's like hey so by the way [304:16] what do you guys think about Microsoft [304:17] and working with them and then they just [304:19] like have a really nice conversation [304:20] supposedly for the next couple hours and [304:22] shoot the [ __ ] and talk about Microsoft [304:24] and Computing history it's this very [304:26] cordial relationship and I think the [304:29] sense I get is that Steve really liked [304:30] Mark and respected him well they're both [304:32] like pretty product Visionaries and [304:35] they're both like very stubborn about [304:37] their views of the future and they both [304:39] were very right about their views of the [304:40] future I can see it totally and clearly [304:45] that is not the feeling today between [304:47] the two companies and I think what [304:48] happened since okay if it had just ended [304:52] at Apple isn't going to let Facebook [304:54] build a platform on top of their [304:55] platform that might have been it but [304:57] then when mobile app installs happened [305:00] and Facebook all of a sudden Facebook [305:02] was making billions of dollars off of [305:06] deciding what apps in the app store get [305:08] downloaded Apple had to have felt like [305:10] hey this is actually this is our turf [305:12] ours yeah yeah and so then after [305:16] everything that happened in 2016 2018 [305:18] Cambridge analytica privacy to your [305:21] point Ben part of this is that Apple has [305:23] always cared about privacy and it's [305:24] always worked well for them and it's [305:25] part of their main product strategy part [305:27] of it too is like oh great here is kind [305:31] of the ultimate competitive Vector that [305:34] we can have against somebody we really [305:36] don't like who we feel has profited off [305:38] of our platform unfairly yeah you end up [305:40] with apple feeling like everything that [305:43] happens on our platforms is ours and [305:45] these are our users to protect and no [305:47] one's going to do wrong by our users in [305:49] any way so you have apple who is [305:52] protective as all hell and then you have [305:55] Facebook where Mark Zuckerberg more than [305:57] anything in the world wants as much [305:59] freedom to operate as possible and you [306:01] have Apple trying to constrain and you [306:03] have Mark who hates feeling held captive [306:06] yep so now here we are in 2020 and [306:10] Apple's like we're making our move yeah [306:13] Facebook's been reeling we've been [306:14] beating the Privacy drum for the last [306:15] couple years since 2018 in Cambridge [306:18] analytica now it's time to make the move [306:21] so what ends up happening actually the [306:23] first couple quarters not much [306:25] Facebook's talking about it on earnings [306:26] calls hey we think this is going to have [306:28] impact it's not huge but then February [306:31] 2022 which is technically the end of [306:34] year 2021 earnings call for meta they [306:38] dro the bomb interestingly in question [306:41] and answer I listen to the whole [306:42] earnings call and the CFO kind of [306:44] casually says in a response to someone [306:46] oh we think that the impact from ATT to [306:50] our ad business will end up costing us [306:52] on the order of $10 billion for [306:54] 2022 what that is eye popping like uh [306:59] okay so what ends up happening is [307:02] there's a 26% draw down in a single [307:06] day remember we said to put a pit in the [307:09] uh single day draw Down Records the [307:11] market cap went from 900 billion to 700 [307:14] billion the actual number is they lost [307:16] 232 billion in market cap the new [307:19] largest in history surpassing their [307:21] previous record also on this call they [307:25] announced their first ever quarter over [307:27] quarter user decline keep in mind what's [307:29] happening is on top of meta basically [307:32] saturating most of the Internet [307:34] connected World by this point Tik Tok is [307:37] also really really peing so [307:40] he has some marginal users kind of using [307:42] the app less because they're moving to [307:43] Tik Tok and Facebook is reacting to Tik [307:46] Tok and trying to put reels in so [307:48] they're cannibalizing their own Revenue [307:50] by encouraging people to watch these [307:52] short form AI recommended videos that [307:54] actually don't yet monetize as well as [307:56] the newsfeed there's kind of three [307:58] fronts that are destroying them here [308:00] there's ATT there's Tik Tock competition [308:03] and there they're making Revenue [308:05] cannibalizing changes in their own app [308:08] so the hit continues it traded all the [308:11] way down [308:12] 46% by April 27th and then ultimately it [308:16] bottomed on Halloween that year so this [308:19] is what eight months later with a 72% [308:24] draw down oo they lost 72% of their [308:28] value between February and Halloween o [308:32] all in the year 20122 it's crazy and [308:36] that is like Nvidia levels of draw down [308:38] back in the crazy Nvidia Journey that we [308:40] talked about it's totally insane the [308:43] interesting thing is at this time the [308:46] real threat wasn't actually ATT the [308:50] biggest of those three threats was Tik [308:52] Tok stealing users yes the whole at [308:56] thing is about how much money can we [308:59] make off of an ad because it is so well [309:01] targeted it's an optimization that is [309:04] useless if you do not have users to [309:06] advertise to in the first place yes so [309:09] the actual real existential threat is [309:12] Tik Tok and Ben Thompson makes this [309:14] really great point and I know I've [309:16] quoted Ben over and over this episode [309:17] but I think he's just been so astute on [309:19] Facebook at many points throughout [309:21] history this decision to make these [309:24] product changes to respond to the Tik [309:25] Tok threat in the face of ATT to do [309:27] these at the same time is a founder-led [309:30] decision if you had a professional CEO [309:33] the correct thing to do to preserve your [309:35] job and shareholder value is to wait 6 [309:38] 12 months before you start reacting to [309:40] Tik Tok to let the whole ATT thing blow [309:43] over Mark's like I can't get fired I [309:46] think the right thing to do is react to [309:48] Tik Tok now because every day this [309:52] problem compounds and gets absolutely [309:53] worse I don't care that there's this [309:55] horrible narrative going on right now [309:57] with ATT that is going to cause us [309:59] apparently to lose 10 billion of Revenue [310:00] we otherwise would have gotten this year [310:02] we must cannibalize Revenue in addition [310:04] to that to make these product changes [310:07] what it ends up with is a 72% draw down [310:09] and what it also ends up with is the [310:11] chance for it to 5x from there which is [310:14] what has happened the company saved [310:18] Itself by acting correctly in this and [310:21] they had to go through this wild [310:23] tumultuous two-year journey in the [310:25] process it's just another episode of [310:30] Mark making these sets of decisions [310:32] because he's mark because he controls [310:34] the company and because he has the [310:36] stomach to withstand it yep and what [310:39] ended up actually happening from ATT [310:41] it's been value destructive overall [310:44] because I don't think the amount of [310:46] money that has shifted away from [310:47] Facebook has been captured by Apple I [310:50] get the sense that the app install [310:53] business for the App Store those search [310:54] ads is going great but it's not like [310:57] equivalent to what the monetization over [310:59] at meta is on app install ads or was on [311:02] app install ads y I think Apple was sort [311:04] of hoping in its heart of hearts that [311:08] this would shift a lot of those dollars [311:11] into ads in the App Store and the [311:13] reality is people just don't go to the [311:15] App Store searching for apps anymore not [311:17] to mention this knapped meta on a lot of [311:20] ads that are not app install ads this [311:22] knapped a lot of the like we have a [311:24] friend that's in the retail business who [311:26] was saying that their ads doubled in [311:27] price in 2022 when this happened and [311:31] they were sort of scratching their head [311:33] and they have nothing to do with the App [311:34] Store so all that kind of happened is [311:38] that entrepreneur said said I guess I'm [311:40] going to keep advertising on Facebook as [311:42] a platform I hope it gets better I'll [311:44] pick some other platforms but like I [311:45] can't advertise with apple there's [311:47] nothing for me to do it's a retail good [311:48] it's a physical item so I guess I'll [311:51] just pay more money to acquire customers [311:53] now like it ends up actually hurting the [311:55] business when go to market channels get [311:58] less efficient unless you can sort of [312:00] shift that spend to a place where your [312:01] customer is also hanging out yep in [312:05] practice meta is fine when a shift like [312:07] this happens the most scale player with [312:10] the most engineers and the most data [312:12] turns out they're still the winner meta [312:15] launched this thing called Advantage [312:16] Plus and now if you're an Advertiser [312:19] they use a whole bunch of other data [312:20] signals and you still have a customer [312:22] acquisition budget you're going to spend [312:24] it you're probably going to go spend it [312:25] on meta the same where you were before [312:27] and it's just not quite as efficient as [312:28] it was oh and by the way anyone else [312:30] that got hurt from the changes with at [312:34] just doesn't have the ability to react [312:36] and build around it the way that meta [312:38] had because they don't have the [312:39] resources meta does yep so all of this [312:43] finally brings us to the last piece of [312:45] the story here that I'm sure listeners [312:48] at this point you are probably saying [312:50] like what the heck we're however many [312:52] hours in you guys haven't said the words [312:54] reality Labs or Oculus or ar or VR once [313:01] oh is Facebook in that industry [313:03] yeah yeah you know I guess they are all [313:06] right so what happened here uh way back [313:09] back in call it 2012 Facebook starts [313:13] getting interested around the same time [313:15] that they're starting fair for AI and [313:17] what is the platform of the future they [313:19] start doing some Hardware prototyping on [313:20] their own they're not really set up for [313:23] that yet but they do want to do the same [313:25] sort of approach that they did with AI [313:27] which is focused research not general [313:29] research pick a particular thing where [313:31] we have an opinion about something we [313:33] think is going to be the future and then [313:35] invest deeply in it yep so a crazy thing [313:38] happened in February of 2014 they opened [313:42] up the purse strings and paid $19 [313:44] billion for WhatsApp they had so much [313:47] conviction that the future was AR and VR [313:51] that 34 days later they paid another 2 [313:54] billion for Oculus I had no recollection [313:57] that these things were a month apart no [314:00] I didn't I knew they were both in 2014 [314:01] but I didn't realize they were a month [314:03] apart it's wild I think Mark got the [314:05] demo of oculus and was just like oh this [314:09] has the most credible potential to be [314:11] the future out of anything that I've [314:13] ever tried so yes but I think actually [314:16] the right way to think about this is [314:18] through the same lens as Fair it's a [314:20] research lab in the same way that fair [314:22] is a research lab and it was like oh [314:24] great two billion dollars to Kickstart [314:26] the research lab fantastic all right so [314:30] you and I have both tried Orion it is [314:33] unbelievable [314:35] fantastic very clearly a path to the [314:38] next generation of computing device I [314:40] don't know if this is going to be the [314:41] winning company I don't know if that's [314:42] going to be the exact winning device but [314:44] like never have I been so sure that a [314:47] main frame very far away from me to [314:50] converting to a PC which lives 3 feet [314:53] from me converting to a phone that lives [314:56] in arms length from me the next logical [314:58] step is glasses that live on my face and [315:01] that used to sound ridiculous and then [315:03] you and I tried Orion and now I'm like [315:05] oh yep that is going to replace or [315:07] augment my phone in whenever these [315:09] things are commercially available and [315:11] consumer grade I'm glad we waited to [315:14] this episode because I think I would [315:15] have had a pretty different take having [315:17] only tried VR headsets and Vision Pro [315:20] and big goggly things over the years I [315:23] think I would have been unconvinced [315:26] frankly because I didn't think it was [315:27] possible to put something in that small [315:28] of a form factor so that is the current [315:32] product experience that you and I have [315:35] recently had now let's look at the [315:37] business strategy and the finan [315:39] of how we have gotten here and why so [315:42] there are two ways to look at reality [315:44] Labs the first way is to answer the [315:46] question what would have to be true [315:48] about the business to be great on its [315:51] own and justify all this investment and [315:54] let's level said on how much investment [315:57] has gone into it yes so based on their [316:00] spend already since they started [316:02] reporting reality Labs as a separate [316:04] segment in 2019 they spent right around [316:07] $60 billion that is in operating losses [316:10] for the segment and does that include [316:12] the 2 billion for Oculus or I mean [316:14] either way it's a rounding error who [316:15] cares I don't think it does with that [316:17] level of investment it already needs to [316:20] be essentially the most successful and [316:22] profitable consumer product in history [316:25] to pay itself back that is the only [316:28] possible outcome here where we even get [316:30] our money back I know this sounds wild [316:33] but like that is the bet no other [316:35] outcomes are acceptable fortunately if [316:38] we want model this out we know the [316:40] financials of a product like that the [316:42] most successful and profitable consumer [316:45] product in history the iPhone so as a [316:48] thought experiment what if meta managed [316:51] to launch such a product say the Orion [316:54] glasses tomorrow and say that such a [316:56] product grew at the exact same rate and [316:58] with the exact same Profit Stream as the [317:00] iPhone well I did that napkin math and [317:03] if [317:04] 2024 this year was Apple's 2007 so you [317:08] just take all the iPhone's cash flows [317:10] and you start the clock right now meta's [317:14] cumulative cash flows from reality Labs [317:17] would be net negative until at least [317:20] 2035 okay okay so we're talking 11 years [317:24] you would get back to break even on your [317:26] investment 11 years from now if starting [317:28] tomorrow they manage to create the most [317:30] profitable and widely adopted product in [317:33] human history which obviously is not [317:35] happening tomorrow that also assumes [317:37] generously that could build a Services [317:40] business attached to it the size of [317:41] Apple's Services business which probably [317:44] generates right around an equal amount [317:46] of profit so basically take all the [317:48] profits from iPhone and double it that's [317:51] actually what you would need it's [317:53] actually fair to attach a similar size [317:54] Services business I think that's very [317:56] fair knowing meta and all of its [317:58] capabilities and everything that we just [317:59] talked about yes that is the BET let's [318:03] just be super clear anything else is a [318:06] complete incineration of cash well okay [318:09] that is the BET looking at through the [318:11] lens that you posited of a purely [318:13] Financial investment yes perspective the [318:16] second way is actually kind of financial [318:18] too I think there's probably a third [318:19] more emotional way the second way though [318:21] is if you're Mark and you constantly [318:23] live under the thumb of platform control [318:26] you'd do almost anything to get out of [318:29] it and that's not irrational I mean [318:31] Apple made a 10 billion dollar dent in [318:33] their revenue just two years ago with [318:35] ATT and I mean they could I don't they [318:39] would but they could at any given moment [318:40] just pull you out of the App Store and [318:42] you'd have little recourse that is a [318:45] existential business risk and it's an [318:47] unlikely one but every day you could [318:49] wake up and all of your access to all [318:53] Apple customers could be over which by [318:56] the way we haven't talked about it but [318:58] Google could do the same thing Google [319:00] could totally do the same thing they [319:01] have the right to distribute or not [319:03] distribute anyone's app in their store [319:04] at any given time look at Apple and epic [319:07] with fortnite totally so there is [319:10] actually an expected value calculation [319:11] you can run which is my entire company's [319:14] market cap times the likelihood that it [319:16] could happen which you know it's [319:18] extremely low likelihood but because the [319:20] market cap is $1.5 trillion the expected [319:24] value is still a very big number so then [319:27] if you're thinking this way what is a [319:29] reasonable percent of your market cap to [319:31] invest every year in a hedge that might [319:34] might get you out from under the thumb [319:37] of big Tech platforms [319:39] and at meta being worth 1.5 trillion if [319:41] they're spending 15 20 billion a year on [319:44] reality Labs losses that's a little over [319:46] 1% right is that worth it if you truly [319:50] believe that this is the most effective [319:53] way to offset your most existential risk [319:56] in the next two decades hell yeah it is [320:00] hell yeah it is now am I certain that [320:02] this is the best way to hedge that risk [320:04] no I don't know and like is there a huge [320:06] amount of execution risk along the way [320:08] well they're running many Hedges on this [320:09] of course I mean this is part of what [320:10] fair is this is what llama is etc etc [320:13] but it's a 1% tax on your I mean it's [320:16] generous to frame it off of market cap [320:18] you probably should frame it off of [320:19] Revenue but still a 1% tax off of your [320:22] entire Enterprise Value every [320:25] year I think there is also potentially a [320:27] third way to frame it and it's funny I [320:29] think you put it in the emotional lens I [320:31] I actually think less emotional but more [320:35] upside oriented of Facebook did have the [320:39] taste of being a platform for those [320:42] couple years there from 2007 to call it [320:45] 2010 [320:46] 2011 and if they own the next [320:50] platform not only will they do [320:54] everything that you're just talking [320:55] about of like Oh They'll have an iPhone [320:57] siiz business because if it turns into [320:59] be glasses in reality Labs they're going [321:01] to be selling the hardware and maybe [321:03] they can make that profitable like the [321:04] iPhone is profitable etc etc then [321:06] there's also your point number two which [321:09] is like this is a defensive play our [321:11] Core Business right now is at risk [321:13] because it's built on somebody else's [321:14] platform and Mark made the comment on [321:16] stage with us that they've run the [321:17] analysis they think they could be twice [321:19] as profitable in their core media [321:21] products if they were on their own [321:23] platform sure great but then there's [321:26] like the big big play is the third part [321:28] of like well what if we actually built a [321:30] platform again where we have millions of [321:32] developers who are developing on us and [321:35] we are participating in their business [321:38] in the way that we used to participate [321:39] when they were developing on us in the [321:41] past that's an even bigger plate and to [321:44] underscore your point David just a few [321:46] weeks before we recorded this episode [321:48] meta held a multi-hour keynote at [321:51] connect in September 2024 and of those [321:54] multiple hours zero minutes were [321:56] dedicated to their core products of [321:59] social media apps their hundred billion [322:02] dollar business of selling [322:03] advertisements on these social media [322:05] products so at least with Developers [322:08] it's not about any of that at all [322:11] literally the entire keynote yeah I did [322:13] it was great it's meta AI it's the Llama [322:17] models underneath it it's the open [322:18] source strategy it's developers building [322:21] for the quest it's announcing new [322:23] products like the next iteration of the [322:26] Rayband metag glasses it's revealing [322:28] Orion zero minutes to their current [322:31] products or business and they're one of [322:33] the biggest businesses in the world yeah [322:36] wild right they are all out this [322:38] platform future so here we are trying to [322:42] make the case of like they did it they [322:44] decided they were an ads business and [322:45] they built one of the two greatest [322:48] advertising systems ever known to man [322:50] and one of the most amazing business [322:51] models ever and at their annual [322:54] developer keynote they're talking about [322:56] how they're going to One Day become a [322:57] platform [323:00] company there we have [323:02] [Laughter] [323:03] it that is the story of meta I should [323:06] probably say for a minute here too [323:08] my experience using Orion you were [323:10] talking about yours a minute ago I [323:12] thought it was unbelievably compelling [323:16] and before trying it I never would have [323:20] believed that any type of AR or VR would [323:25] really have been the platform of the [323:26] future and then after trying Orion I was [323:29] like oh yeah I could wear these on my [323:32] face all day and like I could replace a [323:35] lot of what I do on the phone with just [323:39] these lenses being in front of me all [323:40] the time and it would be way more [323:42] efficient and way more enjoyable I would [323:44] never have guessed that in this year you [323:45] could fit that into glasses glasses and [323:48] a wireless Puck totally yeah listeners [323:51] we've never shared our Impressions on [323:52] this it is wildly compelling and if you [323:54] asked Mark he probably would frame this [323:56] whole thing differently than we have [323:58] it's not about this hedge and he would [324:01] say some things about platform control [324:03] but I think for him is just this General [324:06] belief that we want to make awesome [324:09] products and I believe there's an [324:11] awesome product to be made here and I'm [324:12] going to assemble the best people I can [324:14] to go work on it and I think that's the [324:15] like most interesting duality of this [324:17] company is it's both it's what is the [324:22] best strategic move to make to Marshall [324:24] my resources and a much more touchy [324:26] feely like I want to make products that [324:28] are great and bring people closer [324:30] together because that's the mission of [324:31] the company yep well well put should we [324:34] wrap the story there we should all right [324:37] so we will catch you up on the business [324:39] today just to put some numbers to all of [324:41] this and then we will move into analysis [324:44] so as of the end of reporting last [324:47] quarter there are 3.3 billion daily [324:51] active people across the whole family of [324:55] apps okay astonishing that's up 7% [324:59] year-over-year the family of apps [325:02] Revenue per person is about [325:04] $12 the end ofe stats from last year [325:07] when you just look at the Facebook app [325:09] not the whole family of apps the daily [325:11] active users are 2.1 billion so of those [325:15] 3.3 billion daily active people across [325:18] all the apps 2.1 billion are on Facebook [325:22] and the monthly number of Facebook is [325:26] 3.1 billion monthly active users so 2.1 [325:30] daily 3.1 monthly wow WhatsApp and [325:35] Instagram are in the neighborhood of two [325:36] billion monthly active users WhatsApp [325:39] has a 100 million now in the US which [325:43] this is sort of a big narrative [325:44] violation that WhatsApp will never catch [325:46] on in the US and iMessage is dominant [325:49] and even after that it's you know text [325:51] message this is crazy 100 million people [325:54] a month in the US use WhatsApp now I [325:56] mean it's been a slow burn over time [325:58] growing and growing and growing but to [325:59] me that kind of came out of nowhere the [326:01] other narrative violation here there [326:03] actually is a lot of growth among young [326:05] adults using the Facebook app itself in [326:08] the US I think a lot of people think [326:10] that's a sort of Boomer thing it's funny [326:12] one thing that may be contributing to [326:14] that which we didn't put in the story is [326:15] market place I mean for me personally [326:17] like I have totally boomeranged all the [326:20] way back around on the Facebook Blue app [326:22] and I am now a loyal daou because of [326:25] marketplace it's awesome which is crazy [326:28] because you don't use any other form of [326:31] social media besides like posting on [326:33] Twitter for acquired right correct [326:36] they've announced that meta AI is on [326:38] track to be the most used AI assistant [326:40] by the end of the year it's worth [326:42] disambiguating meta AI from llama llama [326:45] is the name of their family of open- [326:47] source models those models do power meta [326:49] AI but meta AI itself is a branded [326:52] consumer experience that lives both [326:54] Standalone and in a bunch of meta's apps [326:57] yep their revenue in 2023 at the end of [327:00] the year was 135 [327:02] billion operating income of $47 billion [327:06] so that's a 35% oper margin worth [327:09] knowing just like all tech companies [327:12] they have become capex heavy the last [327:14] few years they now uh last year spent [327:17] $28 billion in capex which you should [327:21] mostly read as data centers they operate [327:24] a hyperscaler sized data center [327:27] footprint give or take you know you [327:29] think AWS Azure at Microsoft and Google [327:32] Cloud meta is the fourth one they just [327:36] don't sell it to anyone else it's only [327:39] consumed by internal teams so huge [327:42] amounts of investment in AI hardware and [327:44] just other data center expansion the [327:46] balance sheet is Fortress they have 50 [327:49] billion in cash and 58 billion including [327:52] cash equivalence and marketable [327:53] securities they have 71,000 employees [327:57] and their market cap is 1.5 trillion up [328:00] from 230 billion in just October of 22 [328:05] wow incredible totally incredible so two [328:08] really insane observations about the [328:10] state of the company today all of their [328:13] products seem to increase user [328:15] engagement over time and all of these [328:19] products have different use cases and it [328:22] happens across [328:24] geographies there is something in the [328:26] water at this company why is it that [328:29] WhatsApp and Instagram are both [328:31] increasing in user engagement over time [328:34] it's the growth function I mean it's [328:35] this like purpose-built [328:38] heat seeking missile of attention and [328:40] metrics that the company pays attention [328:41] to where across a whole broad product [328:44] Suite engagement increases it's not like [328:47] they have one thing that happens to do [328:49] really well it's a process I would say [328:51] it's a whole set of things is these you [328:54] know the company calls them centers of [328:56] excellence internally it's the growth [328:58] function it's AI you know it's the ads [329:01] and the revenue Team all of this [329:04] contributes to it yep Fair the second [329:07] insane thing [329:08] at IPO the US and Canada Market had an [329:12] average revenue per user of [329:14] $11 that number is now [329:17] $227 wow yeah and globally that average [329:21] revenue per user when you include all [329:23] the emerging markets and less valuable [329:25] markets for them is $44 so they really [329:28] really monetize now so how much of the [329:31] world does meta really have [329:33] left to kind of put a book end on this I [329:36] almost asked you this question all the [329:37] way way in the beginning when you were [329:39] teeing up the episode but I'm I'm glad [329:40] we'll save it for here so according to [329:42] the UN last year there were 5.4 billion [329:46] people online this includes China right [329:50] so they're basically saying 2third of [329:51] humans are online and meta does not [329:53] include China yes meta has 4 billion [329:57] monthly active people across the family [329:58] of apps China alone is 1.4 billion and [330:02] well this isn't like totally exact I [330:03] think you just apply the same multiple [330:05] to China and say well two-thirds of [330:07] humans are online third of China is [330:08] online so that means that there's 940 [330:11] million people online in China I mean [330:13] that's probably being conservative I [330:15] think you could probably argue a much [330:17] higher percentage of China is online [330:19] yeah but let's be conservative that [330:21] leaves about 450 million people or 6% of [330:25] the human population who have access to [330:28] the internet but are not yet meta [330:31] monthly [330:32] actives that doesn't mean they're not [330:34] meta users that just means they weren't [330:36] monthly active users as of the end of [330:39] the last reporting period So Meta [330:42] addressable users who aren't yet users [330:44] is less than 6% of humans just wild just [330:48] wild and then still I think the craziest [330:50] thing about the company is that if you [330:53] look at the daily active user figure [330:56] it's not that much lower right and [330:58] across all products and so once you [331:01] frame it this way and you're like huh [331:02] there's only 6% of the population left [331:04] either through reactivation or signing [331:06] up that they could get you sort of [331:08] understand why they put so much effort [331:10] behind Emerging Markets behind doing [331:12] things like zero rating doing custom [331:14] deals with Telos rolling out fiber [331:17] bringing countries online for the first [331:19] time even when they have no near-term [331:21] monetization potential internet.org was [331:23] the name of their ini year for a long [331:24] time around this to basically say look [331:26] we are saturating humans we got to [331:29] figure out how to get more humans on the [331:30] internet yep move on to analysis yep [331:35] okay so in the analysis the first thing [331:37] we're going to do is power and then [331:38] Playbook so we are going to do a seven [331:41] Powers analysis of what enables meta to [331:45] achieve persistent differential returns [331:48] or you know to put it another way to be [331:50] more profitable than their closest [331:51] competitor and do so [331:53] sustainably this is interesting who is [331:56] their closest competitor I think that's [331:58] probably worth defining first and [331:59] foremost ultimately they're in the [332:02] business of selling advertising so I [332:05] think their closest competitor is Google [332:07] yeah [332:09] they are certainly the most similar [332:11] looking businesses in their current [332:13] forms but like who do they compete [332:15] against for the same profits or the same [332:17] potential profits right I'm not sure [332:20] that it's actually Google I mean it [332:22] almost certainly is YouTube within [332:24] Google but I actually think it is other [332:28] social platforms yeah you're right I [332:31] suppose it's other places people spend [332:33] time yeah it's where people spend their [332:34] time right online and offline right and [332:37] and I think really what has to happen [332:39] here is an analysis of each set of [332:41] stakeholders individually like you [332:42] almost kind of want to do a seven Powers [332:44] analysis on the user side also of why [332:47] would someone pick a meta product over a [332:49] snap product and even though they're not [332:51] voting with dollars it's almost like [332:53] their attention is axy for dollars [332:54] because you just assume that those [332:55] companies should do a comparable job [332:57] monetizing the attention yeah but what [333:00] makes sense I think is to just walk [333:02] through each of them so counter [333:03] positioning they probably don't have [333:05] much counter positioning in the current [333:08] state of meta and hold on AI for the [333:09] moment in their startup days they did a [333:13] lot of counterposing against other [333:15] global social networks by being a [333:18] college only authenticated social [333:20] network they were accepting lower growth [333:25] they were accepting a capped ceiling of [333:27] number of users and they were doing that [333:30] because they wanted to make the trade [333:31] that they felt you know a closed [333:33] community is more important yep with [333:35] higher engagement scale economy [333:38] scale economies 100% absolutely this [333:42] company's in the business of scale [333:43] economies yes and I think that's scale [333:45] economies on the infrastructure side on [333:47] the GPU investment side I think that [333:50] scale economies on The Advertiser side [333:52] and on the ad experience side like just [333:56] by having so much more scale they are [333:58] the default standard way to spend money [334:00] for advertisers think about on the [334:02] tooling side think about the experience [334:03] of being an engineer at meta and the [334:06] Thousand engine engineer years of work [334:10] that comes out of that company every [334:11] year on making the developer experience [334:13] of working there better I mean it's [334:15] crazy with the revenue scale that they [334:18] have how much they can amortize these [334:20] fix costs totally everything at this [334:23] company is scale [334:25] economies yes okay scale economies for [334:29] sure switching costs as a user you're [334:32] pretty locked in once you have followers [334:36] oh I hadn't thought about that yeah that [334:38] really speaks to again you know another [334:39] shift we didn't really talk about in the [334:41] evolution of what we call social media [334:45] is the creator aspect of all of this the [334:49] whole idea of like a Creator an [334:50] influencer didn't exist in the early [334:52] days of social media and Facebook but [334:55] now you and me our platform on the [334:57] podcasting platform like we have [334:59] incredible switching cost somebody who [335:01] has a ton of Instagram followers has [335:02] incredible switching costs yep now [335:05] fortunately in the land of creators you [335:07] don't actually [335:08] need to switch it's just an and Y you [335:11] know it's not like you're ripping out [335:12] one vendor and putting another vendor in [335:15] I do have a sunk cost in building a [335:17] following on a given platform but that [335:18] doesn't actually prevent me from also [335:20] launching on another platform if you [335:22] have the time to kind of do that yep I [335:25] think there might actually be process [335:27] power normally there's not but I keep [335:29] kind of going back to this like there's [335:31] something in the water their products [335:32] grow in engagement over time their [335:35] growth team does 10,000 Little Things to [335:38] open up every step of the funnel as wide [335:40] as it can be and make the most [335:42] frictionless fluid experience for users [335:45] interesting yeah I was actually having [335:46] this discussion with Hamilton recently [335:48] about the difference between process [335:50] power and cornered resource and a good [335:53] test for this is when you bring somebody [335:55] brand new into the organization and plug [335:58] them in does the power transfer to them [336:01] as well if no then it's a cornered [336:04] resource if yes then it's process power [336:07] and in this case I think that's [336:09] potentially true yeah you can bring in [336:12] new Engineers you can bring in new [336:13] product people and transfer the [336:15] advantage to them you know for a while [336:18] they had legitimate process power in how [336:20] they shipped I mean stat is started on [336:22] the premise of this it blew my mind when [336:26] someone from Facebook came and gave a [336:27] talk at Microsoft in [336:30] 2014 on how they ship product and how it [336:34] rolls out and not only the feature [336:36] flagging elements of it but they can [336:37] sort of like watch Performance Auto roll [336:40] things back if they're causing negative [336:42] performance metrics how like there's [336:45] just deployed code all the time with a [336:47] whole bunch of experiments turned off [336:49] how there's these experiments that are [336:51] running in different ways in different [336:53] markets and they can statistically [336:55] significantly [336:56] disentangle which results are from which [337:00] experiment when they have multiple [337:01] experiments that are concurrently [337:02] running with the same user base I mean [337:04] that was like real voodoo that only face [337:07] did for feels like a decade Y and I [337:10] suppose Google probably does similar [337:13] type things in different ways but nobody [337:16] else really does I think open AI kind of [337:20] does I think open AI has so much former [337:22] Facebook DNA and kind of thinks the same [337:25] way that I think they do a lot of this [337:26] sort of thing too yep okay I can buy [337:29] that there's some process power here [337:31] yeah branding branding is such an [337:34] interesting one I mean can they have [337:36] negative Brand Power yeah right [337:38] anti-power branding weakness there's no [337:40] love for the brand meta there's no love [337:43] for the brand Facebook there is love for [337:45] the brand Instagram yeah but diminishing [337:48] as it gets larger certainly not like it [337:50] once was yeah there's like reliability [337:54] with it all like I feel pretty I mean [337:56] there's branding that comes with being [337:58] any big company that you're sort of like [338:00] large and trusted and institutional yep [338:03] but no I think this is a clear example [338:07] of how strong the other powers in this [338:10] company are that like despite the brand [338:13] taking a 20-year hit in Marx's parin [338:16] they're still today a you know one and a [338:18] half trillion dollar company yeah and [338:21] then cornered resource unless you're [338:23] going to call Mark one which always [338:25] feels a little bit too cute to name the [338:28] founder uh I actually think there are [338:31] some cornered resources here within meta [338:34] and ironically I think one of the [338:36] biggest ones is the integrity and safety [338:41] team and all of the Privacy controls and [338:44] work and security that they have done [338:47] that they've had to harden over the [338:49] years yeah yeah even though this is also [338:52] the source of so many criticisms about [338:54] the company it's now pretty robust if [338:56] you think about anybody that would start [338:58] up to try and compete with them somebody [339:01] I was talking to put this to me like [339:03] imagine a startup trying to build a hate [339:06] speech classifier farsy meta's got that [339:10] and trying to have the relationships [339:12] with the public policy people in 200 [339:14] countries to understand what is [339:17] acceptable speech in each of those [339:19] countries yep so while probably you know [339:24] YouTube to a certain extent Tik Tok [339:27] perhaps also have similar strengths here [339:32] I doubt there is any other company in [339:35] the world capable of actually operating [339:40] privacy integrity and security and [339:42] within that is also anti-spam stuff too [339:44] which is super super super huge Y at the [339:47] level that Facebook is it's interesting [339:50] okay so in looking at this why does meta [339:52] do 47 billion dollars a year in [339:54] operating income and why do we all [339:56] believe that's going to continue for a [339:57] while like what actually is the [339:58] defensibility here is it the network [340:01] economies because they're so tological [340:02] that I mean I would argue it's not that [340:05] because what we've seen is Tik Tok [340:07] showed us there's a way without initial [340:10] strong Network effects to go capture [340:14] people's attention and thus eventually [340:15] the ad dollars yep and they competed [340:19] very successfully with meta in doing so [340:21] yeah if this were five years ago I would [340:23] have been like why even talk about the [340:24] rest Network economies once you have the [340:26] network then you know it doesn't the [340:28] rest doesn't matter that's just not true [340:29] anymore y well I think the network [340:32] economies are still true relative to [340:35] anybody who would compete in a social [340:36] product yes which still exists totally [340:39] social is just now divorced from media [340:41] right and most people do use both and [340:44] sometimes they even use different apps [340:46] for it but it's two different modes of [340:48] let me it may be the same session on the [340:50] couch but you do need both use cases of [340:54] show me who the people that I [340:55] intentionally follow what they're doing [340:58] and then show me entertaining things [341:01] yeah totally so part of their business [341:03] is defensible from the network economies [341:06] but the other part honestly it it kind [341:08] of feels like habit well habit but I [341:10] think scale economies are really big for [341:12] that too especially in AI I think the [341:15] only way that this new paradigm exists [341:18] is because of AI and because of gpus [341:20] specifically and I'm not even talking [341:21] about generative AI I'm talking about [341:23] like feed recommender systems and [341:26] personalization systems and so like [341:28] there is a very very large minimum [341:31] barrier to entry there that keeps going [341:33] higher as bite dance and meta keep [341:36] investing in it yep that's a great Point [341:38] okay Playbook Playbook let's do it okay [341:43] so let's talk about what meta is doing [341:47] with AI and that will lead us into our [341:49] first Playbook theme because we've [341:50] talked about so far the beginning of [341:52] fair them starting all this AI research [341:55] the early 2014 on use of feed [342:00] recommenders and the AI for the ad [342:03] matching system but there's a lot going [342:06] on with Facebook and AI right now that [342:08] we really haven't talked about so there [342:11] are two words for you to know one is [342:14] llama and this is the family of [342:17] foundational models that meta has [342:19] developed and these are competitive with [342:21] open Ai and anthropics Claude and Google [342:25] Gemini Etc yeah and then there's meta Ai [342:29] and meta AI is a consumer brand that is [342:33] the way that you interact with meta's [342:36] self-hosted [342:37] version of llama Or maybe give it a [342:39] little bit more Creed than that it is an [342:41] application that meta has that uses [342:44] llama in the background but provides [342:47] meta specific AI experiences some of [342:50] them bundled into apps like in WhatsApp [342:52] chat or in Instagram the Blue app or [342:55] whatever yep yeah but also there's a [342:57] meta AI website that you can go to and [342:59] interact with it directly but llama is [343:01] the models themselves meta AI is the [343:03] consumer product and llama they have [343:06] spent billions and billions of dollars [343:09] huge amounts of R&D huge GPU clusters to [343:11] train Big Data Centers interestingly it [343:15] is all open source so Mark makes a big [343:19] deal about this yes it's interesting to [343:21] sort of think about why it's not open [343:24] source in the same way really that oh [343:27] Linux is open source and this is the [343:28] free standard that everybody uses and [343:30] there's just sort of a foundation behind [343:32] it all meta is putting huge amounts of [343:36] cap and Opex like huge amount of dollars [343:39] into willing llama into existence it's [343:43] very different than these sort of cheap [343:45] Grassroots open source projects of the [343:47] past yeah it's open source like Android [343:49] is open source yes and so why are they [343:53] doing this well if you ask Mark why they [343:57] have an AI model at all and I'm quoting [343:59] from a great blog post that he put out [344:01] about this we must ensure that we always [344:03] have access to the best technology and [344:06] that we're not locking into a [344:07] competitor's closed ecosystem where they [344:10] restrict what we build okay I understand [344:14] that you believe that the most important [344:16] thing is to control the key technologies [344:18] that make your products possible okay [344:21] that last bit is important basically [344:23] what Mark is saying here is we're going [344:26] to spend a lot of money training these [344:28] foundational models but unlike all of [344:30] the competitors in the AI space we [344:32] actually don't have a business model [344:34] around making money on this so we're [344:36] going to spend all the money we're going [344:37] to give it away free why does that make [344:39] sense it's open compute project all over [344:41] again yes So for anybody who's not [344:43] familiar Facebook made this move in the [344:46] early 2010s where they realized they [344:48] were spending tons and tons of money on [344:50] their data center infrastructure and the [344:52] vendors who they were paying these [344:54] integration partners and the server [344:55] companies and the networking companies [344:57] were making fat margins and they were [344:59] thinking this is dumb we're a really big [345:01] customer why is everybody else making so [345:03] much money on us and they looked around [345:05] and they saw all these other Big Data [345:07] Center companies and they're like gez [345:08] all those people are paying big margins [345:10] too what if we just publish the specs [345:14] for the billions of dollars of work that [345:16] we have done to make our data centers [345:20] and then we start this thing called the [345:23] open compute project and we just get a [345:25] bunch of other people to adopt it too [345:27] well suddenly then the open compute [345:29] project is this standard by which all [345:32] the hardware manufacturers and [345:34] integrators actually have to snap to cuz [345:37] all the customers are saying this is [345:38] what we want and it's a pretty genius [345:41] way to drive margins down for these [345:43] suppliers and also a way to guard [345:46] against somebody in your supply chain [345:48] kind of building platform power yes so [345:52] what they basically learn from this is [345:54] oh we open sourced this thing a lot of [345:56] our costs went down because the whole [345:58] ecosystem started using the thing that [346:00] we open sourced so even though we're not [346:02] making money you know they're not a [346:04] cloud company they're not selling access [346:06] to their data cers anyone right they're [346:08] not a hyperscaler it pays back in the [346:10] form of saving them money it can do the [346:13] same thing in AI they publish a really [346:16] expensive open source foundational model [346:18] that is in the conversation to be as [346:21] good as these other closed Source ones [346:23] well now there's a lot of developers out [346:26] there who are just going to build on the [346:27] open source free one it kind of becomes [346:30] something that the community can build [346:31] on and improve and make better as Mark [346:33] said in our conversation with us on [346:35] stage there's a lot more smart people [346:36] outside your company that inside it but [346:38] effectively what it does is it puts [346:40] pricing pressure on the AI model [346:42] companies to the extent that Mark views [346:44] it as a super important key ingredient [346:47] to the product experiences he wants to [346:49] build in the future it's super bad if [346:52] there's a few Clos Source providers who [346:53] can provide that experience and they a [346:56] lock him into controlling so here's what [346:58] you can build here's what you can't [346:59] build but B take his margin basically [347:02] say it's really expensive every time you [347:04] want to make a call to one of our [347:05] proprietary services [347:07] when you think about it it's actually a [347:09] form of operating leverage where he's [347:11] basically saying there's a big fix cost [347:13] I willing to bear in order to bootstrap [347:16] this ecosystem and commoditize all of [347:19] these complement commoditize all of [347:22] these other Clos Source AI models and in [347:24] exchange what I'm going to get for that [347:25] is just pricing pressure on all of them [347:28] so that in the future my variable costs [347:30] are lower I just get to keep more of the [347:33] dollars that we bring in rather than [347:35] having to pay them out to propri model [347:37] providers in the future it's a pretty [347:40] novel business strategy yeah totally so [347:43] I said commoditize your compliments this [347:45] notion was dreamed up and named by Joel [347:48] spolski in 2002 there's a great blog [347:52] post about it and he makes this analogy [347:55] think about cars and gasoline these are [347:57] complement when sales increase in one [348:00] sales increases in the other you have a [348:02] car you need gasoline well AI models end [348:04] up being a complement to meta products [348:06] and order for them to build meta AI or [348:08] even the feed recommendation stuff we [348:09] talked about they need best-in-class AI [348:11] models and just because cars increase [348:14] the sales of gasoline that doesn't tell [348:16] you about how profitable an automaker [348:19] gets to be versus a gasoline maker so [348:21] imagine the automaker decided that they [348:24] want to get into the gas business so [348:26] that their customers could have access [348:27] to lowcost gasoline or even further [348:30] let's say the automaker decided to make [348:31] lowcost gasoline just to drive all the [348:33] other gasoline prices down well if gas [348:35] is cheaper and you're the automaker you [348:38] can actually charge more for cars since [348:40] consumer willingness to pay is around [348:41] the total cost of ownership not about a [348:44] car or gasoline specifically right as a [348:47] consumer you're always buying a solution [348:49] you're buying Transportation or you're [348:51] buying a technology product doing a job [348:54] for you and if AI is a required input to [348:57] that solution then like yeah you're [349:00] thinking about that as a total purchase [349:02] decision yes or an Advertiser is not [349:05] going to pay you more more or less [349:07] depending on how much you have to pay [349:09] the AI model provider they're going to [349:11] pay you an amount and if you want to [349:12] maximize the amount of that you get to [349:14] keep it behooves you for you to not have [349:16] to pay as much money to AI model [349:18] providers yes better [349:22] put okay so how does this apply to [349:24] Playbook well meta seems to like taking [349:28] all the risk in situations like this [349:30] putting lots of dollars in so they can [349:32] take more of the reward you could [349:34] imagine a more moderate company saying [349:36] oh well there's going to be lots of AI [349:38] vendors out there and uh we could just [349:40] let them take care of that as their core [349:42] competency and we'll just buy off the [349:44] shelf from them but that's not how meta [349:46] Works especially with everything they've [349:49] been through with apple they have gotten [349:51] True Religion and I think always kind of [349:53] wanted to be in this position and they [349:54] just have the capital to do it now they [349:56] will control the key technologies that [349:58] matter to them both for getting to own [350:02] and dictate product road maps and [350:03] products decisions but also for the [350:05] financial upside of making sure that [350:07] they control their own destiny and no [350:09] one in the ecosystem has extreme [350:10] leverage over them right only a company [350:14] of the scale of meta can kind of run [350:16] this Playbook playbook in quotes because [350:19] no startup is going to be like oh okay [350:21] what is my strategy to make sure that [350:23] open Ai and anthropic don't get leverage [350:25] over me good luck with that yep totally [350:29] so my first big Playbook theme is meta [350:32] discovers commoditize your compliment [350:34] and is now looking for ways to use it [350:36] everywhere [350:37] I love it well put okay great I think my [350:40] first one is almost a different way of [350:43] putting that which is really like Mark [350:45] and specifically his superpower of [350:49] placing multiple bets on multiple [350:51] chessboards never wanting to be backed [350:53] in a corner yeah and I think he has done [350:56] incredibly well and I think the story of [350:59] meta shows all throughout even in the [351:02] very very beginning as a tiny project [351:04] not even a startup where degrees of [351:07] freedom were limited he was always [351:09] making choices and playing the game such [351:12] that there were multiple options of how [351:14] things could go right totally Mark is a [351:18] master at maximizing his degrees of [351:20] freedom and setting up the board such [351:23] that in an uncertain future there are [351:26] multiple paths to Victory no matter how [351:28] the world unfolds yep which is [351:31] unbelievably gatesan I feel like we said [351:33] the exact same line about Bill gates in [351:37] our Microsoft episodes the real comp for [351:39] this company is Microsoft I think that's [351:42] right that was actually my next one they [351:44] do iterative product development they [351:46] put the first version out just to kind [351:47] of get feedback and see how they need to [351:49] rev it and get better for the future the [351:51] early days are characterized by hiring [351:53] all the smartest people and prioritizing [351:55] IQ over everything else commoditizing [351:57] your compliment obsession with building [351:59] a platform that other developers build [352:01] on top of this whole thing about [352:03] multiple bets in an uncertain future I [352:05] mean it's funny that just look at they [352:07] were building messenger internally they [352:09] bought WhatsApp I bet they're glad that [352:11] they had that dual prong strategy they [352:13] were developing an app called photos at [352:15] the same time that they bought Instagram [352:17] that's what they do it's gatesan and the [352:19] most interesting thing that I think is [352:22] like hit me in the face like a ton of [352:23] bricks Zuck after all the 2016 election [352:27] Fallout and the shift in public [352:30] perception and having to do all this [352:32] testifying is like watching an alternate [352:34] future for Microsoft where Bill States [352:36] had decided to stay at the helm instead [352:38] of leaving after the doj case that is [352:42] the biggest difference is they both went [352:43] through this hellish period Y and Mark [352:46] came out of it and said I'm more bought [352:48] in than ever Y and bill came out of it [352:51] and said Steve you're the CEO now yep [352:54] that's exactly it yep are you gonna talk [352:57] about Mark's amazing line from the [352:59] Harvard cs50 lecture in like 2005 I'm [353:02] not but lay it on me oh it's so great [353:05] marks what years old at this point maybe [353:08] he spent the summer out in Palo Alto he [353:11] gets invited back to do a guest lecture [353:13] as like a practitioner in the field at [353:15] Harvard cs50 class and he's talking [353:19] about product strategy and he talks [353:21] about how he really admires Microsoft [353:23] product strategy and he thinks it makes [353:25] a lot of sense of the first version is [353:27] getting something out there and it's [353:28] usually not very good but by the third [353:29] or fourth version it's pretty darn good [353:31] and that makes a lot of sense to me as a [353:33] good product strategy it is amazing he [353:35] lays it out right right there we don't [353:37] have to like speculate at it it is [353:38] explicitly stated in the same vein it is [353:42] painfully obvious when you look at this [353:43] company that companies are just Founders [353:46] extended the culture of this company is [353:50] just Mark and it's a huge lever for Mark [353:54] to act this has been true over and over [353:56] again Microsoft Nvidia Nike it's just [354:00] how great companies are in the world [354:02] it's very hard for me to point to a [354:04] truly great company where the DNA isn't [354:09] like a lever on the founders personality [354:12] I think that is right and I think it is [354:13] doubly so with meta because of Shawn [354:17] Parker and because it was set up from [354:19] the very very beginning that nobody [354:21] could ever get rid of Mark that it was [354:23] always going to be his decision and [354:26] there's got to be a lot of companies out [354:27] there like we're looking at the success [354:29] case yeah this works if the founder is [354:32] really really really smart and right [354:34] almost all of the time right as talk to [354:36] people here are the traits that we heard [354:38] over and over again Mark is a genius a [354:41] really good listener a fast learner he [354:43] goes from knowing zero to Mastery in [354:45] months or years he has low ego about [354:47] being right that's not to say that he [354:49] has a low ego but he has a low ego about [354:51] being right he is obsessed with finding [354:53] truth and open to being wrong he's [354:55] intensely competitive he's Relentless [354:57] he's actually a very good product [354:59] designer and understanding the computer [355:00] architecture that will be required to [355:02] accomplish such a product experience at [355:04] scale I mean all these things like if [355:07] you have these [355:09] characteristics and you're then [355:11] empowered to singularly control the [355:13] company at massive scale yeah it's going [355:16] to go well but the first thing is [355:18] actually much harder than the second [355:19] thing yes yes it's not a causal [355:23] relationship here that having founder [355:25] control will lead to the right decisions [355:28] right this is my regular reminder that [355:30] in studying these episodes with extreme [355:33] survivorship bias we are looking at the [355:36] most extreme outliers who in every [355:39] Dimension you could multiply by they're [355:42] at the edge of the distribution Mark is [355:44] I don't know six or seven standard [355:46] deviations from the mean human across [355:49] the important traits that mattered to [355:51] making Facebook oh and by the way with [355:54] the right timing and the right luck and [355:55] the right circumstance and the right [355:56] knowhow my biggest lesson from doing [355:59] acquired is these things are unre [356:01] creatable [356:03] yeah agree all right once next one [356:07] durable executive team it's kind of [356:09] shocking how many of the people that are [356:12] Mark direct reports and their directs [356:14] have just been there for a really long [356:15] time this team knows how to work [356:17] together oh yes I was actually talking [356:21] about this with one of the execs at meta [356:23] the other day and he brought this up and [356:26] was saying that like this really feels [356:28] like something very very unique to meta [356:30] and I thought about it for a minute and [356:31] I was like [356:33] actually no I think this is something [356:36] that is unique to all great companies if [356:38] you look across the landscape out there [356:41] Apple certainly like this if you look at [356:43] Nvidia if you look at Apple if you look [356:45] at you know Microsoft in recent years if [356:48] you look at the original Microsoft [356:49] Costco you look at Costco like this is [356:53] almost always true really is true in [356:55] every great company that I can think of [356:57] the core senior most executive team are [357:00] all people who came up in the company or [357:02] have been there for a very very long [357:04] time it's not a whole bunch of mercen [357:06] there at the [357:07] top yeah that's [357:09] interesting okay that brings me to this [357:12] next one which is something we talked a [357:13] little bit about with Mark on stage and [357:15] I'm curious to hear your answer to this [357:19] now that we've had all this time and [357:20] space to think about it here are all the [357:22] battles where meta has either one by [357:27] buying the company one by beating the [357:28] company one by copying or got into some [357:31] kind of sustainable [357:32] stalemate Myspace and Friendster Google+ [357:36] Twitter Instagram WhatsApp we didn't [357:40] talk about this one but um Mir cap and [357:42] Periscope oh yeah they launched Facebook [357:45] live when that was supposed to be the [357:47] next big thing yep Snapchat and Tik Tok [357:50] that's like seven [357:52] eight why do they keep winning I have [357:55] some thoughts it's funny as we were [357:58] preparing to interview Mark at Chase [358:00] Center last month this was the question [358:03] that we kept asking people in some ways [358:07] I think it's a little bit of a rosack [358:08] test everybody's answer was different [358:11] and it was always like what they think [358:13] is the most important thing yep I think [358:17] my answer for this is going to be a [358:20] version of Mark as the answer yeah and [358:26] it kind of goes back to Adam D'Angelo's [358:30] Friendster testimonial about him being [358:33] way too lucky and then realizing it's [358:35] not lucky at all and I think it is that [358:40] you know as Adam was telling to me as [358:42] he's reflected on this it's a very [358:44] specific way of carrying yourself [358:46] through the world where you don't hold [358:49] too tightly to the path you're on [358:53] obviously Mark is incredibly driven to [358:56] the goal of you connecting all the [358:59] people in the world and doing that [359:01] through meta but he's very very open and [359:04] flexible to like the exact path that [359:06] it's going to take to get there yes and [359:08] that combination I think has led over [359:11] and over and over again to these threats [359:13] emerging and Mark and the whole team [359:15] there at meta saying like Okay what is [359:20] the way that we can neutralize this [359:23] threat or defeat it and we are open to [359:26] anything on that front and I think they [359:29] are open to being whatever they need to [359:31] be to make that [359:34] happen I think meta might look look [359:36] extremely different 20 years from now [359:39] than it is today almost like [359:40] unrecognizably because this company [359:42] moves like water in response to whatever [359:46] the new shape of the world is I mean [359:48] hell after trying Orion I think there's [359:50] a very plausible reality five years from [359:54] now where this company looks very very [359:56] different than it does today yeah I mean [359:58] it really is this idea meta is a [360:00] technology company through and through [360:02] and then they leverage that technology [360:03] to be whatever the hell they need to be [360:05] to adapt to the new world which is so [360:07] different from all of the other big [360:09] technology companies out there they are [360:11] all wedded to a particular vision of [360:13] what they are totally because these are [360:14] all different strengths like they aren't [360:16] afraid of copying okay that is a [360:17] strength but that's also a completely [360:19] different thing than I'm going to place [360:22] bets on what I think is going to be the [360:24] big technology wave of the future and [360:26] spend tens of billions of dollars on [360:27] that that's a different way of winning [360:30] or in some cases I'm going to try to [360:32] leverage my existing Network effect to [360:36] make sure I adopt someone else's social [360:38] mechanic okay that worked in a handful [360:41] of these scenarios but that's not at all [360:44] what works in others they don't just [360:47] have One Singular Playbook that they run [360:49] over and over again right right this [360:51] company moves like water and product is [360:54] an act of Discovery there I'm convinced [360:57] it's not pure invention you asked Mark [361:00] that on stage and there's some stuff [361:03] they have to invent like Orion you have [361:05] to invent but soft sofware where you can [361:06] ship and quickly respond to user [361:08] feedback and iterate it's like they're [361:10] chisling away at the marble to find [361:12] David I don't think they have David in [361:14] mind when they're starting most of the [361:16] time there are a few moments where they [361:18] did Newsfeed is completely one of them [361:21] inventing the social feed is a [361:25] completely distinct and brand new thing [361:27] that meta created it's actually one of [361:29] the few yep when I was a little bit more [361:32] bearish on the future of reality Labs I [361:34] was trying to come up with has meta ever [361:37] successfully created something new that [361:40] has become a profit Center for them that [361:42] is like not adopted from someone else [361:45] well I think the tagging of photos was [361:47] pretty novel yes and Newsfeed yep but [361:52] it's like kind of funny that we're [361:53] naming those things as platform as [361:55] short-lived as it was no that was a we [361:59] wish we had an operating system but we [362:01] don't so let's see if we can convince [362:02] people that this is a sufficient [362:04] platform well I think building a real [362:07] honest to God platform with hundreds of [362:10] thousands if not millions of developers [362:12] making money on it on the open web okay [362:15] fair push back it was totally novel like [362:18] nobody had done it without an operating [362:19] system before and like maybe there was [362:22] going to be a social platform in this [362:24] new era yeah maybe that was an operating [362:26] system of sorts or on par with operating [362:28] systems to be able to create a platform [362:30] on top of but I take your point it's [362:32] certainly not every turn of the game [362:34] where they're inventing they're doing a [362:37] lot of Discovery here right and maybe [362:40] all companies do and you know there's [362:42] not something distinct to meta here and [362:44] and everybody learns from each other and [362:45] that's fine too but that was something [362:47] that I was racking my head on thinking [362:49] through in fact there's even a testimony [362:51] this is kind of funny in the Mark [362:53] Zuckerberg versus the winlos case from [362:56] way way way back when Mark even makes [362:58] some comment about well actually the [363:00] idea for Facebook wasn't even new [363:01] because Myspace and friend existed so [363:03] it's like this interesting positioning [363:05] of the whole thing itself is actually a [363:07] borrowed idea which of course served him [363:09] well in that particular case but the [363:12] point sort of stands is like so much of [363:14] this is borrowed and we just live in a [363:16] world especially in Social where you do [363:19] kind of have to just maybe this is Media [363:21] as a whole observe what the new format [363:23] is and adopt it as quickly as possible [363:25] yep so speaking of trying things I just [363:29] wanted to take a moment to honor all the [363:32] failures I kept a running list as I for [363:35] the failures as I was doing my research [363:38] on products that either died or just [363:40] didn't live up to the hype Facebook live [363:43] Facebook watch the uh drone the solar [363:48] power drone that they developed to [363:49] provide internet access oh yeah that's [363:52] right that was like a whole big thing [363:54] for a while that yeah that's right Mark [363:56] was actually two for three on he was [363:59] forecasting like in 2015 what are we [364:02] going to be focused on 10 years from now [364:04] that was one of them or the other two [364:05] were augmented reality and AI it's kind [364:09] of an impressive call portal portal [364:12] TV workplace which as far as I can tell [364:15] Facebook is the only company in the [364:17] world that actually uses workplace no [364:18] other enterpris has ever adopted it I'm [364:20] sure that's not like exactly true but [364:22] that feels approximately true Beacon [364:25] Facebook deals which was their Groupon [364:27] clone oh that's right think about how [364:29] Nimble Facebook is that they're like oh [364:31] uh oh this social deals thing seems to [364:33] be gaining traction maybe that's a core [364:36] part of the platform that needs to be a [364:37] piece of this in the future I mean they [364:39] just had zillions of these Facebook [364:41] gifts do you remember gifting and you [364:42] could use Facebook credits to to pay for [364:44] them Facebook credits I mean on top of [364:47] Facebook credits uh Libra oh boy yeah [364:51] wow we didn't even talk about that whole [364:53] chapter on crypto yeah they invented a [364:56] whole new you know cryptocurrency and a [364:59] big Consortium around it and a huge set [365:01] of investors and other Fortune 500s [365:04] originally Facebook mess was going to be [365:06] a Gmail killer you could email people [365:09] into Facebook messages and use that as [365:11] an email Suite Facebook places when I [365:14] mean four square was almost on par for a [365:18] period of time with Instagram and [365:20] Twitter as plausibly the next social [365:23] mechanic checking in places and Facebook [365:26] places was a real effort hey now it just [365:28] turns into tagging location on posts but [365:30] like a first class Post in Newsfeed for [365:33] a while was a check-in on Facebook [365:35] places yep and then there's all the [365:38] independent apps lasso poke slingshot [365:40] photos hello Facebook gaming life stage [365:43] moments notify Facebook watch moves all [365:47] the things they acquired TBH Beluga they [365:49] launched [365:50] igtv this company tries everything yep [365:54] they move like water to discover what [365:57] they need to be through an everchanging [366:00] environment so [366:02] true all right that's it on the failures [366:05] what else you got well almost sort of [366:07] the opposite of the failures it's quite [366:10] ironic that Facebook became the [366:14] prototypical startup given that the goal [366:17] was never to be a startup yes the goal [366:20] was first to be a project and then to [366:23] build like a very profitable business [366:25] the largest Empire in the world yes [366:28] there was no in between of like we are [366:31] building a startup here it was like [366:34] college project let's connect all the [366:36] people in the world and let's get big [366:38] really really really fast and skip this [366:40] whole small company thing right that's [366:43] such a good point and all this whole [366:46] generation multiple Generations now of [366:47] startup Founders that have really like [366:49] embraced and romanticized this whole [366:51] startup thing on the one day like it's [366:53] great Silicon Valley is so much bigger [366:55] and there's so much more investment and [366:56] there's so many more VCS and all that [366:58] but it's also just all kind of ironic [367:00] that it's glorifying this startup phase [367:02] when that is not the point deeply okay [367:06] so then we've touched on this one a [367:07] bunch but I kind of want to put a pin in [367:08] it why has Facebook always been in a [367:11] precarious position why do they need to [367:13] keep fighting these existential battles [367:15] why are they so obsessed with building a [367:17] platform I think the answer is they want [367:20] to be as durable as a company that makes [367:23] Hardware with an operating system that [367:25] all the users use and all the developers [367:27] have to Target because all the users are [367:29] there I think that's like what they [367:30] really want but because they've never [367:33] had quite that much defensibility you [367:35] know like oh a network effect is good [367:37] but it's not as good as that incredible [367:39] platform durability they're always [367:41] trying to expand and be more every time [367:46] they bump up against someone else it [367:48] kind of creates a problem for them and [367:50] so I'm using a bunch of weird metaphors [367:52] here but it just seems like the place [367:54] that they occupy in the technology stack [367:57] is just not quite privileged enough to [368:01] do the things that they want to do and [368:03] so they're always at the whim of someone [368:05] else knock them around yep I think [368:07] that's right and it will be very [368:09] interesting to revisit this again in 5 [368:11] years or 10 years yep part two in 10 [368:14] years yeah right and then as we as we [368:19] drift to a close here the engineering [368:22] culture and being a technology company [368:24] at their core has been essential early [368:28] on they really did managed to hire only [368:30] A+ and then stay A+ after that forever [368:34] and ever and ever it was just this badge [368:35] of honor if you were an engineer at [368:37] Facebook for product design too I mean [368:39] they just had such a great talent [368:43] density the set of things that they did [368:47] on the technical side were over and over [368:50] again a way to have their cake and eat [368:52] it too if you can move faster you can [368:55] learn more through your multiple [368:57] iterations and so speed of development [369:01] comes from having great tools one very [369:04] great shiny example of this [369:05] is something called Hip Hop for [369:08] PHP I know this is very esoteric but in [369:11] the late 2000s they had this crazy idea [369:14] that what we should do instead of [369:16] switching to C or C++ or Java we want [369:21] our Engineers to keep writing PHP so [369:23] they wrote a compiler to C so they [369:26] didn't have to take the performance hit [369:28] from running PHP which was an [369:31] interpreted language but it also solved [369:33] this like scalability problem because [369:34] then they didn't need to go hire all [369:36] these systems level programmers they [369:37] could hire web developers who wanted to [369:39] move with that pace and flexibility [369:41] while also having the infrastructure to [369:43] run these massive systems and scale [369:46] really efficiently and then that solved [369:48] the scaling and performance problem but [369:50] again in 2014 they realized oh crap [369:53] we're big enough and we know have enough [369:55] sensitive data that we really should [369:56] switch to a statically typed language [369:59] like Java or C but again they didn't [370:01] want to force their Engineers to learn [370:03] that they didn't want to people who code [370:04] in those languages have a different [370:05] culture than existed at Facebook 2 they [370:07] couldn't really recruit people so they [370:09] invented a new language called Hack That [370:12] was very similar to PHP but had static [370:14] typing over and over again I mean TOA [370:17] was another example this nosql database [370:19] they just keep finding ways where they [370:21] like invent new technology to solve a [370:24] problem that probably only exists for [370:25] them and then they create this like [370:28] whole Boutique system that allows a [370:32] having world class Talent B to have a [370:34] ton of them C everyone gets to move fast [370:37] but then D it's all unbelievably [370:39] performant and efficient and they just [370:40] don't have to make trade-offs it is wild [370:43] yep okay then my second technical one in [370:47] addition to this is they are their own [370:49] customer we've talked about this they're [370:51] a AWS scale technology company but they [370:55] don't take outside customers so they [370:57] only have to build for their own [370:59] internal use cases now this sounds great [371:02] but it actually does have these big [371:03] trade-offs you can't dramatically change [371:06] what your infrastructure is used for it [371:08] is purpose-built but it does let you be [371:10] incredibly efficient and have high [371:12] performance if you have good [371:14] communication between the customer sort [371:16] of the app team or the backend service [371:18] and the designers of that data center [371:21] this is completely the opposite of [371:22] Amazon Amazon uses interface so teams [371:25] don't have to talk to each other at meta [371:28] they require incredibly tight [371:29] communication it's a very different [371:32] organizational philosophy where they're [371:34] like no no not only do we not have [371:36] external customers we want this insane [371:38] type coupling between our infrastructure [371:40] and our internal [371:42] customers anyway my last one this is a [371:45] company that grows [371:48] intentionally it would be easy to look [371:50] at this company and say wow what a viral [371:54] product what a universally applicable [371:57] product that is not the case it is [372:00] unnatural to have connected 4 billion [372:04] humans this is a freak of nature this is [372:06] not just something that people adopted [372:09] and so while I think it is totally fair [372:12] to say wow it just like blew up at [372:14] Harvard Facebook is the story of 50 [372:17] different growth tactics in different [372:18] eras all carefully constructed and [372:20] itrated upon weight list at colleges [372:23] strategically launching at every single [372:25] College picking each one for a reason [372:27] translation internationalization zero [372:29] rating with carriers to bring new people [372:31] online figuring out when they needed to [372:33] acquire versus build carefully split [372:35] testing every change aligning the whole [372:38] company on specific networks on specific [372:40] metrics I mean building relationships [372:43] with governments in all these different [372:45] countries and at the very least [372:46] complying with local laws on where [372:48] should we have certain speech laws [372:50] versus not I mean it is completely [372:53] unnatural for them to have done what [372:54] they've done and it's all been very very [372:58] intentional to connect the world yep [373:01] okay last Playbook theme I've got [373:04] there's always another battle for [373:07] meta yeah it's funny I think you could [373:10] be listening and uh think like oh yes [373:12] there's always another platform battle [373:13] ahead for meta and that's probably true [373:16] but I think you actually mean like [373:18] another societal battle ahead for meta [373:20] right totally yeah now that they're [373:22] through the user privacy issues and the [373:25] many years of the whole 2016 election [373:28] conversation we talked about and I'm not [373:30] going to list them all here the many [373:31] many societal conflicts that they've had [373:33] over and over again the current issue [373:35] for them is around the impact of social [373:37] media on mental health and in particular [373:40] Teen Mental Health yep totally and it [373:43] feels fitting to put this near the [373:45] conclusion of the episode because while [373:47] going deep on this wasn't a part of our [373:51] understanding of how and why meta as a [373:53] business works so well it is a really [373:56] important topic there's a lot of people [373:58] making arguments that social media is [374:00] bad for our brains and the consequences [374:01] if that ends up being globally true is [374:04] catastrophic for meta probably a bigger [374:06] challenge than they've ever faced at any [374:08] other point in history so if you're [374:10] asking yourself what are the things to [374:12] keep an eye on going forward for them it [374:14] is of course all the product Innovation [374:16] and the growth of the existing business [374:18] and trying to invent the next platform [374:19] and everything we've talked about oryan [374:21] and everything yep yeah but it's also [374:24] how the mental health issue [374:26] understanding all that unfolds and how [374:29] they handle it yep totally all right [374:32] time to land the plane let's land the [374:34] plane how are you liking this by the way [374:36] this land the plane way of finishing [374:40] episodes ah well I actually have a [374:42] proposal for you Ben oh I don't think we [374:45] have landed pun intended on the right [374:48] nomenclature here or the right um [374:52] construct so I actually we've been [374:54] calling this land the plane takeway the [374:57] Splinter in our minds like what's the [374:59] one thing that's knocking around for you [375:01] yeah what is the essence of this company [375:03] so I propose that we change the name of [375:06] this final segment to [375:08] quintessence what is the quintessence [375:11] yes this is my inner French literature [375:14] major coming to Bear here un acquired [375:17] listeners what is the essence of this [375:19] company we've just spent all this time [375:21] all these months studying this company [375:24] what is the very essence of this company [375:26] that makes it different from any other [375:28] out [375:30] there all right I like it the [375:33] quintessence ultimately my big takeaway [375:35] is the company Moves Like Water it is [375:38] the company that has connected the world [375:40] that will always gear up for the next [375:43] battle and be whatever they need to be [375:46] in the next era and whether it's them [375:47] defining the next generation of [375:49] computing or creating all these AI [375:52] experiences or fending off the next Tik [375:54] Tok or the current Tik Tok like they [375:57] just move like water yep I totally agree [376:00] I can't think of a better [376:02] characterization of this company and how [376:05] they got to be so darn important in the [376:07] world perfect place to end it and [376:10] ultimately it is still very much a Mark [376:13] Zuckerberg production yes that too all [376:16] right quick carve outs quick carve outs [376:19] my quick carve out is a Google product [376:22] actually notebook LM this is freaking [376:24] wild our friend Ben Goen over at the [376:27] Wall Street Journal who wrote The Great [376:29] Piece on acquired a few months back [376:30] texted us what was this two weeks ago [376:32] maybe yeah a week ago and uh was like [376:35] have you guys tried this thing notebook [376:37] LM and I think you had and me being me I [376:39] hadn't I was like no let me check it out [376:41] I uploaded just the links to the sources [376:45] that I used for my side of the research [376:47] for the Microsoft part one episode just [376:51] links and when it spat back out at me I [376:53] was like holy crap yeah at this point I [376:56] think we're well past the Turing test [376:58] but it is sort of the most convinced [377:00] that something is actually a person I've [377:02] ever been and if I didn't upload the [377:05] sources and know that it was like [377:06] unbelievably tailored content to the [377:08] thing I just uploaded I'm not sure I [377:10] would know that it was AI it's pretty [377:12] amazing I have two one is a documentary [377:15] on Netflix called Mr McMahon I was not a [377:18] pro wrestling person growing up I kind [377:20] of want to go be a pro wrestling person [377:22] now this documentary is incredible it is [377:25] some of the best storytelling I've ever [377:26] seen and interestingly it's a [377:28] documentary that is told with no [377:30] narrator so there is story arc all [377:33] throughout the episode exclusively with [377:36] interview [377:38] answers and you almost don't notice at [377:41] some point you finish an episode and [377:42] you're like wait there was no narration [377:45] in that there was no cheesiness it was [377:47] all first party accounts and then cuts [377:50] to like old footage of things that aired [377:52] on TV and then you the credits come up [377:54] and of course it's a ringer production [377:57] Bill Simmons is the executive producer [378:00] it is remarkable and uh Mr McMahon is a [378:04] singular figure in the world certainly [378:06] not to be glorified but one to try and [378:08] understand oh man be fun to do a WWE [378:11] episode someday yes yes my second quick [378:14] one is the dwares podcast I love the [378:16] dwares podcast I also love dwares and I [378:19] think that if you like this show you'll [378:21] love listening to the interviews he does [378:24] uh most recent one or maybe a couple ago [378:26] is with uh Daniel Jurgen who is the [378:28] author of The Prize which is a book [378:31] David yeah you and I almost read a whole [378:34] bunch of it for standard oil and then we [378:36] realized Standard Oil is sort of over at [378:38] least the chapter of Standard Oil that [378:39] we were covering within like the first [378:41] two chapters of his book and so it's [378:42] basically everything from the end of our [378:45] Standard Oil episode forward on the [378:47] geopolitics of oil that end up shaping [378:50] and forming our world today and uh [378:52] dwares is just an amazing interviewer [378:54] and [378:55] conversationalist all right well with [378:56] that listeners a huge thank you to our [378:58] partners JP [ __ ] payments cruso stat [379:01] Sig and Huntress you can click the link [379:03] in the show notes to learn more [379:05] we talked to a ton of people for [379:07] research on this one and while we can't [379:09] mention everyone in part cuz the list [379:11] would just be too long in other part [379:13] made some folks ask not to be thanked we [379:15] do have some specific ones that uh we [379:17] want to give a shout out to so Alex [379:20] Schultz the CMO head of growth analytics [379:23] internationalization great to talk with [379:25] BOS Andrew Bosworth the CTO step ly who [379:29] wrote the book Facebook the inside story [379:31] was generous with his time Jim Brier who [379:33] led excel's series a investment in [379:36] Facebook Yan laon met as Chief AI [379:39] scientist Alex Heath at The Verge for [379:41] spending his time with me to Friend of [379:43] the show arvind Navar rotham from [379:46] worldly Partners who wrote an excellent [379:48] research report kind of chronicling [379:50] everything well it's almost like a [379:52] written version of this podcast it's [379:53] like a 100 page PDF that was awesome to [379:56] consume kind of to help me remember all [379:58] the big beats of the story and that his [380:00] research report is linked in the show [380:02] notes to Ariel Zuckerberg Mark's sister [380:05] Cheryl Sandberg obviously longtime coo [380:08] Mike shreer the former CTO and now [380:10] senior fellow to Pete hunt early [380:12] engineer who transferred from Facebook [380:14] to Instagram post acquisition Naomi [380:17] glight who we talked about founding [380:18] member of the growth team and the [380:20] longest employee longest 10e meta [380:23] employee at this point besides Mark yep [380:25] to Mike Vernal former meta VP of product [380:27] and engineering and former Sequoia [380:29] partner VJ rajie former engineer in VP [380:32] now of course CEO of stat Sig a part [380:34] Romani a VP in AI data and developer [380:37] infrastructure at meta to Owen vanada [380:41] Facebook's early coo and David I know [380:43] you have a few as well yeah a few for me [380:45] to Adam dandelo obviously Facebook's [380:47] first CTO Dan Rose early Partnerships at [380:51] Facebook and big big final thank you [380:54] that we owe to Chris Cox met as Chief [380:57] product officer and leader of the entire [380:59] family of apps over there really this [381:03] whole past couple months at acquired [381:05] would not have happened without Chris [381:07] cold emailing us what about a year ago [381:10] maybe and saying hey guys I love the [381:12] Nintendo [381:13] episodes and because of that we met [381:16] Chris and because of Chris Mark joined [381:18] us on stage at Chase Center and now here [381:19] we are doing this episode so thank you [381:22] Chris for making it all happen yes [381:24] essentially this episode came about [381:25] because we had done too much research [381:26] for the mark interview and we were like [381:28] we probably should do the actual meta [381:30] episode [381:31] to listeners it is time for our acquired [381:34] ual survey so if you have 3 to 5 minutes [381:36] please click the link in the show notes [381:38] or go to acquired. fm/ survey you might [381:41] win meta Ray bands you might win some [381:43] acq dad hats this is our one big ask of [381:46] the year and it really really helps make [381:48] the show better to hear your suggestions [381:49] feedback and to help show sponsors just [381:52] how impactful the acquired audience is [381:55] that is acquired. FMS survey check out [381:59] acq 2 in any podcast player if you like [382:01] this episode listen to our Nvidia series [382:04] listen to our Microsoft series uh I [382:07] don't know maybe go listen to our [382:08] Standard Oil series uh lot of grade [382:10] acquired in the back catalog and uh [382:13] discuss it with us in the slack [382:14] acquired. fm/ slack with that listeners [382:19] we'll see you next time we'll see you [382:20] next time who got the [382:23] truth is it you is it you is it you who [382:27] got the truth now [382:31] [Music]