Paul Graham Says Operating Like Nonprofits Gives Startups Market Edge
startupschool
Startups that prioritize user benevolence over early monetization survive high-mortality downturns and secure decisive long-term pricing power against entrenched incumbents.
AI-assisted summary of Paul Graham at Startup School 08 — original source. Check the original for full context.
Source published:
Key ideas
- Combining Y Combinator's core tenets—make something people want and defer monetization—reveals that the most successful early-stage startups functionally operate like benevolent nonprofits. (0:02)
- Benevolence acts as a competitive moat and moral compass, protecting founder morale, attracting world-class engineering talent, and establishing a 'weather gauge' advantage over revenue. (2:04)
- Launching quickly triggers a 'Tamagotchi effect' where real, daily dependent users motivate founders to persist through severe emotional downturns and near-death experiences. (10:44)
- Operating purely in the interest of users provides a 'stateless' decision heuristic that dramatically reduces cognitive overhead and prevents unscalable corporate exceptions. (17:27)
Detailed notes
## Background Published on YouTube by the startupschool channel on January 30, 2009, this video captures Y Combinator co-founder Paul Graham speaking at the Startup School 08 conference. According to the video description, the talk provides foundational advice from Graham on how to create a successful startup.
## TL;DR Paul Graham argues that combining Y Combinator's core advice—making something people want while deferring monetization—reveals that the most resilient startups initially operate like benevolent nonprofits. Drawing on examples such as Craigslist, early Google, and Octopart, he demonstrates that acting purely in the interest of users protects founder morale through the 'Tamagotchi effect,' attracts top engineering talent, and provides a clear, stateless heuristic for decision-making. Urging founders to surpass 'Don't be evil' by actively striving to be good, Graham frames benevolence not as moral righteousness, but as an indispensable competitive strategy. During the closing Q&A, he discusses the risks of dropping out of college prematurely, how to locate startup ideas by observing broken problems, and why commercial execution remains a prerequisite for achieving social good.
## Atmosphere & style The presentation carries an unpolished, intimate, and refreshingly candid vibe, opening with Graham dryly recounting how he had to reconstruct his presentation on the podium due to unreleased presentation software malfunctioning. He delivers his insights with self-deprecating humor and pragmatic intellectual clarity, disarming audience cynicism by admitting he is no saint and merely speaks good as a second language. The session feels like an unfiltered fireside chat within early Silicon Valley culture, oscillating smoothly between high-level philosophical thought experiments and frank, spontaneous audience sparring over university degrees and venture capital realities.
## Bulletin - [00:02](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=2s) Combining Y Combinator's core tenets—make something people want and defer monetization—reveals that the most successful early-stage startups functionally operate like benevolent nonprofits. - [02:04](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=124s) Benevolence acts as a competitive moat and moral compass, protecting founder morale, attracting world-class engineering talent, and establishing a 'weather gauge' advantage over revenue. - [10:44](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=644s) Launching quickly triggers a 'Tamagotchi effect' where real, daily dependent users motivate founders to persist through severe emotional downturns and near-death experiences. - [17:27](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1047s) Operating purely in the interest of users provides a 'stateless' decision heuristic that dramatically reduces cognitive overhead and prevents unscalable corporate exceptions.
## Key takeaways - [00:02](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=2s) Y Combinator's core advice to build what people want without immediate monetization paradoxically mirrors the operational blueprint of a nonprofit organization. - [02:04](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=124s) Securing the 'weather gauge' of revenue means building massive user goodwill upfront so that monetization can be engaged at will rather than forced via aggressive sales. - [02:04](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=124s) Google operated for its entire first year with zero advertisements, functioning essentially as a benevolent web index before monetizing. - [02:04](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=124s) Startups and benevolent public projects naturally converge at scale, as seen when open-source initiatives require corporate structure to survive. - [08:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=519s) Young startups must rely on goodwill and charm because they lack the market leverage to bully customers the way large incumbents do. - [08:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=519s) Paul Buchheit's motto 'Don't be evil' functions as an elixir of corporate youth that protects companies from organizational lethargy and hostility. - [08:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=519s) Founder morale is the single most critical survival variable; emotional despair paralyzes founders and becomes a self-fulfilling prophecy. - [10:44](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=644s) The 'Tamagotchi effect' occurs when an early launch introduces live users whose reliance keeps founders working through catastrophic downturns. - [12:42](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=762s) Indifference to outside capital makes startups hard to kill and paradoxically increases their appeal to venture capitalists. - [14:41](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=881s) Idealistic missions act as talent magnets because elite software engineers have their choice of employers and prefer projects that improve the world. - [17:27](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1047s) Focusing strictly on what is best for users is an inherently 'stateless' strategy that removes the mental friction of managing hidden agendas or corporate exceptions. - [21:52](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1312s) Crafting corporate mission statements should never delay launching a product into the hands of real users. - [21:52](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1312s) Undergraduates are generally better off remaining in school to mature and learn rather than impulsively dropping out to pursue early, half-formed ideas. - [26:08](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1568s) Viable startup ideas are best discovered by observing broken problems in daily work rather than attempting to forecast future industry buzzwords. - [27:24](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1644s) Infrastructure startups appear underrepresented in accelerator portfolios primarily because they require far longer development cycles before public launches. - [29:03](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1743s) Commercial execution ability is the fundamental limiting factor on benevolent impact, as failed startups never get the opportunity to do good.
## Detailed walkthrough ### [00:02](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=2s) The Startup Accidental Nonprofit Paradox
Paul Graham opens his presentation with a candid behind-the-scenes look at the high stress of startup life, noting that he had to remake his presentation right at the podium while the audience watched. He reveals he was using an unreleased web-based presentation software called 280 Slides, built by the startup 280 North, dryly adding that it was fortunate their team was standing around when it malfunctioned.
Graham then reflects on the earliest days of Y Combinator. About a month after launching the accelerator, the partners settled on a phrase that became their permanent motto: "Make something people want." Graham observes that although they have changed their minds on many startup topics over the years, this principle remains untouchable, remarking that if it were possible to compress it to fewer than three words, he would be the first to try. The second most frequent piece of advice Y Combinator gives to founders is not to stress excessively over their business model or monetization at the outset—not because making money is unimportant, but because monetizing is far easier than building something great. However, Graham explains that combining these two core tenets leads to an unexpected revelation:
> “What you've got is a description of a nonprofit.”
When faced with such a paradox, Graham suggests there are only two possibilities: either it is a logical defect proving by *reductio ad absurdum* that one of the premises is false, or it represents an entirely new discovery. He concludes it is the latter, explaining that viewing startups through the lens of benevolent nonprofits makes a multitude of otherwise puzzling startup realities suddenly fall into place.
```mermaid flowchart TD accTitle: The Startup Non-Profit Paradox accDescr: Flowchart showing how combining Y Combinator's two main rules leads to the concept of a nonprofit and the synthesis of benevolent companies.
subgraph Advice["YC Core Advice"] A["Rule 1: Make something people want"] B["Rule 2: Don't worry about money at first"] end
subgraph Paradox["The Synthesis"] A & B --> C{"Combined Outcome"} C --> D["Operates like a Nonprofit"] end
subgraph Resolution["Theoretical Implication"] D --> E["Not a contradiction / bug"] E --> F["Discovery: The most successful startups behave benevolently"] end ```
---
[02:04](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=124s) Craigslist, Google, and the "Weather Gauge" of Revenue
To illustrate why operating like a charity can be a sound corporate strategy, Graham points to Craigslist. Although Craigslist is technically a for-profit entity, it is run largely like a nonprofit, with an employee count so small in proportion to its traffic that on lists of top websites it looks like a misprint. Graham estimates that Craigslist likely generates more revenue per employee than Google does. While they could extract substantially more money if they focused intensely on monetization, almost any conventional startup would eagerly trade places with them.
To describe their structural advantage, Graham borrows a naval warfare metaphor from Patrick O'Brian’s Aubrey–Maturin novels, where sailing captains constantly battle to secure the "weather gauge"—the upwind position relative to their adversary:
> “If you're up wind, you decide when and if to engage the other ship.”
In this analogy, Craigslist sits firmly upwind of massive revenues; turning on monetization would simply be a matter of managing public relations and spin. That is a vastly superior position to tacking upwind against the gale—spending ten times as much on aggressive sales teams as on engineering just to force an inferior product down the throats of ambivalent customers. Graham clarifies that founders should not necessarily aim to mimic Craigslist’s permanent structure, which stems from unique historical conditions, but should absolutely copy it during their initial phases.
Google followed this exact pattern: during its entire first year, it had zero advertisements. Year-one Google was essentially the mathematical limit of what a purely benevolent web index would have been. Graham recounts his own past experience building spam filters, where he conceived a free webmail service with world-class spam protection simply because the world needed it. He ultimately abandoned the project because he realized that to execute it properly, it could not survive merely as an open project; it had to be organized as a commercial business. Startups and benevolent projects continually converge: when you try to execute a purely benevolent project at scale, it often takes the shape of a company, and the most enduring companies often start out acting like public services.
---
[04:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=279s) How Far Does Benevolence Reach? From Malaria to Microsoft
Examining the boundaries of this dynamic, Graham asks whether all worthwhile nonprofits could function as companies. Skeptics might argue Google succeeded because its user base had money to spend, meaning benevolence toward wealthy people eventually yields lucrative returns. But Graham questions whether benevolence could work even when directed at populations without money. Taking an extreme example, he contemplates whether an enterprise could generate wealth by tackling an unfashionable, devastating disease like malaria. Because endemic malaria acts as a severe drag on the economic output of entire regions, lifting that burden would unlock massive growth that an organization might participate in. While humorously warning the audience not to submit Y Combinator applications to cure malaria unless they hold a genuinely brilliant breakthrough, he uses the thought experiment to show how dangerous it is to bet against the economic utility of benevolence.
Benevolence is not limited to consumer internet software. Graham recalls that in its earliest days, Microsoft operated much like Robin Hood relative to the incumbent monopoly, IBM:
> “When IBM produced the IBM PC, they thought they were going to make money by selling people hardware at really high prices.”
Microsoft broke IBM's model by establishing control of the operating system and opening the hardware standard to any manufacturer. This triggered a price collapse that made personal computers affordable to millions who could never have bought one previously—an outcome remarkably similar to modern Google initiatives.
Over time, Microsoft's posture inverted. Graham jokes that today, all the verbs associated with Microsoft's treatment of users seem to begin with the letter *F*, such as "frustrate" and "foist"—a trend exemplified by forcing Windows Vista onto customers who preferred Windows XP, or simply driving them toward Apple's OS X. Yet Graham points out that bullying has not rewarded Microsoft financially; its stock price remained stagnant for years, whereas during its benevolent, "Robin Hood" phase, its valuation climbed rapidly.
---
[08:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=519s) The Elixir of Corporate Youth and the Three Powers of Benevolence
Graham explains the underlying corporate life cycle: young startups lack market leverage, so they cannot bully their customers and must rely entirely on charm and goodwill. Large incumbents, by contrast, can exploit and mistreat their customer base simply because bullying requires less effort than building great products. Companies grow large by being nice, but they often stay large by being mean—a dynamic evident in the historical conduct of the major record labels. This bullying works only until technological or market shifts provide an opening, at which point their captive users break free and escape.
Because of this trap, Graham argues that Paul Buchheit's famous corporate motto for Google, "Don't be evil," may be his most valuable creation—outstripping even Gmail:
> “Don't be evil may actually be the most valuable thing that Paul Buchheit made for Google, more valuable than Gmail because it may turn out to be an elixir of corporate youth.”
Even if Google's leadership finds the motto restrictive, its real value is safeguarding the firm against the organizational lethargy and hostility that doomed IBM and Microsoft. Best of all, this "elixir" is unpatented and freely available to any startup bold enough to adopt it, though legacy industries such as tobacco firms or record labels could never use it because society would hold them accountable to the standard.
Drawing on Y Combinator's dataset from funding dozens of early-stage teams, Graham breaks down the concrete mechanisms through which benevolence empowers a young company into three distinct advantages: - **It protects founder morale**, keeping the team resilient through volatile swings. - **It makes outside people eager to help you**, turning users and observers into active supporters. - **Above all, it provides moral clarity that makes founders decisive.**
Focusing on the first point, Graham emphasizes that founder morale is the single most critical variable in determining whether a startup lives or dies. Founders inevitably endure a violent emotional roller coaster, oscillating between visions of world domination and crushing despair. The fatal danger of despair is not merely personal unhappiness, but that the sensation of feeling doomed paralyzes founders and stops them from doing the work.
[10:44](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=644s) The Tamagotchi Effect and Surviving the Downhill Slopes
Paul Graham discusses the psychological dynamics of startups, emphasizing how the emotional lows are far more dangerous than the highs because they act as self-fulfilling prophecies. While optimistic momentum naturally drives founders to work harder and improve their odds, despair causes them to stop working altogether, which virtually guarantees collapse. Genuine benevolence serves as a critical defense against this fatal stall: when founders feel they are genuinely helping people, they maintain the motivation to push through even when facing apparent doom. Graham points out that humans possess an innate drive to assist others who depend on them. When founders build an application that requires daily user engagement, they essentially create an emotional dependent:
```mermaid flowchart TD accTitle: The Tamagotchi Effect Feedback Loop accDescr: Diagram showing how rapid launch triggers the Tamagotchi effect and user demand to keep founders working. subgraph Launch["Launch Stage"] A["Launch Quickly"] --> B["Acquire Daily Active Users"] end subgraph Survival["Survival Loop"] B --> C["Tamagotchi Effect Kicks In<br/>(Duty to Care)"] C --> D["Founders Keep Working<br/>Through Lows"] D --> E["Discover True User Needs"] E --> B end ```
Graham interrupts himself to nervously check on the performance of his own systems, asking how the server is holding up and whether the response times are solid. He points to Blogger as the quintessential example of weathering catastrophic downturns. When Blogger ran out of money, the entire team left until Evan Williams was the only person who returned the following day. Graham explains that the sole anchor keeping Williams at his desk was the undeniable fact that active users still relied on the service: “Beneath everything else like the hope of making money from a startup or the fear of looking like a failure to your friends, you have like this server with all these users that they need you and you know the response time has to be good.” Launching rapidly is invaluable because it immediately triggers this "Tamagotchi effect," compelling founders to discover exactly what users want.
[12:42](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=762s) Indifference to Investors and Becoming Hard to Kill
Helping users breeds authentic confidence, which radically shifts the balance of power when dealing with venture capitalists. Graham shares an anecdote about Chatterous (a chat service built specifically for events like Startup School). The two founders sat down and made a resolute pact: what they were building was so critical that they would see it through regardless of outside support, even if it forced them to relocate to Canada and live in their parents' basement. The moment they embraced that mindset, they stopped worrying about what prospective investors thought. Paradoxically, the moment they walked into pitch meetings unconcerned with whether they received funding, investors became substantially more interested because they realized the founders did not need them. Graham notes that even Y Combinator operates under this filter:
> “Frankly, we're looking for startups that would do better with us, not startups that would die without us.”
Graham insists that being lean and deeply committed makes a startup virtually indestructible, urging founders to adopt Craigslist and cockroaches—rather than bloated failures like Webvan—as their true spiritual models.
- Extreme frugality combined with mission-driven conviction makes a company "very hard to kill." - Virtually every startup, even the most legendary success, faces near-death experiences, and founders are lucky if they only encounter that crisis once. - The intrinsic resilience gained from a meaningful mission far outweighs any short-term advantage gained from chasing selfish, shallow ideas.
[14:41](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=881s) Good vs. Evil: The Octopart Story and Hacker Magnetism
Graham argues that benevolence naturally inspires outside support because humans are evolutionarily wired to help those who do good. He points to Y Combinator-funded Octopart, an industrial component search engine founded by two Berkeley physics Ph.D. dropouts seeking to solve an issue they faced in their own lab research. Although parts search might sound unglamorous, Octopart created a functioning discovery tool that users embraced. However, they found themselves locked in a classic battle against Digi-Key, the dominant industry distributor. Despite Octopart sending free, paying customer traffic to Digi-Key, the distributor actively fought to force Octopart to remove their pricing data. Digi-Key's existing business model relied on overcharging customers trapped by incomplete pricing information, making transparent search a threat to their margins.
Graham explains that Digi-Key's bullying triggered an aggressive desire within him to champion Octopart, leading him to dedicate far more time and public advocacy to their cause than to an ordinary company. Benevolence attracts an entire ecosystem of advocates—customers, investors, partner companies, and especially top engineering talent. Top hackers are rarely desperate for employment; because they can choose to work anywhere, they gravitate toward idealistic missions that improve the world. As Graham frames it: “So, if you have a project that makes the world better, you're like a magnet for all the smart people to come and work for you.”
[17:27](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1047s) The Moral Compass and Stateless Decision-Making
Benevolence solves the agonizing dilemma of resource allocation. Early-stage teams face thousands of tasks with only two or three founders available to execute them, making vacillation fatal. Graham offers an absolute heuristic: “Here's the answer: just do whatever is best for your users, right? You can hold on to that like a rope in a hurricane and it will save you if anything can.” Focusing entirely on making users happy even solves unrelated startup hurdles, such as fundraising. While charismatic founders might attempt to smooth-talk investors, most engineers lack sales polish; the most direct route to investment is generating explosive word-of-mouth growth among real users. Showing the Octopart traffic chart on the screen, Graham highlights their classic exponential hockey-stick curve (pointing out a brief seasonal dip by joking that nobody buys industrial components as Christmas gifts).
Graham highlights the operational efficiency of altruism: doing good is an inherently "stateless" strategy. Just like telling the truth frees a person from tracking past lies, acting purely in the best interests of users removes the mental overhead of hidden agendas. This stateless principle is essential for scaling:
| Entity | Scale / Challenge | Stateless Operating Principle | Operational Advantage | |:--- |:--- |:--- |:--- | | **Y Combinator** | Advising 57 active portfolio companies (out of 80 total investments) simultaneously | “Our rule is just try and do whatever is best for the startups, right?” | Eliminates ulterior motives that advisors would otherwise have to memorize across dozens of boardrooms. | | **Google** | Ranking millions of search results across the entire web | Strictly refuses to manipulate or "munch" organic search results for commercial favors | Avoids an unscalable web of bespoke exceptions, maintaining automated algorithmic consistency. |
Closing this line of thought, Graham clarifies that he does not preach virtue out of personal righteousness, acknowledging candidly that he was never an exceptionally noble kid himself.
[20:44](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1244s) Speaking Good as a Second Language and the Case for Benevolence
Paul Graham closes his talk by confronting the natural cynicism people feel when someone preaches morality from a stage. Recalling his childhood, he notes that whenever adults used the word "good," it always seemed synonymous with "quiet," making him grow up deeply suspicious of the term. He readily admits that people do not describe him as a saintly figure: the best compliment he usually receives is that “he means well.” Acknowledging his own limitations, he explains that he is merely average and speaks good as a second language.
Because of this, Graham insists his advice should not be dismissed as self-righteous posturing. He advocates goodness not out of moral superiority, but because it functions as an effective, practical strategy. Benevolence can serve as a guide to corporate strategy and a functional design specification for building software, leading directly to a higher standard than Google's famous motto: “So don't just not be evil, be good.”
[21:52](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1312s) Company Mottos and the Value of Staying in School
Transitioning to questions from the audience, Graham consults the event clock and jokes that he can fetch a question from memory and execute it. An attendee named Gabor references Jim Collins's book *Built to Last*, noting that enduring, massive companies frequently center themselves around credos or mission statements, and asks whether early-stage startups should prioritize adopting mottos. Graham concedes that Google has one and that it might be worth encouraging, but cautions against letting it become a distraction. If agonizing over a motto delays a launch by even a few hours, founders are better off waiting until after their product is out in the wild. When someone in the crowd quips about organizing the world's information, Graham chuckles and moves on.
```mermaid flowchart TD accTitle: Practical Startup Guidance Flow accDescr: Flowchart illustrating Paul Graham's advice on education, idea generation, and founder selection.
subgraph CollegeDecision["Education Decision"] A["In College with an Idea"] --> B{"Drop Out or Stay?"} B -->|"Recommendation"| C["Stay in School: Learn, Meet People, Mature"] B -->|"Common Trap"| D["Drop Out on a Premature Idea"] end
subgraph IdeaFinding["Idea Generation"] E["Observe Daily Work & Friction"] --> F["Identify Broken Problems"] F --> G["Let Problems Drive Solutions"] end
subgraph YCSelection["Startup Funding Criteria"] H["Candidate Evaluation"] --> I["Likelihood of Startup Success"] I --> J["Ability to Execute Benevolent Impact"] end ```
Another audience member brings up an article on Graham’s website regarding whether it is better to obtain an undergraduate degree versus a graduate degree, asking whether an undergraduate with a compelling idea should drop out immediately to make money. Graham advises against prematurely rushing into the startup world, pointing out that initial ideas are rarely as brilliant as founders think they are in the moment. College offers distinct value beyond wealth generation—allowing students to learn, socialize, and build friendships. Graham emphasizes that waiting rarely hurts: “And it's certainly even more likely that if you wait till you finish school, there'll be some other idea you'll think of that will be even better, right?”
- Founding teams should launch software first rather than delaying to craft brand mission statements or corporate credos. - Dropping out of college is often romanticized because famous founders did so, but college provides time to explore concepts and mature as a thinker. - Better startup ideas naturally surface later as personal experience and technical knowledge expand.
[25:01](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1501s) The Debate Over Dropping Out to Pursue Startups
A member of the audience strongly objects to Graham’s defense of staying in school, arguing that fifteen years of consulting alongside executive advisers at companies like Apple, Cisco, and Nokia proved the opposite. The commenter insists that degree holders inevitably end up working for the dropouts who have the courage to execute, asserting that school will always be there, whereas founders must follow their passion and ignore their parents. Graham responds with lighthearted, noncommittal humor, refusing to be drawn into an ideological war against higher education while speaking from the stage of a university. As the commenter urges everyone to follow their passion regardless of school, Graham simply smiles and replies, “Maybe.”
[26:08](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1568s) Finding Ideas by Finding Broken Problems
The session shifts to the mechanics of ideation when an attendee asks how founders can anticipate technological trends three or four years out—such as transitioning from Web 2.0 to Web 3.0—without pitching redundant concepts that competitors are already prototyping. Graham rejects the approach of trying to forecast buzzwords or consciously inventing futuristic solutions from scratch. Instead, he argues that the best way to uncover viable startup concepts is to focus on genuine friction in the present day.
```mermaid flowchart LR accTitle: Problem-Driven Ideation accDescr: Flowchart contrasting the flawed top-down approach to ideation with Graham's problem-driven methodology.
subgraph IneffectiveApproach["Top-Down Ideation"] X["Speculate on Future Buzzwords"] --> Y["Invent Theoretical Solution"] Y --> Z["Discover Competitors Already Built It"] end
subgraph EffectiveApproach["Problem-Driven Ideation"] A["Work in an Applied Field"] --> B["Notice Broken Everyday Tools"] B --> C["Octopart: Component Search"] end ```
He recommends searching for things that are noticeably broken and letting those problems dictate the resulting product. To illustrate, he points to Octopart, a Y Combinator company created by applied physics researchers who struggled with the fragmented, cumbersome process of purchasing electronic components for their lab. Graham quips that if the founders had been theoretical physicists rather than applied researchers, they probably would have launched a website to buy yellow legal pads. As Graham puts it, “Don't don't try and think of solutions. Just try and find good problems and let them drive the solutions.”
[27:24](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1644s) Web 2.0 Versus Infrastructure and What VCs Actually Fund
A questioner challenges Y Combinator's portfolio composition, asking why there appears to be such an overwhelming emphasis on lightweight Web 2.0 applications instead of deep, fundamental technical problems such as computer security, core search, or software that accelerates hardware performance. Graham pushes back on the premise that building an electronic component search engine is not a fundamental problem, and clarifies that YC has indeed funded heavy infrastructure projects, including two database companies. The reason infrastructure startups lack the high visibility of social consumer software is simply that they take far longer to build and launch.
Graham explains that Y Combinator ultimately funds what comes through the applicant pool. The portfolio reflects user submissions, and a significant portion of applicants inevitably pitch either a variation of Facebook or an application geared toward independent music. When the audience member asks why traditional venture capitalists fund so many Web 2.0 companies, Graham explains that investors can only catch the fish swimming down the stream in front of them rather than inventing companies out of thin air.
- Infrastructure companies (such as database engines) require lengthy gestation periods before making public launches, creating the illusion that incubators only fund consumer web tools. - Startup applications consistently cluster around prevailing cultural trends, especially social networks and independent music platforms. - Investors and incubators are inherently constrained by inbound applicant distribution.
[29:03](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1743s) Benevolence and Success in Startup Selection
The final question addresses how Y Combinator balances the benevolence factor against purely commercial metrics when choosing which applicants to back. Graham explains that funding founders they genuinely like began as an act of personal self-indulgence. However, the overarching selection filter remains whether the founders possess the capability and drive to actually succeed. Pragmatically, benevolence alone is completely inert without commercial execution: a well-meaning founder who fails never gets the opportunity to do good in the world.
Graham compares investing to athletics, noting that just as a soccer player aims to kick the ball into the opponent's net rather than their own, an investor must back companies capable of winning. Execution ability sets the absolute ceiling on any charitable or social outcome. Graham wraps up his address by connecting commercial competence back to core moral intentions: “So basically succeeding as founders is a limiting factor on how benevolent we can get.” Because benevolent founders naturally build things that work, choosing founders poised for success ultimately drives the most ethical and productive outcomes.
## Notable quotes > [12:42](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=762s) Frankly, we're looking for startups that would do better with us, not startups that would die without us. — Paul Graham
> [26:08](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1568s) Just try and find good problems and let them drive the solutions. — Paul Graham
> [17:27](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1047s) Here's the answer: just do whatever is best for your users, right? You can hold on to that like a rope in a hurricane and it will save you if anything can. — Paul Graham
> [08:39](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=519s) Don't be evil may actually be the most valuable thing that Paul Buchheit made for Google, more valuable than Gmail because it may turn out to be an elixir of corporate youth. — Paul Graham
> [14:41](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=881s) So, if you have a project that makes the world better, you're like a magnet for all the smart people to come and work for you. — Paul Graham
> [20:44](https://www.youtube.com/watch?v=q7K0vRUKXKc&t=1244s) So don't just not be evil, be good. — Paul Graham
## Mentioned Paul Graham, Y Combinator, 280 Slides, 280 North, Craigslist, Google, Patrick O'Brian, IBM, Microsoft, Windows Vista, Windows XP, Apple, OS X, Paul Buchheit, Gmail, Evan Williams, Blogger, Chatterous, Octopart, Digi-Key, Jim Collins, Built to Last, Facebook, Webvan, Cisco, Nokia
## Video description Paul Graham, founder of YCombinator, speaks at Startup School 08 about how to create a successful startup.
[Source](https://www.youtube.com/watch?v=q7K0vRUKXKc)