At a Glance
Garry Tan (Founder and Managing Partner of Initialized Capital) discussed the "wrapping paper vs. gift" framework in venture capital, how to systematize investment decisions, and core lessons learned from Paul Graham in Episode 267 of Invest Like the Best. Core thesis: The true scarce resource in early-stage capital is not money, but trust — "In a world of infinite capital, the real scarcity is trust" (Garry Tan), and trust can only be accumulated through deep community engagement and long-term relationships, not replaced by algorithms.
Theme 1: From "Global Brain" to "Local Brain" — A Shift in Investment Perspective
Garry Tan argues that investors should not overemphasize the macro impact of global platforms like Facebook, but instead focus on the restructuring of the "local brain" within each organization.
- Historical Context: During a YC speech, Yuri Milner noted that approximately 10% of global energy flows to data centers, a proportion similar to the energy consumed by the human brain relative to the body. Tan thus likened the internet to a "global nervous system," but now believes a more pragmatic perspective is to focus on the "local" — every organization is undergoing a transition from traditional management to software-driven, data-structured operations.
- Mechanism Breakdown: Excel spreadsheets are merely "semi-structured data," essentially no better than paper folders. The real transformation occurs in companies like Flexport, which upgrade traditional industries such as logistics from "phone calls and emails" to structured data systems, achieving "better, faster, cheaper."
- Implication: Tan views this transformation as a "one-time shift," where every organization has the opportunity to redesign its own operations. Falsification Condition: If traditional enterprises can maintain competitiveness through simple digitization (e.g., deploying SaaS), then the value of the "local brain" theory is overestimated.
Theme 2: Systematic Venture Investing – From "Lone Wolf" to "Hive Mind"
Tan advocates replacing traditional VC's "individual heroism" with software and collective decision-making processes, centered on a two-step mechanism of "blind voting + dissent commitment."
- Mechanism Breakdown:
- Two-Round Blind Voting: After the founder leaves, all partners first vote anonymously in the software (first round), then engage in discussion ("Have you considered the PR risk?" "Did you see that blog post six months ago?"), followed by a second round of voting. Key point: Consensus is not the goal; instead, a "silver bullet" mechanism is retained—any partner who is extremely confident can unilaterally push a project forward.
- Speed Requirement: All partner meetings are arranged within 48 hours, and due diligence and investment decisions are completed within 72 hours. "Speed kills deals"—even with Initialized's brand, it cannot assume founders will wait.
- Analogy: Tan compares the decision-making process to an "internal dialogue while eating a burger"—one part of the brain wants to eat, while another says, "You're on a diet." A good investment decision is an "integrated decision," requiring partners from diverse professional backgrounds (lawyers, designers, journalists, engineers) to each express their genuine judgment.
- Data Support: YC's 2% acceptance rate (two batches per year, tens of thousands of applications for a few hundred spots) demonstrates the feasibility of systematic screening. Initialized has invested in hundreds of companies, producing 25 unicorns.
- Inference: Tan argues that traditional VC's "too much money chasing too few good ideas" is a self-fulfilling prophecy. Falsification condition: If a purely algorithm-driven early-stage investment platform emerges in the future (e.g., "press a button to get equity capital") with returns comparable to human VCs, Tan's "trust scarcity theory" would face a challenge.
Theme 3: How to Identify the "Gift" – Examine the Invention, Not the Wrapping Paper
Tan emphasizes that when evaluating startups, the focus should be on "the quality of the invention itself" (the gift), rather than the founder's resume, educational background, or other "wrapping paper."
- Historical Case: Instacart founder Apoorva Mehta insisted on joining YC even after the application deadline, sending a case of beer. After Tan downloaded the app, the key judgment came from "technical details"—scrolling through thousands of items was "smooth and fluid, with images loading in the correct thread." Dozens of similar apps existed at the time, but only Instacart achieved this. Steve Jobs' analogy: "Cabinetmakers don't look at the front; they turn it over and look at the back."
- Mechanism Breakdown:
- "The invention is the inventor": Matt Kohler's comment—"Who cares about the inventor? Look at the invention. If it's a good invention, the inventor is likely good too."
- "Talent magnet" effect: A good product attracts other great builders. "People who are truly good at building enjoy working with other builders"—this is the real currency in Silicon Valley.
- "Shelling point": A VC brand should become "the shelling point of shelling points"—attracting founders, who in turn attract engineers and designers, creating a self-fulfilling prophecy.
- Data Support: During his time at YC, Tan read tens of thousands of applications and found that founders with "Stanford CS degrees + Google/Facebook work experience" (gold foil wrapping) were often the worst. Counterintuitive Judgment: The most successful founders often come from non-traditional backgrounds.
- Deduction: Tan proposes a test criterion—"If I weren't doing my current job, would this company be one of the top 10 I'd most want to work for?" If all partners give an affirmative answer, it's worth investing in.
Theme 4: Two Major Causes of Failure and the "Idea Maze"
Tan summarizes the most common reasons for startup failure and introduces Chris Dixon's "Idea Maze" framework to explain the path to success.
1. Lack of Team Skills: The most common is "a non-technical founder who cannot code and has no technical co-founder." Tan suggests learning to code at least to the level where one can "converse with a technical co-founder."
2. Choosing a Non-Existent Market: Founders often "look for keys under the streetlight"—exploring areas that have been thoroughly studied, rather than delving into the unknown "Idea Maze."
- Mechanism Breakdown: Tan uses the "fog of war" analogy from StarCraft—when a startup enters a new market, the map is blurry, and only the immediate area can be explored. Key Insight: Sometimes, a "wall" that has existed for 20 years suddenly disappears (e.g., the proliferation of smartphones made Instacart's grocery delivery possible), and founders standing in front of that wall gain a significant advantage.
- Deduction: Tan advises founders to ask themselves, "What do we believe that others do not?" (a Peter Thiel-style question). If a business lacks an "anti-memetic" quality, it is usually not worth investing in.
Theme 5: Media as an Investment Tool — From "Few to Many" to "Many to Many"
Tan argues that media (especially Paul Graham’s blog) is a key lever for VCs to expand trust networks and reduce information asymmetry.
- Historical Context: 20–30 years ago, accessing capital required "joining the right club, being born in the right place, being the son of the right person." Today, the internet has unlocked the "labor potential of billions of people."
- Mechanism Breakdown: Paul Graham’s blog taught a generation that "you don’t have to have a job; you can be a founder." Tan likens each blog post to "teaching people to understand atoms" — once you grasp a fundamental principle, you can predict the next element. Each such insight could give rise to a company serving a billion people.
- Data Support: Naval Ravikant’s "many-to-many" media theory — in the past, a few broadcast to the many; now, everyone can participate. Tan believes "every business is now a meme."
- Implication: Initialized’s goal is to become "the Shell point of Shell points" — using media content to attract like-minded founders, then building a "mutual trust network" through community.
Theme 6: From Posterous’s Failure to Initialized’s Success — "Thriving in Chaos"
Tan attributes Initialized’s success to personal growth — from denying childhood adversity to embracing the trait of "thriving in chaos," and designing organizational culture accordingly.
- Personal Background: Tan is a descendant of Chinese immigrants, grew up in poverty ("sometimes dinner was bread dipped in milk"), and had an alcoholic father. He spent 20 years trying to be "normal" and ultimately accepted that he was "never normal."
- Mechanism Breakdown:
- "High Love, Low Structure": Through executive coach Cameron Yarbrough, Tan realized he needed a "high love, low structure" environment — establishing order amid chaos. However, most people require "high love, high structure."
- Posterous Lesson: In 2018, when the team grew from 20 people, Tan remained fixated on "writing all the code, doing all the design," failing to transition into a managerial role. The company ultimately ended in an acquisition (peak monthly active users of 100–200 million), falling short of the "billion-user" goal.
- "Integration": Tan uses "integration" to describe self-awareness — denying childhood experiences leads to a state of losing control, such as "face flushing without knowing why you’re angry." A more integrated state is "understanding the source of emotions, accepting them, and channeling them."
- Data Support: Initialized manages billions of dollars in assets, with six partners from diverse backgrounds (lawyer, NYT app developer, designer, journalist, etc.).
- Extrapolation: Tan believes "every founder has some form of adversity," even if they deny it. Falsification Condition: If a founder with no history of adversity emerges in the future and achieves great success, Tan’s "adversity-driven theory" would need revision.
Mentioned Positions
| Position |
Guest Stance |
Key Data |
| Coinbase |
Bullish (early-stage investment case) |
In 2011, Brian Armstrong sent 0.1 BTC as a demo; started from 1 person |
| Instacart |
Bullish (YC investment case) |
Apoorva Mehta insisted on joining YC after the application deadline; the app scrolls through thousands of products "smoothly and seamlessly" |
| Flexport |
Bullish (industry transformation case) |
Upgraded logistics from "making phone calls" to a structured data system |
| Formic |
Bullish (invested in Seed round) |
Lux led the Series A; target is to deploy thousands of robots with $100 million in annual net revenue |
| Rippling |
Neutral (Parker Conrad's "compound startup" model) |
Grew from 100 to 1,000 people |
| Posterous |
Risk alert (Tan's own failure case) |
Peak MAU of 100–200 million; acquired by Twitter; failure reason: founders failed to transition from builders to managers |
| BarkBox |
Neutral (background case) |
Rob Schutz grew revenue from $0 to $100 million/year in under 18 months |
| Upstart |
Neutral (analogy case) |
Programmatic debt capital model |
Judgments Worth Remembering
1. "In a world of infinite capital, the true scarcity is trust" (Garry Tan) — Early-stage investing cannot be replaced by algorithms because trust requires a "hard-to-join society" and long-term shared experience to accumulate. Support: YC's 2% acceptance rate proves that even with tens of thousands of applications, screening still relies on human judgment and community trust.
2. "Look at the invention, not the inventor" (Garry Tan, quoting Matt Kohler) — When evaluating startups, examine the quality of the product itself ("the cabinetmaker looks at the back"), not the founder's resume. Support: The Instacart case — among dozens of similar apps, only Apoorva's app achieved "smooth scrolling."
3. "A Stanford CS degree + Google/Facebook experience is gold foil wrapping, not a gift" (Garry Tan) — The worst founders often come from the most glamorous backgrounds. Support: Experience from tens of thousands of YC applications shows that people with the best resumes "are the worst founders."
4. "Speed kills deals" (Garry Tan) — In competitive early-stage investing, scheduling all partner meetings within 48 hours and completing due diligence within 72 hours is a prerequisite. Support: Initialized's "blind two-round voting system + silver bullet mechanism" ensures fast decision-making while preserving room for dissent.
5. "Every business is now a meme" (Garry Tan, quoting Naval Ravikant) — Media (especially blogs) is a key tool for VCs to expand their trust network. Support: Paul Graham's blog taught a generation that "you can be a founder," which is more effective than any sales pitch.
6. "Two main reasons for failure: lack of team skills and choosing a non-existent market" (Garry Tan) — Non-technical founders without a technical co-founder is the most common mistake; founders often "look for keys under the streetlight" (exploring already well-researched areas). Support: Chris Dixon's "idea maze" framework — when a startup enters a new market, the map is blurry, and it needs to "find its way by hitting walls with its face."
7. "I need to thrive in chaos, but most people need high love and high structure" (Garry Tan) — Tan realized his own traits through an executive coach and designed Initialized's organizational culture accordingly. Support: The Posterous failure — when the team grew from 20 people, Tan failed to transition from builder to manager, causing the company to stall at 100–200 million monthly active users.
8. "Every founder has some form of adversity" (Garry Tan) — Even if founders deny it, childhood experiences shape their drive. Support: Tan himself spent 20 years trying to be "normal" and eventually accepted that he was "never normal," which became part of his investment judgment.