Sequoia Capital's Investment Committee Mechanism Dissected: Partner Julien Bek Deeply Reviews Top Deals and Project Judgments

Julien Bek
Sequoia

Original Statement

1. The Internal Operation Mechanism of Sequoia Capital and the Truth about the Investment Committee 1. Decision-making Process of the Investment Committee: A Combination of Fast and Slow Thinking • Monday IC Meetings and Asynchronous Memos: Sequoia's Monday Investment Committee, which has been running for decades, is evolving; the team is implementing an "asynchronous collaboration mechanism"—after the investment memo is sent out, partners independently write evaluations and questions in the document. • Asynchronous Mode (Slow Thinking): Provides space for thoughtful and independent judgment, avoiding groupthink. • On-site IC (Fast Thinking): Partners directly defend their positions against founders, facilitating high-frequency interactions. • Voting Mechanism and "1 Point Decision": Partners score from 1 to 10. Even if a senior partner gives a very low score of 1, if the project sponsor has the highest conviction, they still have the final authority to greenlight the project (provided the sponsor takes full reputational responsibility for the outcome). • SpaceX's Classic Comeback: Partner Shaun Maguire strongly advocated for SpaceX despite significant internal resistance, with some partners scoring it as low as 1; Shaun insisted on not giving up, requiring all partners to visit the site to witness it firsthand, ultimately resulting in one of Sequoia's most lucrative investments in history. 2. Everyone is a "Hunter" • Breaking the Myth of "Waiting for the Phone Call": The early team consisted of only 11-12 people, operating like a high-intensity football team, with everyone actively seeking non-consensus opportunities. • Heavy Investment in Long-term Relationships: For example, partner Constantine built trust with Ken Griffin during college, maintaining continuous follow-up over the years, which eventually led to Ken's acceptance of external institutional investment. • Updating Priors: In decisions regarding giants like Anthropic, Sequoia demonstrated the ability to quickly overcome the pride of having previously rejected them, decisively entering with a higher valuation ($2.5 billion round) after recognizing the reality of AI's exponential growth. 3. "Our Value Depends Only on Our Next Investment" • A core warning phrase is printed on the wall at Sequoia: "We are only as good as our next investment." Regardless of past myths created, each fund review resets history and refocuses on the next generation of great companies. 2. The Identification Map and Penetration Rules of Top Founders 1. Evaluation Framework: Direction and Magnitude • Pat Grady's Vector Model: Outstanding talent is like a vector: • Direction: Why are they doing this? Are their motivations pure and steadfast? • Magnitude: How ambitious are they? Can they endure extreme pain to keep pushing forward? • Shaun Maguire's Four-Dimensional Assessment: Beyond traditional IQ and EQ, it focuses on judgment (the ability to find solutions in complex systems) and political quotient (PQ, the ability to break through in complex organizations/networks of interests). • Distinguishing Top Executives from Exceptional Founders: Alfred Lin warns against mistaking "top operators from star companies" for "exceptional founders who can break through from 0 to 1." 2. 30-Minute Icebreaker and "5 Whys" • Trading Vulnerability for Sincerity: Julien breaks the cold commercial pitch by sharing personal family experiences, guiding founders to reveal their genuine growth trajectories and inner motivations. • Penetrating Deception and Disguise: By continuously asking "why" about decision details, one can observe speech speed, micro-expressions, and body language. A founder suspected of providing false data revealed inconsistencies under questioning, ultimately being exposed and warned by the industry. • Cross-Cultural Calibration: • German/French clients and founders are usually very restrained, with an NPS score of 7 often corresponding to actual high satisfaction (requiring an internal +1-2 points). • The American team expresses with great enthusiasm and showbiz style (requiring an internal -1-2 points for calibration). 3. Founder Background Check Rules (Doug Leone Methodology) • Who is your best reference? Who would give you the worst evaluation?: When a founder provides their best reference, immediately ask, "Who would give the worst evaluation and why?" This counterintuitive question reveals their self-awareness and honesty, allowing for deep cross-referencing. • ELO Rule: Only a grandmaster with a rating of 2400 can accurately identify another grandmaster in the crowd; one must seek references from truly exceptional individuals. 3. Paradigm Shift and Business Simulation in the AI Era 1. "Agents are a New Form of Customer" • Machine Traffic Surpassing Human Traffic: In the coming years, the traffic from agents on the internet will reach a thousand times that of human traffic; shifting from optimizing UI/pixel conversion rates over the past 20 years to building a "Bits-Perfect" underlying platform that adapts to agent calls. • Bias of Agents and AEO (Answer Engine Optimization): Agents have biases due to pre-training and post-training data (e.g., during development, they may prefer to call Cloudflare and Vercel); companies optimizing their weight in AI Q&A and agent decision-making (AEO) will give rise to a new ecosystem worth billions of dollars. 2. The Ultimate Form of Software Companies: "Software Companies in Service Clothing" • From Co-pilot to Autopilot: The first wave of AI sold tools for human use (capturing $1 of tool expenditure), while the real trillion-dollar opportunity lies in directly selling the final delivered results (outcomes), thus consuming the traditional outsourcing and human service budget of $6 (e.g., Sierra charges based on resolved tickets in the intelligent customer service sector). • "A Lot of AI + A Few Human Professional Judgments": Utilizing AI to complete 90% of automated execution, supplemented by human final aesthetic and business judgment at key points, companies will disrupt the massive traditional service industry with the high gross profit structure of traditional software. 3. Vertical Market vs Replacement Market (Greenfield vs Replacement) • Explosive Potential of Greenfield Markets: New capabilities (e.g., Lovable, Cursor, Harvey) can achieve tens of millions or even hundreds of millions in ARR at an incredible speed. • Penetration of the Real Economy (Project Iowa): For example, when traditional non-tech entities like car washes and auction houses begin to fully procure and consolidate core financial data, their businesses will truly possess cross-cycle stickiness and high barriers. 4. Top Investment Principles and Classic Case Reviews 1. Investment Preference: Extreme Traits (Spikiness) Over Mediocre Balance • Don Valentine's 2x2 Matrix: "Founders you like" and "Founders who can make money" are not the same. Top founders are often controversial, arrogant, or have distinct personalities; the core responsibility of investors is to identify whether their extreme strengths (spikes) are sufficient to support building an empire. • Reflection on Missing Trade Republic: Previously missed early investment due to misjudging that Revolut would quickly crush competitors; deeply realizing that trillion-dollar super tracks can often accommodate multiple giants coexisting, not an absolute zero-sum game. 2. The Family Angel Investment Legend of Revolut • Julien keenly sensed the extreme focus and fierce execution of Revolut founder Nikolay Storonsky just two weeks into his venture capital career; after the institution missed the lead investment, he persuaded his mother to co-invest by setting up an SPV (with a valuation of less than $200 million), and as the company's valuation soared past $100 billion, it resulted in thousands of times extraordinary returns. 3. Technology for Good and Ultimate Vision • Frontiers of Life Sciences and Brain-Computer Interfaces (BCI): The ultimate value of AI lies in solving chronic diseases, neurodegenerative disorders, and promoting breakthroughs in cutting-edge biotechnology, liberating humanity from mechanical work and leading to a more creative and healthier future. Video Source: https://www.youtube.com/watch?v=N8CBejLRztg

ABAB AI Insight

This episode is highly valuable. On the surface, it discusses "how Sequoia invests in projects," but it actually covers four larger topics: First, why top VCs must allow a minority to strongly disagree with the majority. Second, why true venture capital alpha comes from "non-consensus + high ownership + long-term holding," rather than collective voting accuracy. Third, AI agents are expanding the internet's "customers" from humans to machines. Fourth, the biggest business opportunity in AI may no longer be selling software, but directly consuming the global service industry and labor budget. If the previous episode on Lobster Capital was about "how micro VCs can compete with YC," this episode looks at the other end: How a top institution prevents itself from becoming mediocre due to its success, size, and history. This is the most worthy aspect of Sequoia to study.
J
Julien Bek
Sequoia
·
17 min read
分享: