20VC: Inside Sequoia's Investment Committee: Lessons from Don Valentine, Doug Leone and Alfred Lin | How the SpaceX and Citadel Deals Went Down | What Sequoia Specifically Looks for in Founders with Julien Bek
Julien Bek, a Sequoia partner, discusses how Sequoia identifies and invests in exceptional founders and companies, emphasizing that everyone at the firm is a 'hunter' rather than waiting passively for deals. He explores the shift from co-pilot to autopilot AI applications, the emerging 'services economy' as a trillion-dollar opportunity, and how to read founders through vulnerability, references, and understanding their trajectory rather than pattern matching.
Summary
Julien Bek, a three-year Sequoia partner who previously worked at Excel, provides behind-the-scenes insights into how Sequoia operates as a partnership. He emphasizes that contrary to popular belief, Sequoia doesn't passively wait for companies like Anthropic to call—instead, all partners actively source deals like hunters. He illustrates this with the story of Constantine Buehler's persistent relationship-building with Ken Griffin over years, eventually leading to Citadel Securities investment, and Sean McGuire's conviction on SpaceX, which initially received a '1' vote in the investment committee but became one of the firm's best returns after partners were convinced to visit and see the company themselves.
Bek explains Sequoia's investment philosophy centered on conviction over consensus. The investment committee votes on a scale of 1-10, and sponsors can still greenlight deals that receive low scores if they have sufficient conviction—the best investments historically come from the sponsor with the highest conviction. He discusses the evolution of valuations and market dynamics, noting that billion-dollar valuations now function like Series A rounds used to, with companies potentially growing to $20+ billion. He expresses skepticism about investing in new 'neolabs' (AI companies), comparing them to Quora and StumbleUpon when Facebook emerged—suggesting most will fail unless backed by N-of-1 founders building fundamentally different architectures.
On founder assessment, Bek emphasizes reading people requires vulnerability and curiosity about their personal story, not just their pitch. He begins conversations by sharing his own background—growing up between divorced parents, his mother's cancer diagnosis, his father's neurological condition—to create psychological safety for founders to open up. He discovered one fraudulent founder by asking 'why' multiple times about inconsistencies in their narrative (turning down Stanford acceptance), and recommends checking references, particularly worst references, which reveal character through how founders discuss difficult relationships.
Bek highlights lessons from Sequoia's senior partners: Doug Leone asks about best and worst references (noting how founders' body language changes); Pat Grady frames people as 'vectors' with direction (motivation) and magnitude (ambition); Alfred Lin warns against confusing excellent operators with exceptional founders; and Sean McGuire introduced the ELO framework—that 2400-rated chess players can recognize other outliers in 10 moves while 2000-rated players cannot. He notes country and cultural context matter significantly—German and French customers tend to give ratings one or two points lower than Americans, requiring calibration.
Regarding AI's market implications, Bek argues that 'agents are the new customer' and the economy is splitting into parallel human and agent economies. He predicts the next trillion-dollar company will be 'software masquerading as a service'—selling outcomes rather than tools. He gives the example of Sierra (AI customer support) replacing $50-per-ticket human agents with AI at one-fifth the cost, starting as a copilot but moving to autopilot as trust builds. Today's model is co-pilot (AI assists humans), but as models improve, it shifts to autopilot (AI completes tasks end-to-end), then to outcome-based pricing where AI captures value from results rather than software licenses.
Bek addresses margins and sustainability in the AI era. While initially margins compress, he argues that switching costs remain due to data gravity, enterprise controls, and trust-building—particularly in B2B contexts like Relit (finance automation). He advocates for both infrastructure investments (Fireworks, ClickHouse) and application-layer companies simultaneously, as the human brain struggles with exponentials but both can be massive. He notes that frontier model performance matters for applications where AI hasn't reached human parity (hiring, judgment calls) but matters less for commodity tasks (changing flight bookings).
On founder reads and portfolio decisions, Bek recounts missing Revolut at the seed round despite intense founder signals, only to later invest personally with capital from his mother, who achieved exceptional returns (entering at $180-200M valuation, $100B+ current). He emphasizes that the firm intentionally maintains diverse viewpoints—no single 'Sequoia view' on AI—allowing partners like himself, David Khan, and Pat Grady to publish contrasting perspectives, which he argues creates brand differentiation and attracts founders seeking specific visions.
Finally, Bek expresses excitement about AI reaching 500 IQ levels (versus current 120 IQ), which could unlock breakthroughs in chronic disease treatment and life sciences—making today's concerns about AI displacement seem insignificant by comparison. He views this as the most transformative potential of frontier AI research.
About this episode
<p>Julien Bek is a Partner at Sequoia Capital, one of the most renowned venture firms in the world. At Sequoia, he has partnered with companies including Rillet, Tacto, and Auctor. Before joining Sequoia, Julien spent five years at Accel, where he worked with companies including Miro, Melio, and BeReal. He is also an angel investor in Revolut and Attio.</p> <p>AGENDA:</p> <p>06:35 – What did Julien only discover about Sequoia after joining the firm?<br /> 08:05 – What does everyone get wrong about Sequoia?<br /> 13:00 – Is Series A the hardest stage at which to invest today?<br /> 14:00 – Is Sequoia less focused on ownership as outcomes become larger?<br /> 19:20 – Does Sequoia simply pay more than everyone else to win deals?<br /> 22:00 – Is the "triple, triple, double, double" growth model dead?<br /> 25:00 – What is Sequoia's investment process really like behind the scenes?<br /> 27:15 – Can a partner still invest when the rest of Sequoia votes against them?<br /> 30:00 – Why can a flawless founder pitch actually be a warning sign?<br /> 31:00 – How do you judge whether a founder is exceptional in just 30 minutes?<br /> 33:00 – How can investors tell whether a founder's story is genuine?<br /> 35:10 – Is arrogance a bad trait in a founder?<br /> 36:00 – Which great founder did Julien completely misread?<br /> 37:15 – How should investors adjust their founder assessment across different cultures?<br /> 39:00 – What does a founder's childhood reveal about their future trajectory?<br /> 40:30 – What has Julien learned from Doug Leone, Pat Grady, Alfred Lin and Shaun Maguire?<br /> 47:00 – What does it mean when "agents become the new customer"?<br /> 49:00 – Does UI become irrelevant in an agent-first economy?<br /> 50:15 – Will answer-engine optimisation become larger than SEO?<br /> 52:00 – Will AI agents destroy software margins and brand loyalty?<br /> 53:00 – How quickly will enterprises allow agents to make purchasing decisions?<br /> 54:00 – Is AI infrastructure a safer investment than applications?<br /> 56:00 – Do software margins matter less when the potential outcomes are larger?<br /> 58:00 – Could the next trillion-dollar company masquerade as a services business?<br /> 01:00:00 – Which service industries can AI truly automate end-to-end?<br /> 01:02:00 – Are enterprises simply crying out for more help implementing AI?<br /> 01:03:10 – Will AI lead to dramatically smaller teams?<br /> 01:06:15 – Is traditional private equity screwed?<br /> 01:07:00 – The most overfunded and underfunded categories in venture<br /> 01:08:15 – What is the best AI agent company outside Sequoia's portfolio?<br /> 01:09:00 – Which missed investment still haunts Julien?<br /> 01:09:45 – Which founder trait will Julien never compromise on?<br /> 01:10:00 – Which competing fund makes Sequoia bring its A-game?<br /> 01:10:45 – How Julien missed Revolut—and helped his mother retire by investing anyway<br /> 01:14:00 – What does Julien believe that other Sequoia partners might disagree with?<br /> 01:15:00 – What is Julien most excited about over the next five years of AI?</p>
Key Insights
- Bek argues that Sequoia's competitive advantage stems from all partners being active hunters rather than passive capital allocators, with examples like Constantine Buehler's multi-year relationship-building with Ken Griffin before landing the Citadel investment.
- The investment committee model at Sequoia allows sponsors with highest conviction to override low votes (even a '1' rating) if they believe sufficiently in a company, as demonstrated by Sean McGuire's SpaceX investment that initially received a one-vote but became a top historical return.
- Bek claims that founders who memorize investor expectations and retrofit narratives to match are a red flag, noting that when all IC votes are 7-8, it signals founders have delivered a 'safe' pitch rather than revealing their true character.
- Bek identifies that asking 'why' multiple times about inconsistencies in a founder's narrative—like turning down Stanford—can uncover fraudulent founders, as inconsistencies accelerate their speech tempo and nervous body language.
- According to Bek, checking worst references is more informative than best references because it reveals how founders discuss and learn from difficult relationships and conflicts.
- Bek contends that cultural context significantly affects investor assessment: German and French customers provide ratings 1-2 points lower than Americans on the same experiences, requiring conscious calibration.
- Bek argues that 'agents are the new customer' and will become the dominant traffic source (Cloudflare data: agent traffic now equals human traffic and will reach 1,000x human traffic in five years), creating a parallel economy requiring different business models.
- Bek predicts the next trillion-dollar company will be 'software masquerading as a service,' selling outcomes (e.g., 'closed books' for $15K) rather than tools (QuickBooks at $2K), capturing the $6 service spend rather than competing on the $1 tool spend.
- Bek asserts that early AI applications benefit from frontier model performance when human parity hasn't been reached (hiring, judgment-based decisions) but switching costs from data gravity and enterprise trust prevent price-to-zero collapse in mature applications.
- Bek claims that investing in services-to-software transitions is unlikely to succeed because frontier AI talent refuses to join legacy service businesses, even if they have valuable data, making it better to build application software from first principles.
- Bek recounts missing Revolut at seed despite recognizing founder intensity as obvious, and later investing personally with his mother's capital at $180-200M valuation, which returned $100B+ valuations, illustrating the cost of conviction failures.
- Bek argues that Sequoia deliberately maintains diverse partner viewpoints and public positions (himself on services-as-software, David Khan on $600B questions, Pat on AGI) to remain spiky and attract founders seeking specific investment philosophies rather than bland consensus.
Topics
Transcript
Everyone thinks that we're just waiting for the phone to ring for the next anthropic to call us to invest. It's completely false. Everyone at Sequoia is a hunter. If you look at founders you like versus founders who make money as a two by two matrix, your job is to figure out in which part of the quadrant we make money. The best investments in all the funds are always the companies where the sponsor had the highest conviction. We are only as good as our next investment. That's not an easy jump. If you want an easy jump, you go do something else. Credits to Sean. When he brought in the SpaceX investment, we vote on companies. I…
Full transcript available for MurmurCast members
Sign Up to AccessMore from The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch
20VC: Is Seed Investing Dead Without a $1BN Fund? | Does Ownership and Price Matter When Companies Can Be $1TRN Exits | Are AI Revenue Numbers Real and What to Watch Out For with Venky Ganesan, Menlo Ventures
Venky Ganesan from Menlo Ventures discusses how venture capital has fundamentally changed in the AI era, arguing that seed investing still exists but founders must understand that venture is now focused on outlier companies with explosive growth rather than steady revenue growth. He emphasizes the importance of IRR over ownership percentages, the dangers of gaming metrics, and the need for VCs to play the current market game while maintaining authentic values.
20VC: The Future of Datacentres: What You Need to Know | Why Everyone Gets GPU Depreciation and AI's Energy Costs Wrong | Who Really Makes Money From AI & Why Most Moats Don't Exist with Chase Lochmiller
Chase Lochmiller, CEO of Crusoe Energy, discusses how data centers and AI infrastructure are becoming critical bottlenecks, arguing that energy-distributed computing rather than centralized hubs will power AI's future, and that vertical integration across data centers, GPUs, and inference services creates sustainable competitive advantages in the AI infrastructure space.
20VC: $1BN ARR in 18 Months; The Untold Story of Higgsfield | Spending $4M Per Month on Models | Why Moats in AI are BS | Scaling a Content Team to 150 People with Alex Mashrabov
Alex Mashrabov, founder of Higgsfield, discusses how his AI video generation startup achieved $1 billion ARR in 18 months—faster than Coursera—by focusing on product-market fit, strategic model routing between open-source and proprietary AI models, and building a 400-person team primarily based in Kazakhstan. He shares insights on AI moats being largely meaningless, the $4 million monthly spend on model inference, and his vision for Higgsfield to become bigger than Shopify by owning the distribution layer for AI-generated content.
20VC: Jensen Huang Declares AGI Has Arrived | GPT Astra and Fable 5.1 Accelerate the Model Race | Tesla Launches Cybercabs | Index Pulls Out of Town & Anthropic Pulls From Descartes Acquisition
In this 20VC episode, Rory O'Driscoll and Jason Lemkin discuss major tech news including Jensen Huang's AGI claims, AI agent risks and rule-breaking behavior, Elon's Cybercab launch, venture capital conflicts around AI assistants, and high-growth companies like Wonderful raising at massive valuations with significant secondaries.
20VC: The $100 Billion AI Assistant Race: Town vs Instinct vs GrokBot | We Spend $75K Per Engineer on AI Tools | Why the AI Assistant Market Is Not a Bubble & AI Assistants Will Replace Every App on Your Phone with JD, Founder of Town
JD, founder of Town, discusses how his AI assistant platform operates within email and calendar to automate work tasks, argues that network effects at the agent level will be the true moat in the AI assistant market, and explains why the speed of market development means companies can only learn at human speed while building at machine speed.