Sequoia Just Made the Case for Solo Founders

What this article answers (plain language): A Sequoia partner argues that we consistently underestimate how big outcomes get and how long they take. If he's right, the implications for early-stage AI startups are profound — and they point toward solo founders, not bigger teams.

Alfred Lin, partner at Sequoia, published a piece this week called "Size of the Prize & Run the Distance." His framework reduces investing to four variables: the size of the prize, the probability of success, the investment cost, and the time to exit. His thesis is that we're decent at estimating cost, overconfident about probability, and consistently, dramatically wrong about two things — how long outcomes take and how big they get.

His proof point is Nvidia: three decades to reach $1 trillion in market cap, then barely two years to go from $1 trillion to $5 trillion. Compounding looks like nothing for a long time, then looks like everything all at once.

He's right. But I don't think he's followed his own logic to its most interesting conclusion.

What Happens When You Apply This to Early-Stage?

Alfred's framework is aimed at growth-stage and public-market investing — the kind of bets Sequoia makes at scale. But the variables he identifies don't stop being true at the seed stage. They just point in a direction that most VCs aren't comfortable with.

If the prize is bigger than we think, then TAM estimates based on existing market spend are structurally too small. Standard venture analysis sizes markets by looking at what customers currently spend on adjacent solutions. But AI-native companies aren't optimizing existing categories — they're creating new ones. The TAM for autonomous knowledge work isn't a slice of Asana's market or Monday.com's revenue. It's closer to the TAM for smartphones in 2005: technically zero if you only count existing sales, enormous if you count what it actually replaces and creates.

If outcomes take longer than we think, then the founder who survives longest wins. This is where the math gets uncomfortable for traditional venture. A solo founder burning $5,000 a month has a decade of runway on $600,000. A 30-person startup burning $400,000 a month has 18 months on the same raise. Alfred's own compounding thesis says the magic happens at year 10-15. Who's still alive at year 10?

The Metrics That Break

Alfred's framework doesn't just challenge how we size markets. It breaks several metrics that early-stage VCs rely on daily.

Burn rate as signal. VCs pattern-match low burn to low ambition. "They're only spending $10K a month — are they serious?" But if the prize requires patience and the timeline is longer than anyone expects, burn rate isn't a measure of ambition. It's a measure of survival probability. High burn with venture funding is a bet that you can compress time. Low burn with AI leverage is a bet that you can outlast time. Alfred's framework says time is the variable we get most wrong. Bet accordingly.

Team size as execution proxy. Headcount has been the default heuristic for velocity in venture: "they've hired 20 engineers, they're executing." But AI collapses the ratio between people and output. A solo builder shipping with AI tools might push more product per week than a 10-person team coordinating across Slack channels. The question shouldn't be "how many people" but "what's your output velocity per dollar of payroll?" A two-person company doing $1M in ARR is a fundamentally different asset than a 40-person company doing $1M in ARR — same topline, wildly different margin structure and compounding potential.

Revenue multiples that ignore AI leverage. Traditional valuation treats two companies at $1M ARR roughly the same. But the AI-leveraged company can scale revenue without proportionally scaling headcount. That's not a marginal efficiency gain — it's a structural change in how value compounds. If Alfred is right that compounding is the force everyone underestimates, then the company with the highest compounding potential per dollar invested should command a premium. Right now, the market gives that premium to the company with the biggest team.

Time-to-moat vs. time-to-revenue. VCs optimize for time-to-revenue because fund structures demand it — 10-year funds, J-curve pressure, LP expectations. But for AI products that accumulate context and improve with use, the real value metric is time-to-moat. How long until switching costs make the product irreplaceable? Revenue follows moat, not the other way around. Alfred's own argument supports this: he says we underestimate both how long things take and how big they get. So optimize for the thing that compounds — the moat — not the thing that's easy to measure early.

The Dilution Math Flips

Here's where it gets concrete. If AI lets a founder reach the same product milestones that used to require $5M in hiring with $500K in tools and compute, the cap table math changes entirely.

A founder who raises $500K at a $5M valuation gives up 10%. A founder who raises $5M at a $20M valuation gives up 25%. The first founder owns more of a company that, by Alfred's compounding logic, might reach the same terminal value — just on a longer timeline with dramatically better capital efficiency.

But VC deal structures haven't adjusted. A $500K pre-seed still often takes 10-15% equity, same as it did when that capital needed to cover 5 salaries. The capital efficiency of AI-native startups means founders who need less money should have more leverage than the market currently gives them.

The Irony

Alfred Lin's piece is a love letter to patience, compounding, and scale of ambition. Run the distance. The prize is bigger than you think. The first trillion is the hardest.

But the structural implication of his own framework — applied to the current moment where AI is collapsing the cost of execution — is a case for exactly the kind of company that most VCs pass on: the solo founder, the $500K round, the company that looks small for years while quietly compounding capability and context.

The venture industry prices for speed. Alfred's framework argues for duration. AI makes duration survivable. That's the gap in the market that nobody's pricing correctly.

If we actually believe compounding works — if Nvidia going from $1T to $5T in two years after three decades of building isn't a fluke but a pattern — then the question every early-stage investor should ask isn't "how fast can you grow?" It's "how long can you compound?"

The solo founder with AI and a decade of runway might be the highest-expected-value bet in venture right now. The math just hasn't caught up yet.