Sam Lessin: the VC factory model is dead and zombie unicorns are haunting the ecosystem

Summary

Sam Lessin, partner at Slow Ventures, argues that the venture capital factory model is structurally finished — and that what has replaced it is not a reformed version of the same system, but several disconnected capital markets running in parallel with incompatible valuation logic.

The factory is broken

For roughly a decade, early-stage VC operated as a conveyor belt: seed to Series A to growth, with standardized metrics at each gate ($1M ARR earns a Series A at a predictable multiple, triple-triple-double on the way to growth), and the whole thing terminating in a public market IPO where retail buyers absorbed the offering. That system depended on public markets hungry for high-growth subscale companies. That hunger is gone. Institutional and retail capital has concluded it can just buy more Amazon or Meta — companies with no effective ceiling — rather than take risk on a $3B newcomer with thinner margins and less liquidity.

The result is a cohort Lessin calls zombie unicorns: companies that cut burn, got marginally profitable, and can sustain themselves indefinitely — but cannot go public, attract a buyer, or return capital to investors. The liquidation preference stacks make M&A unattractive. There is no functioning off-ramp. They simply exist.

Fragmented markets, not one system

Lessin's 2025 update to his October 2023 thesis is that venture capital is no longer a single integrated capital system. It has fractured into distinct ecosystems — public markets, late-stage private markets, and early-stage markets — each with its own valuation logic, each largely non-interoperable. Large LPs invest across all three without forcing convergence. The problem is that founders and early-stage funds still behave as if the factory is intact.

At the growth stage, the question "who is the marginal buyer?" has no clean answer. Private-to-private transactions will become more common — funds selling positions to other funds, as happens in private equity — but the buying fund needs its own exit thesis, which compresses prices well below what a DCF-to-public-market framework would imply.

At the early stage, valuations are still clearing around belief rather than metrics. But the path from pre-seed to a legitimate Series B and beyond now requires a credible "infinity thesis" — a narrative that the company could be boundlessly large — rather than hitting milestone numbers. Downside protection plus an option on infinity is how late-stage buyers are pricing things. The average of infinity and zero is infinity, which is how inflated valuations persist even without rational DCF support.

Large funds, small funds, misaligned incentives

A $50B venture fund is not optimizing for DPI. It is optimizing for fees and AUM growth, because that is how public markets value asset managers when they eventually list. Making 3x on $50B is close to impossible; the business model is deploying gross dollars and charging 2% on them. Early-stage funds running $30–50M have the opposite problem: they must actually return capital or they stop getting re-upped. Marks are not money.

The distortion Lessin documented in 2023 — late-stage funds sending junior staff to write seed checks simply because they had nothing better to deploy into — has mostly corrected. AI has given large funds somewhere to put billions, so they have retreated from early-stage markets. But the damage to seed pricing discipline during that period was real.

The recursive version of the same problem exists at every level: Slow Ventures writes $100K angel checks off a single meeting because the amount is immaterial to the fund, which distorts pricing for angels who actually need that round to pencil.

AI acquisitions and the Windsurf question

On the reported OpenAI acquisition of Windsurf for $3B, Lessin is skeptical of a broader M&A wave. When you acquire an AI company, you are buying talent, technology, or distribution. Talent acquisitions can happen — he acknowledges some individuals will command $200M packages — but it is not a repeatable investment thesis. Technology is increasingly commoditized. Distribution is the most defensible rationale, but only if the target has genuinely locked-in contracts or a structural tailwind.

Large incumbents, he argues, have enough capital and engineering depth to build rather than buy almost anything they need. What happened in the 2010s acqui-hire era — Lessin's own first company was acquired by Facebook — probably was not a good use of capital in most cases, and the structural conditions for repeating it are weaker now. Any deals that do happen will be highly bespoke, not a market-wide trend.

Marketing stunts and early traction

On companies like Artisan announcing a $25M Series A via a billboard, and Roy Lee's viral "cheat on everything" launch for Cluely, Lessin is skeptical but nuanced. His prior on companies that launch with marketing stunts and become important businesses is approximately zero. The structural risk is that a broad, undifferentiated early audience locks your brand into a product definition you may abandon in two months — exactly the challenge facing founders who go viral before product-market fit.

His preferred model is high friction early: customers who show up despite barriers are true believers who tolerate rough edges and become advocates. Too much attention too quickly from a non-fervent audience creates noise, distraction, and churn. His example is Jelly, where a memecoin blew up prematurely, most users bounced immediately, but a kernel of committed early users remained — and that kernel is the only part worth building on.

Tesla as an infinity story

On Tesla, Lessin is blunt: it is a meme stock, and has been for a long time. The DCF value is not a fraction of the trading price. What Elon Musk has built is the purest expression of the infinity thesis — a story large enough that retail investors treat it as a call option on an unbounded future. The fact that Tesla's stock moved 7.5% on a single statement about Musk stepping back from DOGE, against a backdrop of declining profits and revenue, is, Lessin says, the market correctly pricing what his attention is actually worth: the $56M compensation differential is almost defensible on that math. Musk is the best marketer, capital raiser, and storyteller of his generation. The question is just whether the underlying business ever catches the narrative — and on a DCF basis, it has not come close.