Jeremy Giffon on rescuing 'trapped' founders buried under venture pref stacks

Summary

Jeremy Giffon runs a fund built around a specific failure mode in venture: founders who have built real businesses — $20–30M in revenue, solid teams, customers who love the product — but are buried under pref stacks so large that the equity economics have collapsed. A founder with 8% of a company sitting under a $200M liquidation preference, as Giffon describes it, has built something heroic and wound up with almost nothing to show for it. His fund buys into these situations, restructures the cap table, and tries to hand founders either renewed control or a clean exit.

The supply of deals is growing. Venture dollars deployed over the past decade have roughly 10x'd, and the middle of most portfolios is generating almost no DPI. Growth funds that wrote checks at peak 2021 valuations are quietly looking for any path to capital return. That pressure is creating more motivated sellers and more founders willing to have conversations they were previously too stuck to start.

AI is changing the calculus for some of those founders before they call Giffon. Companies built four or five years ago are realizing they can replace meaningful headcount cost with AI tools, grow 100% a year, and do it without raising another round. Giffon argues this is the beginning of a bootstrapper renaissance — capital-efficient businesses that don't sacrifice growth but do let founders own more of what they build. The alternative, slapping an AI label on the deck and going back to market for two more million, he thinks is a mistake.

Deal anatomy

The tension in any restructuring is that investors at different points in the cap table have completely different incentives. Early seed investors are often happy to roll over — no reason to sell, no fund pressure. Late-stage growth investors who wrote checks at inflated valuations just want capital back, even at cost. Getting everyone to the same table means matching each party's actual cost of capital to an outcome they'll accept, rather than pretending everyone shares the same upside horizon.

Giffon's observation on ownership is direct: a VC fund starts as an ideal owner, becomes a neutral one as the fund ages, and can eventually become detrimental when incentives diverge badly enough. The point at which that happens depends on the fund's model — a firm that needs power-law outcomes to justify zeros elsewhere demands something very different from a business than an investor with lower return thresholds.

Operator talent

When a founder is trapped and would rather leave than abandon what they've built, Giffon argues the market for replacement operators is deeper than most people think. His framework, citing Andrew Wilkinson, points to senior executives at companies two or three times the size of the business in question — people with real capability and some entrepreneurial drive, but not the risk appetite or circumstance for a zero-to-one founding role. Taking a $20M-revenue business to $100M requires a different profile than building from scratch, and that talent pool is largely untapped.

Venture-to-PE drift

The big venture firms are already functionally converging with asset management. Andreessen Horowitz and General Catalyst are the clearest examples — GC buying a hospital network is straightforwardly PE behavior. Giffon's read on which firms go fully public or stay private maps to a simple distinction: firms started by founders tend to build toward scale, because that's what founders do. Marc Andreessen and Ben Horowitz are founders. David Falco at GC is an entrepreneur. The institutional ambition follows from that.

On the wave of venture-backed rollups — firms raising venture dollars to pursue PE-style acquisition strategies — Giffon's view is that the economics are genuinely attractive right now. Raising $20M at 20% dilution with no hurdle rate beats the standard 2-and-20 PE structure on paper. But he expects that window to close, and suspects the MBA pile-in is a timing signal that it already might be.

AI startup durability

Asked about the cohort of AI companies hitting revenue milestones at unprecedented speed, Giffon is skeptical that fast growth signals durable businesses. His framing is that anything built very quickly can be undone just as fast. Switching costs are low, the leading model changes week to week, and consumer intent is to buy "an AI version of X" rather than any specific product. He draws the internet parallel: Google was the last search engine, not the first. Most of the companies rushing to capture AI demand now won't be the ones that survive to the equilibrium. He pays for six AI model subscriptions today and expects to pay for one in three years.

Firm building

Giffon is making one hire — an investment principal, first on the team. His description of the role is direct: more deal flow than he can handle, a small team intentionally, significant economics to share, and day-to-day work alongside what he calls "30 AI agents allocating capital."