Is small better?
#14 | On everything I know that I don't know about where VC is headed
VC is a young industry, and has already changed a lot since the time of the Traitorous Eight. Why are we assuming the next decade will be like the last one?
My buddy Chris edited this edition of Unbeaten Path listening to the thumping drive of Alex Farell’s “American Nightmare” (it’s not that scary)
The venture capital industry is consolidating.
The dichotomy between the approach of small specialist firms and humongous generalist ones has been generously discussed1. Right now, it seems the market prefers the big “agglomerators” if the numbers are to be believed. According to the Financial Times, “more than half of the $71 billion raised by US VCs in 2024 was pulled in by just nine firms”.
Some might say the past few years should be considered a transition period because the industry was correcting the excesses of the ZIRP era and, after all, nobody ever got fired for investing in a16z.
Whatever the reason, I think it’s interesting to explore the consequences.
The first problem is in investment strategy.
Lack of liquidity is one of the main problems VC firms have faced in recent years. The IPO window has been closed for some time, short-sighted regulators have killed the M&A market, and secondary buyers know LPs are hungry for money and thus push for big discounts on assets’ prices.
Big capital agglomerators can live with a longer time to exit because their performance targets are lower than small specialist funds. They have an abundance of capital which can continue to be deployed in IPO-sized Series J rounds.
How should “normal” funds, i.e. those that can’t raise $5B, think about their investment strategy? Can we still build a traditional $200M, 2-and-20 fund, with 50% of capital for follow-on investments and underwriting $2B exits in a 10-year time horizon?
Using Fred Wilson’s VC-Poker analogy, how much cash should funds reserve to keep “seeing more cards” when the winners can raise an infinite number of rounds and dilute you to hell?
I think now more than ever that aligning with the consensus and being right will give mediocre outcomes. Consensus deals are expensive from the start, and expensive deals paired with a long time to liquidity destroy every VC fund model.
The second problem is the investing opportunities.
The SaaS era was very profitable for VC because SaaS was characterized by zero-marginal costs, allowing scaling with limited resources possible. SW margins were great, and the traditional wisdom says that $100 million in revenue was enough to reach a $1 billion valuation.
AI is changing everything.
The part of this tweet we should focus on is: “people use it much more than we expected.”
This would be a SaaS founder’s dream! User Retention, Daily Active Users, Stickiness, where do I sign? Losing money out of a subscription because your users use the service too much is something you would expect from a All You Can Eat sushi restaurant, not from a SW company.
Truth is that AI, especially the new “reasoning” models, doesn’t have zero-marginal costs and requires huge upfront investments in CAPEX. This to us is very similar to something we know well: investing in HW!
Investing in hardware is very different than investing in software: it requires a different commitment, a different fund structure, different specializations, different timelines, and different reasons driving acquisitions.
Maybe the shift to a world in which only giga-funds and micro-funds exist is more structural than people think. Maybe you can either be small and very contrarian hoping you are right (”investing at the edge of large markets”), or invest in what’s hot now knowing you have so much firepower that you can set the rules of the game. Maybe in this scenario, being somewhere in between doesn’t work anymore. Maybe small is better.
I don’t have answers to these questions as of now, but I believe this is something people working in VC should think of.