Analysis of 408 European unicorns and 2,053 associated investors
Bondo Foundation's analysis of every European unicorn founded between 1970 and 2021 produced a finding that is uncomfortable for the venture-capital industry and consequential for everyone else.
Seventy-two percent of investors who participated in funding a European unicorn participated in funding exactly one. Fourteen percent funded two. Only fourteen percent funded three or more. The most prolific investors in the dataset — those who participated in funding twenty or more — number fewer than ten globally, and almost all of them are large US firms.
Investors who have been making investments for thirty years are no better than investors who have been making them for five at picking the firms that will become unicorns. The size of an investor's fund is uncorrelated with the rate at which the fund's investments reach unicorn status. The stage at which an investor invests is uncorrelated with the probability of unicorn outcome.
Experience does not change this. Fund size does not change this. The selection function of the venture-capital industry is, at the population level, much closer to random than the industry's self-presentation admits.
The most natural interpretation of these patterns — and the one the broader academic literature increasingly supports — is that selection in venture capital is mostly a property of luck rather than skill. What looks like skill, on closer examination, often turns out to be the consequence of initial success: an investor who happens to back a successful firm early in their career acquires reputation that brings better deal flow subsequently, generating a self-reinforcing pattern that resembles skill without straightforwardly being it.
This finding matters because the venture-capital industry's value proposition is built on the claim that it selects well. The fees the industry charges, the carry it takes, the deference it expects from its limited partners all rest on the assumption that the partner's selection judgement is worth paying for. If selection is mostly random, the industry is mostly selling reputation rather than analysis, and the resources it currently spends on selection would produce more value if they were spent on something else.
That "something else" — supporting the ventures that have already been funded, building the infrastructure that connects funded ventures to the resources they need, ensuring that the most promising ventures get the support they require to reach scale — is the work that the venture-capital industry, structurally, is not configured to do.
Building the infrastructure that does it is the work of the Innovation Operating System.
The implications extend beyond the industry itself. If investors are selecting randomly from the visible tip of the innovation ecosystem, then the entire upstream system — universities, research institutes, pre-incorporation teams — is also being served randomly. The 90% of the ecosystem that lies below the waterline is not just unseen; it is unserved.
The result is a systematic misallocation of capital at European scale, compounding year after year. Fixing the selection infrastructure does not require changing how investors think. It requires giving them access to a dramatically larger and more representative pool of opportunities — which is precisely what bondo.ai is designed to provide.
One unicorn is worth ten to thirty unicorns — not in revenue terms, but in spillover effects. The alumni networks, the talent attraction, the capital recycling, the technology diffusion, the ecosystem credentialling that follow a successful unicorn produce ten to thirty times the direct economic value of the unicorn itself.
Europe produced approximately 400 unicorns to 2021; the United States produced approximately 1,600 in the same period. The direct value gap is large. The multiplier gap — the cumulative value of all the alumni-founded ventures, the inventor attraction, the capital generation, the technology diffusion — is much larger.
The unicorn measure is not the only measure of innovation output. Europe has built different kinds of strength in different sectors. But the unicorn measure tracks something specific and important: the capacity to produce new firms at scale rapidly, in response to emerging technological transitions. Europe is structurally weak on this dimension. The cost of the weakness compounds through multiplier effects over decades.