Between the Lab and the Cap Table

One fifth of the world's research. Less than a tenth of its deep-tech unicorns. That's Europe's balance sheet, and it has surprisingly little to do with the founders.

The science is there. So is the will to found. What's missing is the machinery in between.

The paradox: world-class research, too few companies

The EU and the UK together author around 20 percent of the world's research publications. Yet only 8 percent of the world's deep-tech unicorns are born in Europe. That's already double the 4 percent of 2021. Still: the science is here, the companies emerge elsewhere.

The capital picture is similar. From 2021 to 2025, US deep-tech companies raised around $389 billion in venture capital, European ones around $94 billion. Roughly four times as much. And in his report on European competitiveness, Mario Draghi notes that in the past fifty years not a single EU company with a market value above €100 billion has been built from scratch.

The misconception: it's the founders. Or the money.

Everyone knows the usual explanation. Europeans are too risk-averse, researchers don't want to found companies, and when they do, the money is missing. Both fall short.

At the start, it isn't the will to found or the capital that's missing. Dealroom's European Deep Tech Report 2025 finds that early-stage funding is ample. The problem sits one step earlier: between the lab where the technology works and the cap table an investor could look at.

Don't get me wrong: at growth stage, Europe really does lack money. 70 percent of large late-stage funding for European deep-tech companies comes from investors outside Europe. In rounds of $15 million and more, only 54 percent of the capital is European, while in the US 80 percent is domestic. Dealroom puts the annual gap at $4 to $24 billion. That's a real problem. But it's a problem for companies that already exist. And there are too few of them.

Where the chain breaks

Here's what that looks like in a single case. An institute at a European university. The sensor prototype has been running stably for months, the patent is filed, a doctoral researcher would like to found a company. The technology transfer office has two people, responsible for a few hundred inventions. The university's first offer: a third of the equity. The negotiation takes almost a year. An investor she meets along the way is friendly: "Come back when you have a company, a CEO and your first customers." In the end she takes a job offer from industry. The prototype keeps running. In the lab.

The numbers show it isn't an isolated case. In 2020 the European Patent Office surveyed universities and research institutes across Europe. Only around a third of their patented inventions are commercially exploited. For a fifth there isn't even a plan. Draghi cites exactly this figure as evidence that Europe's innovation chain breaks at the step from research into commercialisation.

Why, the same survey shows. By their own assessment, technology transfer offices have resources for only 37 percent of their patented inventions. More than half have ten or fewer employees, and in 52 percent, one to three people handle patent commercialisation. That isn't a criticism of those people. There simply aren't enough of them for the job.

Then come the terms. In 2023 the Independent Review of University Spin-out Companies measured for the UK what many founders know from experience. Investors report that large university stakes make spin-outs less attractive. The university's first offer averages 34 percent of the equity, and for 29 percent of founders it's 50 percent or more. Getting a spin-out deal done takes eleven months on average. Only 39 percent of founders consider their deal fair in the end.

The insight: an institutional problem, not a founder problem

There's a lot of work between the lab and the cap table. Assess the technology. Find the first market. Complete the founding team, usually with someone who can sell. Structure the licence, equity and incentives. Build a story an investor understands. That's professional venture-building work, and in Europe it mostly has no home.

Universities have no budget for it. Transfer offices lack the operating muscle. Investors only come in once there's a company. And the doctoral researcher is supposed to learn all of it on the side while she's still writing her dissertation.

The valley between research and market isn't a law of nature. It's simply the place where nobody is in charge.

The concession: things are moving

To be fair, Europe's spin-outs are growing in number. Since 2019, according to Dealroom, 37 percent of new deep-tech startups have come out of universities and research institutes, considerably more than in the decade before. In total, more than 7,300 deep-tech and life-sciences spin-outs have emerged since 1990, together worth around $398 billion.

But a large part of that is concentrated in a handful of universities. Oxford, Cambridge, ETH Zurich, EPFL and TU Munich lead the list. In other words, exactly the institutions that have built their own venture machinery. That doesn't refute the thesis. It proves it. Where the machinery exists, companies emerge. Where it's missing, prototypes stay put.

A guardrail: terms are friction, not a death sentence

Before anyone makes the transfer offices the scapegoat: terms are improving. In the UK, the average university stake fell from around 25 to around 18 percent in ten years, and the leading universities are now often at 5 to 15 percent, on a par with the US. And research cited in the UK review finds only a small effect of stake size on the chance of raising VC.

The real costs are the eleven months, the lost trust and the founder who started somewhere else in the meantime. Fair terms are necessary. But they don't replace the work it takes for an invention to become a company.

What changes: capital follows investable companies

There's a second effect that's often overlooked. The growth-stage capital gap is partly a consequence of the gap before it. Produce too few investable companies and you attract too little capital. And with too little capital you don't build the large funds that can lead Series B rounds.

Sweden shows it can be different. According to McKinsey, 65 percent of startup funding there went into deep tech through the end of 2024, and pension capital is invested in venture capital at twice the European average rate. Capital comes when there's something worth investing in.

The zoom-out: a trillion on the table

In 2025 McKinsey calculated what would happen if the large European economies adopted the proven practices of the most successful deep-tech locations: up to a trillion dollars in additional enterprise value and more than a million new jobs by 2030.

Draghi states the problem without hedging: Europe fails to translate innovation into commercialisation. He gives several reasons, from regulation to fragmentation. Much of that needs Brussels. The machinery between the lab and the cap table doesn't. It can be built, today, one institution at a time.

The principle: institutionalise the work in between

What would a system look like that actually works between the lab and the cap table? It would have four properties.

First: a real building team, not a coach. People who assess technologies, find markets, complete teams and found companies, as a profession, not as a side job.

Second: fair standard terms from the start. A small, clearly defined stake for the institution, a cap table on which the founders remain the largest group, and a deal done in weeks rather than eleven months.

Third: the first cheque is ready. Whoever builds should also be able to invest. Then fundraising doesn't start cold, and the gap between building and financing closes.

Fourth: a network, not an island. No single university has enough industry partners, pilot customers and co-investors. A network across several institutions and regions does.

Coming full circle

The doctoral researcher was no worse a founder than the teams in Oxford, Cambridge or at ETH. Her technology was no weaker. She was simply in a place where nobody was in charge of the path.

Europe isn't short of inventions. It's short of the machinery that turns inventions into companies.

Sources

• U.S. National Science Board / NCSES, "Science & Engineering Indicators 2025: Publication Output by Geography and Scientific Field" (2025): EU-27 17.0%, EU-27 plus UK 20.0% of world publications (2023) — nsf.gov
• McKinsey & Company, "Europe's Deep-Tech Engine Could Spur $1 Trillion in Economic Growth" (2025): 8% of deep-tech unicorns born in Europe (2024, 2021: 4%), Sweden 65%, up to $1 trillion and 1 million jobs by 2030 — mckinsey.com
• Dealroom / Lakestar / Walden Catalyst, "The 2026 European Deep Tech Report" (2026): US vs. Europe 2021–2025 (p. 38), 70% of late-stage capital from outside Europe, 54% vs. 80%, $4–24bn gap (p. 52), 37% spin-outs (p. 72) — dealroom.co
• Dealroom / Lakestar / Walden Catalyst, "The 2025 European Deep Tech Report" (2025): ample early-stage funding — dealroom.co
• Dealroom et al., "European Spinouts Report 2025" (2025): 7,300+ deep-tech and life-sciences spin-outs, $398bn combined value — dealroom.co
• Mario Draghi / European Commission, "The Future of European Competitiveness, Part A" (2024), pp. 6, 29: no EU company above €100bn built from scratch in 50 years, only around a third of university patents exploited — commission.europa.eu
• European Patent Office, "Valorisation of Scientific Results: Patent Commercialisation Scoreboard" (2020): 36% exploited, resources sufficient for 37%, 52% with one to three people on commercialisation — epo.org
• Tracey, Williamson / UK DSIT and HM Treasury, "Independent Review of University Spin-out Companies" (2023), pp. 12–21: university stake 25% → 18%, first offer averaging 34%, eleven months to a deal, 39% consider their deal fair — gov.uk