A Cherry Ventures report on company formation
Heartbeat is our view on how Europe's inception-stage companies are forming. 10,000 companies tracked, 89 cities, 10,000 founders, €5B+ raised. Across all industries, bootstrapped and VC funded.
Almost everything written about European startups covers the same slice: companies that already raised a round, or showed up in some database. The rest are starting up in the same cities, in the same categories, at the same time, without passing any checkpoint that would record them. Nobody has been counting them.
Starting a company in Europe has got cheaper, so companies are starting smaller. 75% begin with a single founder, rising to 79% among those nobody has funded yet. They are forming almost entirely outside the machinery built to find them. Only 7% of these founders have been anywhere near an accelerator, and where a second founder does appear, the two met at work rather than at a university or a programme designed to introduce them.
Drawn from Cherry Ventures' proprietary dealflow, cross-checked against public datasets
01 / What we found
Our data suggests that today's European founder is likely male, starts by themselves and holds a Master's; in deeptech, a doctorate.
79%vs46%
79% of companies that have not yet raised institutional money have a single founder. Among those that have completed a first round, it is 46%. The "minimum viable company" has shrunk. Most of the institutions around them still assume a team of three.
4.3×
Master's degrees appear 4.3 times as often among founders as MBAs; in deeptech, PhDs outnumber MBAs almost ten to one.
3.2×
Co-founders are 3.2 times more likely to share a former employer than a university among the companies where we could check.
10.1%
Women account for 10.1% of founders across companies founded in this batch.
02 / Formation
7%
Any accelerator affiliation
93%
No accelerator affiliation
A company settles most of what matters in its first year. Who runs it, what it sells, who it hires, whether it stays in Europe at all. Almost none of that is public while it is happening.
Usually it becomes public in one go. A team joins an accelerator, turns up on a cohort list, does a demo day, and is suddenly legible to everybody. That is where coverage of European startups tends to start.
Of the 10,000 founders in this sample, 657 have been through an accelerator of any kind - 7%. Y Combinator accounts for 185 of them, Antler for 58.
The other 93% did it without one. We went looking for them.
n = 10,000 founders
03 / Where Europe builds
Themes for new companies are consistent across Europe. A map of raw counts mostly tells you where the people are. So we indexed each city against the average instead.
Munich forms nearly twice as much climate as a typical European city. Zurich, more than twice as much deeptech. Frankfurt leans fintech, Tallinn infrastructure.
We left London out of the average. One in five European companies forms there, and including it would have pulled the typical city towards London.
The map shows funded and unfunded companies together. Click a city to separate them.
Volume follows population, so the biggest cities produce the most companies and that tells you very little. Measured against the number of people who live there, the map rearranges. Amsterdam and Zurich lead Europe at about 34 founders per 100,000 residents, roughly twice London's rate and three times Berlin's. Paris, which looks densest on a map because its city limits are unusually tight, sits sixth.
Founders in the sample per 100,000 residents of the administrative city. Population figures are standard published administrative values.
04 / What Europe builds
Select a category to read its detail.
AI-first software is the largest category at 22.5%, with conventional SaaS behind it at 20.9%. The two build for the same jobs. Sales, hiring, legal work and analytics come up in both, because these are the things every company needs and the things software has always been sold for. What separates them is how the product is made, and who makes it. Sales and marketing software is the biggest group inside both. Whichever tools founders reach for, the first thing they build is a way to sell something.
749 companies here are building something that exists in the world: robots, materials, machines, power hardware, spacecraft, medical devices. That is 7.5% of everything formed.
Robotics is the largest group at 189 companies, followed closely by advanced materials at 181 and manufacturing at 117. Power hardware and space come next.
Munich and Zurich are where it concentrates. Munich forms this kind of company at twice the European rate, Zurich at more than one and a half times, and both sit next to a university that has been turning research into hardware for decades. London, which has the most of everything else, is below the European average here - 103 companies, but only 5.5% of what it forms.
They start in teams more often than the rest of the sample, 70% solo against 75%. Their founders hold doctorates at nearly ten times the rate they hold MBAs, the widest gap of any group in the report. And 17% have raised from a fund, against 12% across the sample - the money finds them, because building a machine is not something one person does over a weekend.
05 / How Europe builds
One mark represents one company in a hundred.
This is the beginning, not the end state. Some of these founders will add a co-founder within six months, some never will, and a few already have one that nothing has recorded yet. What the number describes is how companies in this sample look at the point they are formed - which is the only point at which anybody is counting them this way.
Cheaper tools and AI-assisted building have moved the single founder from a stage a company passes through to, for much of this sample, the company itself.
Co-founders are 3.2 times more likely to share a former employer than a university.
That puts the pipeline somewhere nobody has thought to look. Campus programmes and accelerator matching schemes exist to introduce strangers to one another, and the founders here mostly arrived already introduced. Europe’s large employers are running its most effective co-founder matching scheme entirely by accident.
McKinsey has produced more founding pairs than any other institution in Europe. Meta, Google, Apple and Mastercard follow, along with Bain.
There is a difference in how the two sides behave. Work produces far more pairs in total, but no single employer dominates: the four biggest account for just 7% of all the pairs formed at work. Universities produce fewer pairs and concentrate them heavily, with Cambridge, Munich, Zurich and Imperial alone accounting for a quarter. If you meet your co-founder at university, it is likely to be at one of a handful of places. If you meet them at work, it could be almost anywhere.
What one person can finish has changed, and the size of a company at the start has changed with it. Teams still form where physics, a laboratory or a clinical licence insists on one.
06 / Who builds
Founders here have less in common than the map suggests. Whether you build alone, and whether you get to build at all, turns out to depend on what you are building.
The more specialised the science, the more likely a company starts with a team - solo founding peaks where one strong engineer can carry the product, and drops off where a lab bench or a clinical licence is required.
AI applications, the largest category here, has the lowest share of women of any large category at 7.5%. Health, the slowest to form and the hardest to enter without a qualification, has the highest at 16.6%.
Where a credential is the gate, more women get through. Where a network is the gate, fewer do.
A note on the figure: this counts people. The 13% usually quoted is the share of European venture funding going to companies with a female founder, which is a different measurement.
Put the two together and the pattern is not really about technology. Infra is 80% solo and has fewer doctorates than MBAs. Deeptech is 67% solo and has almost ten doctorates for every MBA. Across the categories, the more a field requires a qualification to enter, the more likely its companies start with more than one person.
Teams have not disappeared. They have retreated to the places where you still need a credential to begin.
SF–EU diaspora
Talent leaves Europe and talent arrives. Both legs are visible in the same sample, and they are not the same shape.
Out: of the San Francisco founders here who carry a university record, 72% were educated in Europe - 71 of 98. Imperial College London leads, then ETH Zurich, Cambridge, HEC Paris and Oxford. These are companies registered in San Francisco, founded by people Europe trained.
They are also, to a striking degree, one cohort. 56% of them have raised, against 11% of companies in Europe. But 29 of the 44 funded companies raised exactly the same amount - $126,225, the standard accelerator cheque, appearing twenty-nine times. Seventeen of the sixty-two founders with a named school list Y Combinator as their education. Almost all of them build software: AI-first, infrastructure or fintech, with barely any health, climate or consumer.
Back: 86 founders now building in Europe were educated in the United States. Stanford, Berkeley and MIT lead, then Columbia and Carnegie Mellon. A third landed in London. They raise a median first cheque of $3.25m, against $1.63m for Europe-educated founders, and they found in larger teams.
They also bring the team with them. Founders educated in the United States start companies with 2.6 people on average, against 1.78 for those educated in Europe, and they found alone 41% of the time against 50%. The American playbook that arrives in Europe is not only about capital. It is the multi-person company, at the moment the rest of the sample has stopped building them.
The reading: Europe exports through one door and imports through many. The outbound leg is narrow - a single programme, a single cheque size, three categories of software. The inbound leg is broader, better capitalised and spread across the continent. Seven per cent of Europe's founders have been near an accelerator. Among those who left, it is most of them.
The employer has overtaken the university as the place founders are made, and the pipeline does not stop at Europe's border. This sample shows Europe training founders who leave, and taking in founders trained in America.
City index
How we built these lists
As early-stage VCs, most companies we look at are early, some pre-inception. To properly capture the formation layer, we looked back no more than 18 months from a company's founding, and no more than one institutional round raised.
We believe getting to the core truth of a company radiates from the founding team, and that those signs show up long before a company has its breakout success. So we got on the ground, city by city, talking directly to the ecosystem leaders, university professors, angels, operators and micro VCs who see these companies before anyone else does. The names that kept coming up independently, before they showed up anywhere public, were what we went after.
Every one of those names, alongside thousands of other companies we track a year, went through the same internal scoring model: the scale of impact the company could have globally, how genuinely AI-native the product is, the strength of the thesis, and the team, their talent density, where they've built before, how deep that experience really runs. We ended up with ten companies per city that we believe combine all of the above, while capturing the city's own heartbeat and the themes defining it.
