What is a startup? Stages, funding and how European tech gets built
A startup is a company searching for a repeatable, scalable business model under extreme uncertainty. The stages, the funding, the real survival odds - and what all of it looks like in Europe in 2026.
Key takeaways
- A startup is defined by its search for a repeatable, scalable model – not by age, size or sector.
- European funding runs a five-rung ladder; across the rounds we track the medians are €1.6M pre-seed, €5M seed and €16.5M Series A.
- Expect 15-25% dilution per round plus an option pool: founders typically hold 35-45% after three rounds.
- Raising takes 3-6 months in Europe – start with nine months of runway, not three.
- Accelerators typically take 5-8% for a small cheque – but many European public and corporate programmes take no equity at all.
- “90% of startups fail” is folklore. Dealroom’s cohort of 3,075 European seed companies is the honest number: 31% reach Series A within four years.
What is a startup?
A startup is a young company built to grow fast – designed around a repeatable, scalable business model rather than steady local trade. Speed of learning and access to venture capital, not age or size, are what set it apart from a small business.
In practice, the European ecosystem treats “startup” as a phase, not an identity: a company stops being one when growth stabilises, it gets acquired, or it lists publicly. Most of the companies in the Sesamers startup directory are somewhere between their first cheque and a Series B.
A useful mental model: a startup is a series of experiments funded in stages. Each funding round buys the runway to prove the next assumption – and each proof re-prices the company. Europe raised roughly $44bn in 2025, its best year since 2021-22, and €44.1bn across 1,740 deals in H1 2026 – fewer, larger rounds than two years ago (Atomico; Tech.eu).
The startup lifecycle, stage by stage
European rounds follow a broadly standard ladder. What each rung buys – and what it costs in ownership – matters more than the label.
The ladder isn’t law. Angel syndicates now write institutional-size cheques, and outliers break the scale entirely – that’s what makes tracking the medians below matter. Median round sizes rose 32% from 2024 to 2025 (Bloomberg), which is why a 2023 benchmark will mislead you today.
What a round actually costs: dilution, worked through
Every founder learns this arithmetic eventually. Here it is up front, run at typical medians for a two-founder software company that creates a 10% option pool at seed and tops it up at Series A.
| Stage | Raised @ post-money | New investors | Founders after | Founders’ stake value |
|---|---|---|---|---|
| Incorporationtwo founders, 50/50 | – | – | 100% | €0 |
| Pre-seedSAFE priced at the round | €0.8M @ €4M | 20% | 80% | €3.2M |
| Seed+10% ESOP created pre-money | €5M @ €20M | 25% + 10% pool | 52% | €10.4M |
| Series Apool topped back up | €16.5M @ €66M | 25% + top-up | ≈38% | €24.8M |
The point of the last column: ownership falls every round while the value of what’s owned rises roughly 8× across the three. Dilution is only a bad deal when the money doesn’t buy proof.
Option pools are almost always created pre-money – the incoming investor’s percentage is protected, so the 10% dilutes existing holders alone.
A 1× participating preference means investors take their money back and their percentage. On a modest exit, the founder line can be far below 38%.
Stacked SAFEs with different caps can convert to more than the headline percentage. Model the conversion before signing the priced round, not after.
Worked illustration at typical medians – a teaching example, not a benchmark. ESOP norms from Index Ventures’ Rewarding Talent: ~10% at seed in both the US and Europe, ~15% at Series A in the US, flatter in Europe.
What investors actually check at each stage
The burden of proof moves up the stack as you climb. Pitching the wrong proof for your rung is the single most common reason a good company gets a pass.
Why you, why now, why this problem. There’s no data to argue with, so investors underwrite founder-market fit, speed of learning, and whether you can recruit people better than you.
Not “we have users” – which users, how they found you, what they’d do without you. Retention curves that flatten, and a wedge that a competitor can’t copy in a quarter.
Can you spend €1 and reliably get more back? CAC and payback by channel, pipeline coverage, a sales motion that works without a founder in the room.
Is the market big enough to justify the price, and can the team hold together at 3× headcount? Diligence shifts from the model to management.
How many make it?
Of every 100 European startups that raise a pre-seed round, roughly:
The cleanest public cohort: Dealroom followed 3,075 European seed-funded companies from 2016-18 and found 6% reached Series A within 12 months, 18% by 24, 27% by 36 and 31% by 48. It’s a dated cohort in a different rate environment, but it’s real data rather than the folklore below.
Five startup myths that will not die
Nobody can source it. The nearest real numbers are US BLS all-business survival (20% gone by year two, 45% by five, 65% by ten – every business, not startups) and the Dealroom cohort above. Most startups don’t explode; they get acquired quietly or stop growing.
Venture is one financing model for one kind of company – the kind that must grow fast to win. Grants, revenue and venture debt ($5.6bn in Europe in 2025, 12.7% of all funds raised) fund plenty of good businesses that would be ruined by a VC clock.
First movers educate the market and eat the mistakes. In most European categories the winner is the third entrant with better distribution – see the fintech cohort of 2015 versus who’s actually profitable now.
A $1bn paper valuation is a price, not an outcome. A €40m exit with three founders holding 45% beats a €1bn round with a 3× liquidation preference stack, every time.
Ideas are abundant and mostly wrong at the start. What’s scarce is a team that changes its mind fast when the market disagrees – which is exactly what pre-seed investors are underwriting.
The first 18 months: what actually happens
A modal European software startup, month by month. Deep tech runs on a longer clock – grant cycles and lab work push each phase out by six to twelve months.
Entity, founder vesting (4 years, 1-year cliff is standard in Europe), IP assignment. Do the equity split before there's anything to argue about.
One narrow thing ten people need. National grants and EIC-style instruments run in parallel from here: they take months, so apply before you need the money.
Paying users or signed LOIs. Angels and pre-seed funds move on evidence plus team. Expect 3-6 months from first pitch to money in the bank – start when you have nine months of runway, not three.
The pool created at pre-seed gets used here. First hires are generalists who close their own loops; the specialist org comes after Series A.
Either the retention curve flattens and you raise a seed, or it doesn't and you change the wedge. Both are normal. The failure mode is raising to avoid the decision.
Incubators, accelerators and startup programmes
Somewhere in those first 18 months most founders consider a programme. The three formats get used interchangeably and shouldn’t be – they differ in what they take from you more than in what they teach.
| Format | Duration | Money in | Equity taken | Best for |
|---|---|---|---|---|
| Incubator | 6-24 months, open-ended | Usually none; desk space, labs, mentors | Often 0% | Pre-company and deep tech – you need time, equipment and a first network more than cash. |
| Accelerator | 3-6 months, fixed cohort | €20k-€150k typical | 5-8% (private) | A built product looking for its first customers and a demo-day investor room. |
| Startup studio | From day zero, ongoing | Full build cost + team | 20-50% | Operators who want to run an idea the studio originated, with the machine behind them. |
Two European specifics change this maths. First, a large share of the continent’s accelerators are public or corporate and take no equity at all – regional agencies, university programmes and corporate labs run cohorts as economic development, not as a fund. Second, the EIC Accelerator sits in a category of its own: grants up to €2.5M with an optional equity component up to €10M, on a timeline measured in quarters. And infrastructure players like Station F are landlords and networks rather than investors – you pay rent, they take nothing.
Ignore the website. Call three alumni from the last two cohorts – not the showcase names, the ones nobody quotes – and ask what the programme did in the month after demo day. Anything that can’t survive that call isn’t worth 6% of your company.
Equity figures are typical ranges, not programme terms – they change every cohort. Live calls: programmes directory · Deadline Desk.
Who invests, at which stage
Angels & syndicates
Operators and founders investing their own money, typically €10k-€100k, earliest in. Fast decisions, no board seats – and increasingly organised into syndicates that compete with funds.
Venture capital funds
Professional managers investing LPs’ money for equity, from micro-funds (€20-40m) to growth vehicles (€500m+). Browse who’s actively deploying in the investor directory.
Corporate & public money
CVCs invest for strategy as much as return; public instruments (EIC, Bpifrance, British Business Bank) anchor deep tech and first-time funds across Europe. Deep tech took 36% of European VC in 2025, up from 19% in 2021 (Atomico) – much of it underwritten by exactly these instruments.
Where startups get built in Europe
H1 2026 by country: UK €18.7bn across 423 deals, Germany €6.3bn, France €6.0bn across 132, Sweden €2.8bn, Netherlands €1.9bn, Spain €1.7bn (Tech.eu). By sector: AI €5.9bn, fintech €4.7bn, healthtech €4.3bn. Country detail lives on the funding-by-country pages and the ecosystem files.
Where to incorporate
Structure follows where your customers and investors are, not where taxes are lowest. The seven most common European choices, with the startup status each offers:
| Country | Entity | Startup status / scheme | Why founders pick it |
|---|---|---|---|
| France | SAS | JEI (young innovative company) | Flexible governance, BSPCE options, deep public funding via Bpifrance. |
| UK | Ltd | EMI options · SEIS / EIS | The most investor-friendly option scheme in Europe; fastest to incorporate. |
| Germany | GmbH / UG | – | Access to the largest domestic market; notary-heavy but well understood by funds. |
| Netherlands | BV | WBSO R&D credit | English-language administration, strong holding-structure tradition. |
| Estonia | OÜ | e-Residency | Fully remote incorporation and administration; popular for distributed teams. |
| Italy | Srl | Startup innovativa | Register-based status with tax relief for investors and simplified hiring. |
| Spain | SL | Ley de Startups | Reduced corporate rate, better stock-option treatment, non-resident founder visas. |
Status thresholds and reliefs change yearly – check the national texts before choosing. EU SME definitions follow Commission Recommendation 2003/361/EC.
The 12 terms you actually need
Where the ecosystem actually meets
Europe is thirty markets with thirty legal systems and no single capital. That fragmentation is why the continent’s deal flow runs through events to a degree the US never needed: the room in Paris in November is how a Lisbon founder meets a Helsinki angel and a Berlin LP in the same afternoon.
Which is also why “which conference is worth it” is a real strategic question, not a travel one. We review the big ones on the record – cost, who’s actually in the room, and whether the hallway track converts – in Worth It? reviews, and rank the ecosystem’s most active players in the Sesamers rankings. The full companion guide – formats, investor density, how to measure ROI – is our startup events guide.
How we track this, and where to go deeper
Where to go next, depending on what you’re doing: the funding hub for daily coverage and the stage ladder, sector and country breakdowns for benchmarks in your own market; the ecosystem files for a city-level read; most active investors when you’re building a target list; the Deadline Desk and events worth your badge when you’re planning the next quarter.
Follow the ecosystem
The funding stages, explained
The figures below are the median of the rounds we hold on record at each stage - not a range copied from an American guide. A stage with nothing on record shows no figure.
Source: the Sesamers round tracker. Medians rather than averages - one outsized round would drag an average to a number nobody at that stage will see. · the data desk →