What VCs actually look for at seed
The fund maths behind every seed decision - why a reliable 3x gets rejected, what quality of demand really means, and the questions to ask investors back.
Also see more startups coverage: /category/startups/
At seed, VCs are looking for one thing: a credible case that your company could return a meaningful fraction of their entire fund – which in practice means a small set of users or customers who demonstrably cannot live without the product, a team with an unfair reason to win, and a market big enough that winning it matters.
Most fundraising advice fails founders because it describes what VCs say they want rather than the maths that drives what they fund. Start with the maths and everything else – the obsession with market size, the indifference to your respectable growth, the strange questions about what happens if you shut the product off – stops being mysterious.
The fund maths, first
A venture fund is not trying to make good investments; it is trying to make a small number of extraordinary ones. Returns in venture follow a power law: in most funds, one or two companies generate the majority of the returns, a few return their money, and the rest go to zero or nowhere. The consequence is counterintuitive and explains almost every VC behaviour founders find irritating: a company that will reliably triple an investor’s money is a bad venture bet. A reliable 3× cannot pay for the rest of the portfolio’s failures; only potential outliers can. So the seed investor across the table is not asking “will this work?” – they are asking “if this works, is it enormous?” and, only then, “what evidence exists that it might?”
The bar this implies is high, and the funnel data says so: of 3,075 European companies that raised seed rounds in 2016-18, 27% raised a Series A within 36 months (Dealroom cohort analysis – dated now, but the only clean cohort study of its kind for Europe). Seed investors know most of their portfolio will not graduate. They are pricing for it. For what the next gate looks like, see what is a Series A.
The seed burden of proof: quality over quantity
At seed nobody expects scale – across the rounds Sesamers tracks, the median seed is €5M (n=74), which buys the discovery of a repeatable business, not the business itself. What good seed investors demand instead is quality of demand, and it is a different axis from quantity:
- Retention of a small cohort beats acquisition of a large one. Forty users who are all still active in week twelve are stronger evidence than four thousand sign-ups with a leaky bucket. Retention is the one early metric that is hard to fake and directly measures whether the product matters.
- The scream test. The question behind the questions: who screams if you take the product away? If the honest answer is “nobody, politely”, there is no demand yet – there is curiosity. Founders who can name the customers who would scream, and say why, clear the bar most pitches miss.
- Pull, not push. Inbound requests, users hacking the product to do things it wasn’t built for, customers chasing you for the invoice. Evidence that acquisition happened to you is worth ten times evidence you manufactured.
Team signals
“We invest in teams” is the most repeated and least explained line in venture. Concretely, seed investors are pattern-matching for: earned insight – a reason this team sees something about the market that others structurally cannot; speed of iteration – what changed between your first conversation and your second, because the delta is the diligence; founder-market fit – the harder the industry, the more it matters that you have lived in it; and the ability to recruit – the first employees you attracted are read as a proxy for the executives you will attract later. Note what is absent: pedigree helps get meetings, but at the decision, evidence of motion beats CVs.
The market-size interrogation
Because of the fund maths, market questions are existential, and investors have heard every inflated TAM slide ever produced. What they are actually probing: is the credible market – the buyers your product as it exists could plausibly win – large enough that the outlier scenario returns a fund? A convincing answer is built bottom-up (number of realistic buyers × realistic price) and paired with a believable expansion story. A top-down slide (“if we get 1% of a $50bn market”) is not just weak; it signals you have not done the thinking. If your bottom-up market is honestly small, a good investor will pass respectfully – and that pass is information, not injustice: it may mean your company is a fine business that should not take venture money at all.
Red flags that end the conversation
| What founders present | What investors hear |
|---|---|
| “We have product-market fit” (pre-revenue, no retention data) | The founder doesn’t know what PMF is – or hopes I don’t |
| Cumulative sign-ups, app downloads, “registered users” | Vanity metrics chosen to hide the active-usage line |
| Pipeline presented as revenue (“€800k in advanced discussions”) | Optimism accounting; discount everything else in the deck |
| “We have no competitors” | Either no market or no research – both fatal |
| Growth entirely from paid channels, no organic pull | Demand is being bought, not discovered |
| A metric that changes definition between meetings | The number was the costume, not the body |
The pattern under all of these: seed investors do not penalise smallness – they penalise dressing smallness up as scale. “Twelve users, nine still active in month three, two asked to pay” is a fundable sentence. The same company hidden behind a cumulative sign-up chart is not.
How European seed diligence differs
The folklore says European VCs are more conservative than American ones. The qualitative reality is more structural: European seed funds are typically smaller, which changes their maths – a smaller fund cannot spray as many bets, so each cheque carries more diligence weight, and founders should expect more traction scrutiny per euro raised. Processes lean harder on evidence and reference calls, and less on momentum and FOMO, than the US playbook assumes. None of this means European seed is starved: H1 2026 saw €44.1bn across 1,740 deals in European tech (Tech.eu, July 2026), on the back of roughly $44bn raised in 2025 (Atomico, State of European Tech 2025). The money is there; it simply asks better questions. Calibrate accordingly: bring the retention cohort, the named referenceable customers, and the bottom-up market – the things a smaller fund needs to defend the deal internally.
What to ask them
A seed round is hiring a board member you cannot fire. Interrogate back:
- “Where are you in your fund’s life?” A fund in year one behaves differently from a fund in year four hunting for its last deals – reserves, patience and follow-on behaviour all differ.
- “What do you do when a company misses its milestones?” Then verify the answer with founders from their portfolio – including one whose company failed. How an investor behaves in the bad quarter is the entire product.
- “Who from the fund will actually sit on my board, and what have they built?” The brand invests; a person shows up.
- “What would need to be true for you to lead my Series A?” The answer tells you their real conviction and hands you your milestone list for the next eighteen months.
Frequently asked questions
How much traction do I need to raise a seed round in Europe?
There is no threshold number, and anyone quoting one is selling something. The honest standard is evidence of demand quality – retention of a real cohort, users who would scream if you switched the product off – sized against a story big enough to matter. European funds tend to want that evidence more fully formed than US folklore suggests.
Do VCs really reject profitable, growing businesses?
Yes, constantly, and it is not a moral judgement. A business that will reliably 3× is a good company and a bad venture bet, because the power law demands outliers. If that is your business, the right conclusion is usually different capital – not a different pitch.
What single metric matters most at seed?
If forced to one: retention of your early cohort, however small. It is the hardest early number to fake and the most direct measurement of whether the product matters to anyone. Everything else – sign-ups, downloads, even revenue from unrepeatable deals – can be manufactured.
Should I mention that most startups fail?
Investors know the funnel better than you do – in the classic Dealroom cohort, 27% of 2016-18 European seed companies reached Series A within 36 months. You do not need to recite failure statistics; you need to show why the specific risks in your company are being retired, quarter by quarter.
Keep going
For the foundations, start with what is a startup and the mechanics of the round itself in what is seed funding. When the conversation turns into paper, our term sheet guide covers the terms that matter, and what is a Series A shows the next gate. For the numbers behind European rounds – who is investing, at what size – explore our fundraising data hub.
