I watched a founder get a real exit and still lose the outcome to someone else's timeline.
I was the first angel in. For me, it was a strong exit. For the founders, it wasn't, and the gap between those two outcomes is worth sitting with. Most founders think dilution is the risk they're managing when they raise. The actual risk is raising past what you've actually proven. Dilution is just the invoice that arrives later, itemized.
Here's what happened. The company had found a real segment, with a repeatable buyer and defensible margin, and could have exited cleanly while the founders were still majority owners. Instead it raised more capital than its own evidence required, pushed toward opening offices in several countries and diversifying into adjacent products before either move had earned any proof of its own. Most of it didn't work. What it produced instead was a series of bridge rounds on progressively harder terms, each one taking a larger share of the founders' ownership than the round before it.
The exit, when it came, was real. The company mattered enough for someone to buy it. But for the founders it arrived heavily diluted by rounds that had bought time and geography, not proof. One of them said afterward, without needing to elaborate: they could have had a better outcome without the VCs at all.
An exit can be proof the company mattered, and still be evidence that someone else decided how you'd get there.
The math founders don't do until it's too late
Founders track control closely: veto rights, board seats, reserved matters, information rights. That half gets learned inside a year, because it shows up in every meeting. The other half, liquidation preferences, participation, vesting, drag-along, doesn't get read closely until the day a buyer signs. That's exactly why it surprises.
Do the arithmetic once, slowly. It only takes doing once to change how you read every future term sheet.
A company sells for €30M. It raised €20M. Founders hold 40%, late-stage investors 30%, early investors the rest. Assume the preference sits only on the latest round and doesn't participate. The late-stage investors compare their €20M preference against the €9M their 30% would otherwise return, take the preference, and the remaining €10M splits by ownership among everyone else. Founders receive €5.7M. They own 40% of the company. They receive 19% of the price.
Now change one number. The same €20M at a 2x preference is a €40M claim against a €30M sale. The late-stage investors take everything. Founders receive nothing, and so do the early investors. Nobody behaved badly, and nobody broke the agreement. What moved was the finish line: founders now need €40M before they see their first euro, and reaching that original €5.7M would require selling for €50M instead of €30M. One negotiated multiple just moved €20M of required enterprise value.
Preferences stack. Each round adds its own layer whether or not the valuation moves, so a founder's break-even rises with every round that gets celebrated as progress. That's the invoice dilution doesn't show you on the cap table summary.
What you're actually raising against
In the cases I've seen, every founder who raised past their proof did it for the same reason: capital feels like it fixes the gap. A displacement hull has a maximum speed set by its own physics. Install a bigger engine and the speed barely changes while fuel burn multiplies. My partner Risto calls it runkonopeus, hull speed. In a company, capital is the engine. What's actually been proved is the hull. Capital can buy a bigger sales team, more offices, more adjacent products. It can't buy the buyer proof, value proof, or scaling proof that would make any of that spending rational.
The founders in that scar didn't get diluted because they were bad negotiators. They got diluted because the company raised against ambition instead of evidence, and every round that didn't earn a proof layer was capital that only bought time to be wrong at a larger scale, on worse terms.
I've sold two companies I founded, one to a strategic buyer, one to a buyout fund, and the rounds that hurt the least were the ones tied to a specific, provable thing the company had just earned. The ones that hurt were raised because the market was open, or because a competitor had just raised, or because growth felt like it needed fuel.
This is not an argument against raising, or against VCs. The founders who've exited well in the cases I've watched raised because a specific proof layer (a segment, an adoption motion, a scaling mechanism) genuinely needed a bigger hull than they had, and they said so explicitly before the round, not after it stalled. Hull speed cuts both ways. Sometimes the honest answer isn't grinding harder inside your current shape; it's admitting the next layer requires capital or competence you don't have, and bringing in the bigger engine deliberately, eyes open about what it costs in ownership. The failure mode isn't raising. It's raising to avoid naming what you haven't proven yet.
Vetting the investor like you'd vet a buyer
If capital is the thing that can quietly cost you the company, treat the investor relationship as a claim requiring evidence. Call it Investor Proof, and check it through behavior, not the pitch. Call references yourself, including one from a portfolio company that had a bad year, not just the highlight reel. Talk to the actual partner who will sit on your board, not the associate who disappears after the term sheet is signed. Get economics on the table, in numbers, before you sign anything.
Access belongs in the same test, and it's the cheapest claim in the asset class to make and the hardest to verify. Every investor has a slide about the network. A promise of introductions costs nothing. An introduction actually made during diligence, to a named seat in a market you're trying to enter, is behavior, and it's worth paying for in dilution. It's not worth paying for in control over the outcome. Ask, at term sheet stage, whether the preference is stacked or pari passu. That question costs nothing before signature and can't be asked after it.
An investor irritated by that scrutiny has just shown you how they'll handle scrutiny they don't control later, when it matters more. The founders in that scar never ran this check. By the time the bridge rounds started, there was no leverage left to run it with.
One more number is worth tracking alongside your ownership percentage: how many credible buyers your company could plausibly have on the day you'd want to sell. Every round raised past proof narrows that list, because it raises the minimum price a sale has to clear before your own investors will support it, and a shrinking buyer universe is a founder's problem long before it's an investor's. A company with several buyers who each have their own reason to want it has leverage. A company that needs one specific buyer to justify a stacked cap table has none.
Before the next round, name the specific thing the capital is meant to prove (not “growth,” not “runway,” the actual buyer, value, or scaling proof that doesn't exist yet) and size the round to that instead of to what the market will give you. A raise with no named proof behind it is a bridge round wearing a growth round's clothes, and the founders in that scar paid full price to learn the difference.
More scars than trophies.