Raise or Don't: The Dilution Math Founders Do After Signing
Venture money is not a prize. It is a purchase, and what it buys is the right to expect an outcome you have not agreed to yet.
Founders celebrate the round and read the mechanics afterwards. Then, two years later, a decent acquisition offer arrives and they discover it produces almost nothing for them, because the preference stack absorbs it and their ownership is a third of what they assumed.
None of that is a trick. It is written in the documents, priced into the deal, and standard across the industry. It only surprises people who treated fundraising as validation rather than as a transaction with terms.
Raising money does not answer the question of whether you have a business. It only changes the size of the answer you now need.
So run the arithmetic first. Not to talk yourself out of raising, which is the wrong default for a company that genuinely needs capital to move fast in a winner-takes-most market. Run it so that you are choosing rather than reacting to whoever said yes.
Dilution compounds, and nobody feels it until round three
Start at 100 percent between founders. A seed round at 20 percent leaves you with 80. A Series A at 20 percent takes 20 percent of what remains, not of the original, so you are at 64. A Series B at 20 percent leaves 51. Add an option pool of 10 percent created before the seed and another top-up before the A, and two founders who started with half each are now holding somewhere around 20 percent each.
That is a normal, well-run cap table with no bad behaviour anywhere in it. It is also the reason the size of the exit has to grow with every round: 20 percent of a 50 million outcome and 60 percent of a 15 million outcome pay you almost the same, and one of those two paths is far more likely to happen.
The exit does not clear the preference stack. Common shareholders, including you and your team, receive nothing. Over half the exit is absorbed by preferences. At this raise level, only a much larger outcome pays the founders meaningfully. Preferences take a visible bite but the common stock is real. This is the normal shape of a decent venture outcome. Preferences are a rounding error here. You either raised little or exited large, which is the combination everyone wants and few get.
Simplified model with a standard 1x non-participating preference and no participation or interest. Ownership compounds multiplicatively: 50 percent through a 20 percent seed, a 20 percent A and a 15 percent pool leaves 27.2 percent. On a 40 million exit with 8 million raised, investors take their 8 million back first and you receive about 8.7 million. Drag the raise up to 25 million and watch what happens to the same exit.
Play with the last slider specifically. Holding the exit at 40 million and raising 25 million instead of 8 does not reduce your outcome by a bit, it removes most of it. That is the trade you are making when you take a larger round at a higher valuation: you are betting that the extra capital produces an exit large enough to make the preference irrelevant. Sometimes it does. The failure mode is raising like you believe it and operating like you do not.
What the money has to be for
Venture capital converts cash into growth. It does that well when you already know the conversion rate, and badly when you do not.
If you know that 1,000 spent on a channel returns a customer worth 4,000 over three years, and the only limit on doing that a thousand times is cash, then raising is arithmetic. Every month you delay is compounding you did not capture.
If you are still learning which segment converts, money makes the learning more expensive rather than faster. You hire ahead of the answer, build a team around a hypothesis, and now changing your mind costs redundancies instead of an afternoon. Startup Genome's research on premature scaling identified this as the dominant failure pattern in high-growth internet startups, with roughly 74 percent of failures in their dataset attributed to scaling before the fundamentals were in place.
You do not know what works yet
Customers, deposits and your own time. Equity now prices a company whose value you cannot yet defend, and buys you a hiring plan you will regret.
You know it works, cash timing hurts
Revenue-based finance, invoice facilities, bank debt against contracts. Repaid from revenue, no ownership given up, and far cheaper than equity at this stage.
A land grab you can win with money
Venture capital. Large market, proven conversion, a competitor who will take the position if you do not. Take it and run it hard.
Match the instrument to the constraint. Most founders raise equity for the first case, which is where equity is worst: you sell the cheapest shares you will ever sell to fund the experiments you could have run for free.
The expectation you inherit
Fund economics decide what your investors need from you, and it is worth doing their maths once so their behaviour stops feeling personal.
A 100 million fund needs to return several times its capital to be considered good. Spread across roughly 30 investments, most of which return nothing, one or two companies have to produce hundreds of millions on their own. If a fund holds 10 percent of your company at exit, that requires an outcome in the billions.
So when your investor pushes for more aggressive growth after you find a comfortable, profitable niche at 4 million a year, they are not being greedy or short-sighted. They are doing the job their own investors gave them. The mismatch is structural, and the moment to notice it is before you sign, not during a board meeting two years later where everyone is technically right and nobody agrees.
One page, two columns. Left column, the bootstrapped path: what revenue can you reach in three years on customer cash alone, what does the company look like, what do you own, and what would a buyer pay for it. Right column, the funded path: how much you raise, what dilution and preference come with it, what revenue that capital plausibly buys, and what you own at the same three-year mark. Put the founder proceeds at the bottom of both columns for a realistic outcome, not a fantasy one. Show it to a founder who has done both. The page will not make the decision for you, but it will move the conversation from "should we raise" (a question about identity) to "at what point does the funded path overtake the other one" (a question with an answer).
Terms that matter more than valuation
Founders negotiate the headline number and accept the rest. The rest is where the outcomes live.
The preference stack. A 1x non-participating preference is standard and reasonable. Participating preferred, where investors take their money back and then share in the remainder, changes your economics materially in every scenario except the largest. Multiple preferences above 1x should be treated as a signal about the deal, not just a term in it.
Where the option pool comes from. A pool created before the round dilutes existing shareholders only, which quietly lowers the effective price the investor paid. Negotiate the size against a real hiring plan rather than accepting a round number.
Pro rata and information rights. Reasonable and standard. Fine to grant.
Board composition. Who controls the board controls whether you can accept an offer, change direction, or replace yourself. This matters more than two points of valuation and gets a fraction of the attention.
Founder vesting. Yes, even on your own shares, and yes it is normal. What matters is the treatment on a change of control and whether time already served is credited.
A clean, ordinary path with a 10 percent pool and three rounds at 20 percent each. Founders collectively hold about 42 percent by Series B, and each holds roughly 21 percent. Nothing went wrong here. This is what the standard path produces, which is why the required exit size grows with every round you take.
The strongest position is not needing it
Every negotiating advantage in fundraising comes from being able to walk away, and the only way to be able to walk away is to have revenue and runway. That is not an argument against raising. It is an argument about sequencing.
Founders who raise from a position of twelve months of runway and growing revenue get better terms, faster processes and investors who behave well afterwards, because the alternative to their money is visible and credible. Founders who raise with eight weeks of cash left take what is offered, and the terms reflect it. Same company, same market, same product, different month.
Which points at the practical rule: start the conversation when you do not need it, and keep enough operating discipline that the deadline never sets the price.
The takeaway
- Dilution compounds multiplicatively. Two founders on a standard three-round path typically hold around 20 percent each.
- Preferences come off the top. The size of your raise changes which exits pay you anything.
- Raise when cash converts to growth at a known rate. Before that, capital makes learning more expensive.
- Match the instrument to the constraint: revenue for uncertainty, debt for working capital, equity for a land grab.
- Board control, preference structure and pool source matter more than the headline valuation.
- Negotiate from runway. The best terms go to founders who could say no.
Frequently asked questions
How much equity do founders give up in a seed round?
A priced seed commonly costs 15 to 25 percent, often around 20, with a 10 to 15 percent option pool created at the same time out of existing shareholders. Accelerators take less; Y Combinator's standard deal is 7 percent. What matters is the compounded effect across rounds, including pool top-ups.
What does a liquidation preference do?
It lets investors take their money back before common shareholders get anything. With a 1x non-participating preference and 10 million raised, a 15 million exit returns 10 million first and leaves 5 million for everyone else. It is why modest exits can pay founders far less than their percentage implies.
When should a startup not raise venture capital?
When the business cannot reach the scale a fund requires, when the money would fund experiments rather than a known growth mechanism, or when the outcome you would be happy with is one the round makes impossible.
What are the alternatives to venture capital?
Customer revenue and deposits, revenue-based financing repaid as a percentage of monthly revenue, bank or government-backed lending against contracts and receivables, and grants where they exist. Each is cheaper than equity when the constraint is working capital rather than uncertainty.
The round was the high point. Nobody said so at the time.
Kill My Startup follows what happens after the money lands, and why funding so often accelerates the wrong thing.
Buy on Amazon →Sources
- Standard venture financing terms: 1x non-participating liquidation preference, pre-money option pools, pro rata rights.
- Y Combinator's published standard deal terms.
- Startup Genome Report Extra on premature scaling as a dominant failure pattern in high-growth internet startups.
- Brad Feld and Jason Mendelson, Venture Deals, on term sheet mechanics and control provisions.