Why this matters
"$1M, because that's what seed rounds are" is how founders end up either out of money before proving anything or over-diluted with burn creep. The right raise is sized backward from a milestone, not picked from a round-number menu.
What "done" looks like
- The milestone that makes you fundable at a higher price is written down
- A monthly cost model shows what reaching it takes
- Your number covers a context-appropriate runway, often 18 to 24 months, with a buffer
- You can defend the amount and the dilution it implies in one minute
How to do it
- Name the next fundable milestone, the proof point that justifies a better valuation, not "more runway."
- Budget the monthly burn to reach it: people, tools, marketing.
- Multiply by the months required, then add an appropriate buffer, often 25 to 50%. Everything slips.
- Sanity-check runway: 18 to 24 months is a common benchmark, depending on the company and market, so you aren't fundraising again in nine.
- Check the dilution: selling 20 to 25% can be a useful early-round benchmark, but the right level depends on the company and deal. More dilution affects your cap table for every round after.
- State a defensible range, not a single magic number.
Common mistakes
- Copying a peer's round size, their milestone isn't yours
- Sizing to survival ("we need cash") instead of to proof
- Assuming a top-up will be easy later; bridge rounds are hardest exactly when milestones slip
Real-world examples
- Y Combinator's published fundraising advice frames the amount the same way: months of runway times burn, aimed at a milestone that unlocks the next round.
- Paul Graham's essay "Default Alive or Default Dead?" offers a related test: know whether current numbers reach profitability before the money runs out.
From a founder's point of view
The size of this raise sets the difficulty of the next one. A modest round fully deployed against a clear milestone beats a big round absorbed by payroll. Dilution is forever; the buffer is negotiable; the milestone is the point.
Rule of thumb
Your number should equal monthly burn × months to milestone, plus buffer. If it didn't come from that arithmetic, it isn't your number.
