Parallel, Turbopuffer, and Wordsmith run their entire GTM motion on Attio, with agents that chase every buying signal, build pipeline, and move deals forward, 24/7. Try Attio now.
A founder raises 3 SAFEs over 18 months: 250k at a $5M cap, $500K at $10M, $750K at $18M. Then a seed lead sends a term sheet: $4M at $20M post money for 20% .
She owns 90% today, with a 10% option pool. She does the arithmetic everyone does. The lead takes 20%, so she lands at 72%.
She lands at 59%.
Nothing went wrong. No bad terms, no down round, no surprise investor. Every one of those 13 points was signed for. She just never added them up in the order they actually get taken.
The number behind this issue
A 2026 founder guide from Velawood and Horizon Capital found that 83% of founders who sign post money SAFEs without modelling them first take a bigger equity hit at conversion than they expected. Capwave audited more than 500 rounds and found that when a founder signs two or more post money SAFEs at different caps without modelling the stack, the gap between expected and actual dilution averages 8 to 14 percentage points.
Treat those as direction rather than decimals. They come from firms that sell cap table help, and that kind of firm has a reason to find founders confused. I trust the direction anyway, because the math reproduces it. The example above is built from ordinary terms, and it lands at 13 points, inside that range.
Three ways the math goes wrong
Treating SAFEs as "not equity yet." A post-money SAFE is ownership sold at signing and delivered later. The percentage is fixed the day you sign:
amount ÷ post-money cap; $250K ÷ $5M is 5%. Our founder had sold 14.2% of her company before the seed lead showed up. The cap table on her screen still said 90%.
Assuming each new SAFE shares the pain with the last one. Under the old pre-money SAFE, later SAFEs diluted earlier ones. Under the post-money SAFE, which has been YC's standard since 2018, they don't. Every SAFE you add comes entirely out of common stock and the pool. The investors on your stack are protected from each other. You are not protected from any of them.
Forgetting that the priced round has its own terms. The lead's 20% is measured after the round, and most leads ask for the option pool to be topped up before their money goes in. Here the lead wanted a 10% pool after closing, and that cost 1.7 points. At a 12-15% pool the same line costs 3 to 5 points. We took the pool apart in the 24 September issue; in a SAFE-heavy seed it is the smallest of the three pieces.
One more to check in your documents: a most-favoured-nation (MFN) clause. If an early SAFE carries one and you later give someone better terms, the early investor can upgrade to them. The stack you modelled is not always the stack that converts.
The expensive money is the early money
Founders spend weeks negotiating the seed lead's valuation. Then they sign the first SAFE in an afternoon. The chart below shows why that is backwards.
The first cheque was probably the right call. It may have been the only money available, and it kept the company alive. It was still priced like early stage risk, and it stays on the cap table at that price. Your leverage on dilution is highest before the seed, when every SAFE is small enough to feel harmless.
A quick check before your next SAFE
Don't ask what percentage the SAFE is. Ask what it will cost you after the priced round:
cost after seed ≈ (amount ÷ cap) × (1 − lead % − pool top-up)
In this example that multiplier is 0.78. A $500K SAFE at a $10M cap reads as 5%. After the seed it has taken 3.9 points from founders. Then add it to the SAFEs you've already signed and look at the total.
What to do this week
Write down your stack. List every SAFE and convertible note with its amount, cap, discount and any MFN or side letter. Add up amount ÷ cap. That total is the percentage you've already sold.
Set a SAFE budget before the next cheque. Pick a ceiling for total pre-priced-round dilution and stick to it. Capwave suggests 20–22%. The right number for you depends on how many rounds you expect to raise, but choosing one is better than having none.
Ask your likely seed lead about the pool early. Find out the target size, and whether it's measured before or after the money. Come with an 18-month hiring plan, so the pool is sized to that plan and not to a default.
Watch the low-cap money. If you need a bridge, compare the price per point against the round you're bridging to. Sometimes a smaller bridge and a tighter budget cost less than a bigger bridge at a low cap.
The worked example
This is the scenario behind all three charts. It sums to 100%.
Want your own stack modelled before you sign?
Book a call with me here. Bring your SAFEs and your term sheet, or the one you expect. We'll work out where you land after the round and which lever moves it most.
And if this saved you a bad afternoon with a term sheet, forward it to one founder who is mid raise.
Talk soon,
Imane