Every founder knows dilution is coming. Far fewer can compute it. The difference shows up at exactly the wrong moment: a term sheet on the table, a lawyer on the clock, and a pro forma cap table whose bottom line is fifteen points worse than the mental math said it would be.
This is startup dilution explained end to end, not as a vibe ("you'll own less, but of more") but as arithmetic. By the end you'll be able to take a SAFE stack, a round size, a valuation, and a pool top-up, and compute your post-round ownership on the back of an envelope.
What dilution actually is
Nobody takes your shares. Dilution happens because the company issues new shares to someone else. Your share count stays exactly the same; the total grows; your percentage (your shares divided by the total) falls.
That gives you the only formula you need:
New ownership = old ownership × (old share count ÷ new share count).
Say you hold 4,000,000 of 8,000,000 shares, 50%. The company issues 2,000,000 new shares to an investor. You still hold 4,000,000, but now out of 10,000,000: 40%. Every existing holder scaled by the same factor, 8M ÷ 10M = 0.8. Your co-founder's 30% became 24%. The advisor's 1% became 0.8%.
That uniform scaling is the most useful fact in this lesson. A share issuance never singles anyone out. It shrinks everyone who already held shares by the same ratio. So instead of tracking share counts, you can work entirely in percentages: if a round hands 20% of the post-round company to new holders, every existing percentage gets multiplied by 0.80. We'll use that shortcut for everything below.
The two levers of a priced round
A priced round dilutes you through two separate mechanisms, and they're negotiated separately.
Lever 1: new money. Investors buy newly issued shares. Their ownership is simply investment ÷ post-money valuation. Raise 2M at an 8M pre-money (a 10M post-money) and the new investors own 2M ÷ 10M = 20%. Existing holders scale by 0.80.
Lever 2: the option pool top-up. Most term sheets require the option pool to be created or expanded to some target, commonly 10% of the post-round company, before the round, inside the pre-money valuation. That wording matters enormously: shares reserved pre-money dilute only the existing holders. The new investors buy their 20% of a company that already includes the enlarged pool, so they don't pay for it. You do. This is the option pool shuffle, and it's the most reliably underestimated line in a term sheet.
Stack the levers: if new money takes 20% and the pool top-up adds 10% of the post-round company, existing holders keep 100% − 20% − 10% = 70% of the pie, shared in their old proportions. Everyone who held shares before the round scales by 0.70, not 0.80. The second lever cost you half as much as the first, and it never wired you a dollar. (How pools are sized, and how to negotiate them from a hiring plan instead of a default, is covered in the option pools lesson.)
Layering the SAFE stack on top
Real seed rounds rarely start from a clean cap table. There's usually a stack of SAFEs converting at the same moment, and the order of operations matters.
Post-money SAFEs (the standard since 2018, see how SAFEs work) lock in ownership at signing: investment ÷ post-money cap, measured on the company before the new round's money and pool top-up. So the computation runs in two steps. First, convert the stack: SAFE percentages come off the founders' side, point for point. Then run the round: pool top-up and new money dilute everyone, founders and the freshly converted SAFE holders, by the same scaling factor.
Worked example: 3 SAFEs + a priced round + a pool top-up
Two founders own 100% of the company. They've raised three post-money SAFEs:
SAFE 1: 250k at a 5M cap → 5.0%. SAFE 2: 400k at an 8M cap → 5.0%. SAFE 3: 300k at a 12M cap → 2.5%. The stack has committed 12.5%, so immediately before the round, the pro forma reads founders 87.5%, SAFE holders 12.5%.
Now the round: 2M of new money at an 8M pre-money (10M post-money), with a pool topped up to 10% of the post-round company, created pre-money. New money takes 20%, the pool takes 10%, so every pre-round holder scales by 0.70:
Founders: 87.5% × 0.70 = 61.25%.
SAFE holders: 12.5% × 0.70 = 8.75%.
Option pool: 10%. New investors: 20%. Total: 100%. ✓
The founders' journey from 100% to 61.25%, a total of 38.75 points of dilution, decomposes cleanly: 12.5 points to the SAFE stack, 17.5 points to new money (87.5% × 20%), and 8.75 points to the pool (87.5% × 10%). Note that only 2M of the three sources actually arrived at the round; the SAFE cash came earlier and the pool is a reservation, not revenue.
Two habits follow. Keep a running total of committed SAFE ownership (investment ÷ cap, summed) and treat it as spent the day you sign. The SAFE-stacking walkthrough shows how fast a casual stack reaches 20%+. And when a term sheet arrives, always compute the combined scaling factor (1 − new money % − pool increase %) before reacting to the headline valuation.
Cumulative dilution: seed to Series B
One round is a multiplication. A financing history is a chain of multiplications, and this is where intuition reliably fails, because people add dilution percentages when they should multiply retention factors.
Dilution compounds multiplicatively, not additively. If three rounds dilute existing holders by 30%, 25%, and 18%, you have not lost 73 points. You keep 0.70 × 0.75 × 0.82 = 43% of what you started with. Each round dilutes a smaller base, so adding the percentages always overstates the damage, and founders who add them either panic early or, worse, stop checking.
Worked example: the founders' stake across three rounds
Continue from the seed example above: the founders exit the seed at 61.25%.
Series A: 20% to new money, pool refreshed by 5 points of the post-round company. Scaling factor 0.75. Founders: 61.25% × 0.75 = 45.9%.
Series B: 15% to new money, 3-point pool refresh. Scaling factor 0.82. Founders: 45.9% × 0.82 = 37.7%.
From incorporation to Series B, the founders kept 0.6125 × 0.75 × 0.82 = 37.7%, a fairly healthy outcome, and squarely in the range you'd expect: founders typically hold somewhere around 50–65% after seed, 35–50% after A, and 25–40% after B, with wide variance. Rerun the chain with your own round sizes in the free dilution calculator.
The chain also shows where the leverage is. The seed cost these founders 38.75 points; the Series B, a bigger round in absolute terms, cost them only 8.3 points (45.9% × 18%), because the base was smaller and the terms were lighter. Early-round terms dominate lifetime dilution. A 2-point difference in a seed pool top-up is worth more than a 4-point difference at Series B.
When dilution is worth it, and when it isn't
Dilution is not loss. It's the price of the round, and like any price it can be fair or terrible. The test is never the percentage alone. It's value per point. The founders above ended the seed at 61.25% of a 10M post-money company: a stake marked at 6.1M, versus 100% of something unpriced and unfunded. If the capital raised grows the company faster than the percentage shrank, the trade worked.
Dilution goes bad in recognizable ways: raising at a flat or down valuation, so the percentage falls without the value rising; conceding an oversized pool top-up that sits unused (unissued pool at exit is dilution you donated); stacking low-cap SAFEs without keeping the running total, so 25% is committed before anyone prices the company; and heavy anti-dilution or ratchet terms from a prior round shifting a down round's cost onto common, covered in the term sheet lesson. Every one of these is visible in advance to anyone who runs the numbers.
Protecting yourself: model before you sign
Every unpleasant dilution surprise in this lesson has the same fix, applied at the same moment: model the transaction before the signature, not after. Concretely:
- Keep the SAFE running total. Investment ÷ cap, summed across every open SAFE, visible at all times. Decide your ceiling (say, 20%) before you need it.
- Demand a pro forma with the term sheet. A term sheet without a post-close cap table is a price without an itemized bill. Recompute it yourself: SAFE conversions first, then pool, then new money.
- Size the pool from a hiring plan, not from the investor's default. Every point of pre-money pool is a point of your side, spent.
- Chain the future. Model not just this round but the next two at plausible terms. The multiplication is trivial once the retention factors are written down.
Spreadsheets can do all of this, but they rot between rounds, and the SAFE-conversion step is exactly where hand-built models break. This is the job Vquity's scenario modeler was built for: it chains hypothetical rounds and exits on top of your live cap table (SAFE stack, pool top-up, and all) so the ownership outcome of a term sheet is a number you read, not one you reverse-engineer at midnight.
You now have dilution as mechanism: issuance grows the denominator, everyone scales together, two levers per round, multiplication across rounds. The next lesson turns to the number that anchors so much of this math: the valuation itself, and the 409A/FMV machinery behind the price employees pay for their options.