There is no correct percentage. There is a split the founders can defend in one sentence to each other and to an investor, protected by vesting so it can survive someone leaving. Everything else is detail.
This is general information, not legal or tax advice.
The equal-split argument
Y Combinator partner Michael Seibel’s essay on splitting equity among co-founders argues for equal splits. His reasoning: the work is almost all ahead of the team, and a small grant to a co-founder signals to investors that the founders do not value that person. The essay lists the usual reasons founders give for unequal splits (who had the idea, who started first, who is older) and treats most of them as weak.
The argument is strongest for a team that starts together, full time, at roughly the same point.
When an unequal split is defensible
Some differences are real and lasting. They hold up because they describe the future, not the past:
| Factor | Why it can justify a difference | Why it often does not |
|---|---|---|
| One founder is full time, the other is not yet | Different risk and output for months | Fixable by vesting start dates |
| One founder is CEO with fundraising duties | More accountability to investors | Titles change; equity does not |
| One founder brings a working product or IP | Real asset contributed | Can be valued and assigned separately |
| One founder put in cash | Money at risk | Cash should buy stock or a SAFE at a price, not founder equity |
| Idea came from one founder | Origin story | Ideas without execution are not the asset |
Cash deserves a separate note. A founder who funds the company $50,000 should get a security for that money, such as a SAFE on the same terms as an outside investor, not an extra slice of founder stock. That keeps the founder split about work and the investment about money.
How dilution treats each split
Dilution does not change the ratio between founders. It shrinks both stakes by the same factor.
Assumptions
- 10,000,000 shares fully diluted before any outside money
- 9,000,000 founder shares; 1,000,000 reserved option pool
- Pre-seed SAFE: $500,000 at a $6,000,000 post-money cap
(8.33% on conversion; before any priced-round dilution)
- Percentages are fully diluted (they count the reserved pool)
Split Before SAFE After SAFE converts
50 / 50 45.0% / 45.0% 41.25% / 41.25%
60 / 40 54.0% / 36.0% 49.5% / 33.0%
Check: 45.0% x (1 - 0.0833) = 41.25%; 54.0% x (1 - 0.0833) = 49.5%
The 60/40 founder’s majority on a fully diluted basis disappears with the first SAFE. Founders who split unevenly to give one person control should know that control comes from the board and voting agreements, and the percentage will not hold it for long.
The late co-founder problem
A co-founder who joins after the company has raised money is more expensive in tax terms, because the stock is now worth something.
Assumptions
- Co-founder joins in month 9, after a pre-seed SAFE
- 409A valuation of common stock: $0.05 per share
- Grant: 1,000,000 shares of restricted stock, vesting over 4 years
Option A: co-founder pays fair market value
1,000,000 x $0.05 = $50,000 paid; 83(b) election shows $0 income
Option B: shares issued for services, nothing paid, 83(b) filed
Income at transfer: 1,000,000 x $0.05 = $50,000 of ordinary income
Option C: stock options at a $0.05 strike instead of stock
Nothing owed at grant; exercise price $50,000 later
Each choice has a cost. Founders at formation avoid this because their stock is bought at a fraction of a cent. That is one more reason to settle the founding team before incorporation, and before any outside money comes in.
Vesting makes any split survivable
A split is only risky without vesting. With a repurchase right on unvested shares, a co-founder who leaves in month eight walks away with nothing (under a one-year cliff), and the remaining founders can use those shares to hire a replacement. YC’s essay describes four years with a one-year cliff as a typical setup in the Valley.
Vesting does not help unless the 83(b) election is filed. The 83(b) election explained article shows what happens to a founder who skips it, and the 83(b) deadline article covers counting the 30 days.
Write it down
Put the split, the reasoning and the vesting terms in a founder agreement before incorporating. The co-founder agreement template has a clause for each. The founders who found each other through the process in how to find a technical cofounder should have this conversation at the end of the trial project, while both still have the option to walk away.
A quick test for any split
Three questions, answered honestly by each founder separately:
- If the other founder left in year two, would the remaining founder feel the split had been fair?
- Can the split be explained to an investor in one sentence that is about the future?
- Would each founder accept the same split if their roles were reversed?
If the answers differ, the conversation is not finished.