Co-Founder Equity Split by the Numbers and How to Agree on Yours

Most two-founder teams now split their company almost evenly. In 2024, 45.9% of two-person founding teams on Carta divided equity exactly equally, up from 31.5% in 2015, and the median split between the two founders had narrowed to 51/49 (Carta, Founder Ownership Report). The old instinct that the person with the idea, or the CEO title, should hold a clear majority has been losing ground for a decade.

That makes the percentage the easy part. What decides whether a split holds up is what sits behind it: vesting that returns unearned equity when someone leaves, an 83(b) election filed inside 30 days, cash treated differently from work, and a written agreement that says all of this before anyone needs it. A 50/50 split without those is more dangerous than a 60/40 split with them.

This guide gives the figures on how teams actually split, a way to decide your own, the worked arithmetic for vesting and for cash, and the rules that apply when the business is an LLC or partnership rather than a venture-backed corporation. It is general information; a lawyer in your state should read the actual agreement before anyone signs.

Carta's report covers more than 45,000 US startups incorporated from 2015 through 2024, counting as a founder any individual who held at least 5% before any venture money came in (Carta). It is skewed toward companies that intend to raise venture capital, which is worth remembering if the business is a two-person agency or a restaurant. Even so, it is the largest dataset on the question, and its figures are consistent with the main academic study.

Source: Carta, Founder Ownership Report, data as of 1 January 2025.

The academic picture came first. Thomas Hellmann and Noam Wasserman studied 1,476 founders in 511 ventures and found roughly a third split equally, and that 42% of teams settled the split within a day or less (NBER digest; Management Science, 2017). Three things made an unequal split more likely: one founder having had the idea, one having started a company before, and one putting in more capital.

Two other Carta figures belong in the conversation. Team size is shrinking: 35% of startups formed on Carta in 2024 had a single founder, up from 17% in 2017. And founders' stakes fall fast once money comes in: the median founding team holds 56.2% after a seed round, 36.1% at Series A and 23% at Series B (Carta). A split argued over to the last percentage point at formation is usually a split of a much smaller share a few years later.

The case for equal is that the work is ahead. Michael Seibel of Y Combinator argued in 2015 that small differences in what each founder did in the first months do not justify very different stakes over the seven to ten years a company takes, and that a CEO who keeps far more than a co-founder signals that the co-founder is not valued (Y Combinator). If two people are quitting their jobs on the same day to work full time on the same thing, that argument is hard to beat.

The case against is about the founders who are not alike. Hellmann and Wasserman's working paper found equal splits associated with lower pre-money valuations at the first financing, most strongly among teams that agreed in under a day, and put the value at stake at about 10% of the firm's equity (NBER). The fair objection is that the published version of the paper concluded the link with outside funding comes from which teams choose equal splits rather than from the split itself (SSRN). That objection holds, and it narrows the finding to this: a quick, unexamined 50/50 tends to come from teams that have not talked about differences that exist.

A workable rule, then:

• Split equally when the founders start full time together, will carry comparable roles, and neither is bringing assets the other is not.
• Split unequally when one founder is full time and the other part time, when one has already built something (code, customers, a lease, a license) before the other arrives, or when one is taking far more risk, such as giving up a salary the other never had.
• Never use equity to pay for cash. Money a founder puts in should be handled as money (below), so the split can reflect work alone.

Pure 50/50 between two people carries one mechanical risk regardless of fairness: a tie. With two equal holders, any vote that needs a majority can deadlock. Teams that split evenly usually give one founder the casting vote on day-to-day operating decisions, name a tie-break procedure (a trusted third director, mediation, then a buy-sell mechanism), or both. The equity can be equal while the decision rights are not.

Teams that disagree usually disagree about weights, not facts. Writing down the factors, agreeing what each is worth before anyone fills in names, and only then scoring each founder takes most of the heat out of it. The factors that matter are the ones the research and the investors keep coming back to: time committed over the next few years, the role each will carry, what each has already built, what each gives up to do this, and relevant experience. Capital stays out of it.

Here is that method run on a plainly hypothetical three-person team.

The site's calculator does the same arithmetic with your own factors and weights.

Two adjustments are common after the scoring. Some teams round the top two to equal shares when they are within a point or two of each other, which the 2024 data suggests most teams now do anyway. And a founder who arrives later than the others is often given a smaller share and a vesting clock that starts on their own start date; in Carta's data the third founder on a three-person team held a median of 22% in 2024 (Carta).

The mistake that most often skews a split is paying for cash with founder shares. Founder stock is issued at a nominal price, often a fraction of a cent a share, so a founder who puts in $40,000 at that price would buy nearly the whole company. Turning cash into extra percentage points instead mixes two different things and makes both harder to price.

The cleaner approach is to split the company for work and treat the cash as what it is: a loan the company repays, or an investment on the same paper outside investors would get. On a post-money SAFE, the stake bought is the amount invested divided by the post-money valuation cap (Y Combinator).

Without vesting, a founder who leaves after three months keeps everything they were issued. The standard fix is four-year vesting with a one-year cliff: nothing vests in the first year, a quarter vests on the first anniversary, and 1/48 of the grant vests each month after that (Y Combinator). Founder stock is normally issued up front and made subject to the company's right to buy back unvested shares at the original price if the founder leaves, which keeps the founder a shareholder from day one.

Departures are not rare. In Carta's data the share of two-founder teams that split up in their first year nearly doubled between 2015 and 2024, from 4% to 7.5%; fewer than a quarter split within four years, and more than 60% were still together after eight (Carta). Those figures only count companies still operating, so they understate how often a departure ends a company.

Shares that can be bought back while they vest are, for tax purposes, not fully the founder's until they vest. Without an election, each vesting date is a taxable event: the founder owes ordinary income tax on the difference between what the shares are worth that day and what was paid for them. If the company has raised money and the shares are worth real money by then, that is a tax bill on shares that cannot be sold.

The 83(b) election moves the taxable moment to the day of issue. Because founders pay the full fair market value at issue (a fraction of a cent a share), the taxable amount at that moment is zero, and later growth is taxed only as a capital gain when the shares are sold.

Say B above paid $300 for 3,000,000 shares at $0.0001. With an 83(b) filed, B owes nothing now and nothing at vesting. Without it, if a seed round values the common stock at $0.50 a share by B's first anniversary, the 750,000 shares vesting that day carry about $375,000 of ordinary income ($0.50 less $0.0001, times 750,000), with more each month after.

The election also starts the holding period that matters for the federal exclusion on qualified small business stock in a C corporation. For stock issued after July 4, 2025, that exclusion is 50% of the gain after three years, 75% after four and 100% after five, up to the greater of $15 million or ten times basis, for companies with no more than $75 million of gross assets (Grant Thornton). Keep the IRS confirmation or the certified-mail receipt with the company's records; investors and acquirers ask for it.

Most small businesses with two owners are not Delaware corporations raising venture money. They are LLCs or general partnerships, and three things change.

The default is equal, whatever anyone put in. Under the Revised Uniform Partnership Act, adopted in most states, each partner is entitled to an equal share of the profits unless the partners agree otherwise (RUPA section 401(b); for example Maryland's version). States that adopted the Revised Uniform LLC Act make distributions before dissolution equal among members by default (RULLCA section 404; for example Arizona's). A partner who put in $100,000 and one who put in nothing share profits equally if nothing is written down. Not every state works this way: Delaware's LLC Act splits profits and distributions by the agreed value of each member's contributions when the agreement is silent (6 Del. C. sections 18-503 and 18-504). Check your own state's act.

Money, profit and votes can be split separately. An operating or partnership agreement can give equal votes, return capital first, and share profit after that by any ratio. That is usually the fairest answer when one owner brings the money and the other brings the work.

A partner who joins later can get a profits interest. A share of future profits and growth, with no claim on what the business is already worth, is generally not taxed when granted for services under Rev. Proc. 93-27, unless the interest is sold within two years, comes from a substantially certain income stream, or is in a publicly traded partnership. A share of existing capital is taxed as income on its value (The Tax Adviser). Advisers commonly file a protective 83(b) election when a profits interest vests.

How profit is shared between partners, as opposed to who owns what, is its own decision; revenue share and equity partnerships compares the two structures.

The best-known case is Snapchat. A third early participant sued, claiming a share of the company and of the disappearing-message idea; Snap disclosed in its 2017 IPO filing that it had paid $157.5 million to settle, $50 million in 2014 and $107.5 million in 2016 (Forbes). The founders disputed that he was ever a co-founder, which is the point: with no signed split, no vesting and no assignment of the idea to the company, who was a founder became a question for a court.

The ordinary version is smaller and more common. A founder leaves in year two with unvested equity and no repurchase right; the remaining founders do the next five years of work for someone who owns as much as they do. Or two LLC members never wrote down that one put in the money, and the default rule splits everything equally. Buying out a partner is far cheaper when the price and the method were agreed at the start, and the warning signs in why business partnerships fail usually show up before the split is signed.

The site's agreement checklist goes through the rest of a partnership or operating agreement clause by clause.