पाठशाला Pathshala · दल Dal, The team · Lesson 03 · Start

Splitting founder equity: equal, unequal and how to decide

Nearly half of two-founder teams now split equally, and a quick equal split costs the stronger founder at the first round. How to decide from contribution, risk and future role, and write it down so it survives success.

Pathshala, The Founder Library · 11 October 2026 · 8 min read

The equity split is the first negotiation a company has with itself, and most teams settle it in under a day because the alternative is an awkward evening. The research says the awkward evening is worth having, and that the answer is usually closer to equal than the founder who wants more believes, and further from equal than the founder who wants peace will admit.

This lesson gives the data on what founders actually do, the case for equal and the four reasons that justify leaving it, a scoring method, a figure that shows what any split is worth after the first round, and the way to document the decision so that it does not have to be made twice.

What founders actually do

Carta’s cap-table data is the largest record of the question. Across more than 32,000 companies with multiple co-founders incorporated between 2015 and 2024, about 24 per cent split equity equally. For two-person teams the share is much higher and rising: 45.9 per cent split equally in 2024, up from 31.5 per cent in 2015, and the median two-founder split narrowed from 60–40 to 51–49 over the same decade. Three-person teams moved the same way, from 12.1 per cent equal to 26.9 per cent; in the 2019 cohort the lead founder of a trio held 50 per cent and the third 13 per cent, a gap of nearly four times, which had closed to 44 and 22 by 2024.

The academic record is older and points the other way on one thing. Thomas Hellmann and Noam Wasserman’s study of 511 ventures, The First Deal, found roughly a third of teams splitting equally, and found that equal splitters raised their first round at lower pre-money valuations, with the effect strongest in teams that had settled the split in less than a day. Their explanation is not that equal splits are wrong. It is that a quick equal split is often a way of avoiding a hard conversation about who is bringing what, and that investors can read the avoided conversation in the team. Their estimate is that the stronger founder in a quick equal split leaves around ten per cent of the company on the table.

Hold both findings at once. Equal is now the norm for two founders and increasingly for three. An equal split arrived at by refusing to discuss contribution is a different thing from an equal split arrived at after discussing it.

The case for equal

Michael Seibel’s note for Y Combinator is the clearest statement of the equal position, and YC says it almost always recommends equal or near-equal splits. His argument has four parts. Building a valuable company takes seven to ten years, so a difference of a few months’ head start or one person’s idea is small against the work ahead. Most startups fail, so a larger share of a company that dies for want of a motivated co-founder is worth less than an equal share of one that lives. The split signals how the founders value each other, and investors read it: if you do not value your co-founder, neither will anyone else. And startups are about execution, not ideas, so the person who builds and ships deserves as much as the person who thought of it.

The reasons Seibel rejects are the ones most founders reach for: “I had the idea”, “I started three months earlier”, “I am older and more experienced”, “we need a tie-breaker”. None of these is worth equity. A tie-breaker is a governance question, solved by naming the CEO and giving the board one seat before the first large raise, not by a 51–49 split that resolves nothing in practice because no decision that matters is ever taken by a one-point majority between two people who have to work together the next morning.

The four reasons that justify unequal

Equal is the starting point, not the answer. Four contributions differ enough between founders to move the split, and they are the only four.

Cash. A founder who puts ₹20 lakh of their own money into the company has taken a risk the other has not. The cleanest treatment is to account for it separately: the money comes in as a loan or as shares bought at a price, and the split of the remaining founder equity is decided on the other factors. Mixing cash into the split means neither number is clean later.

Time. One founder full-time and unpaid while the other keeps a salary elsewhere is the most common real asymmetry in India, where the second founder often cannot leave a job until the company can pay. The full-time founder is taking the larger risk and forgoing the larger income. A difference of ten to twenty points is defensible while the asymmetry lasts, and the agreement should say when the part-time founder goes full-time and what happens to the split if they do not.

An asset. Not the idea. A working product, a signed customer, a patent application, a dataset, a distribution relationship that exists today and is assigned to the company in writing. An asset has a value that can be argued about with numbers. An idea does not.

Future role. If one founder will be CEO and carry fundraising, hiring and the board for a decade, and the other will lead a function that will eventually be led by a hired executive, the roles are not symmetric and a modest difference reflects it. This is the weakest of the four because roles change, and it should be the smallest adjustment.

A practical method: each founder scores the other, privately, from zero to three on each of the four, then compare. Convert the totals into a ratio, then round towards equal, because the research says founders over-weight their own contribution and because the cost of being slightly generous is small against the cost of a resentful partner. Two founders who score 10 and 8 are at 55–45 before rounding; most should settle at 52–48 or equal and treat the cash separately. The scoring is not the decision. It is a way of making the conversation specific enough to have.

What the split is worth after the money

Every founder stake is diluted by the same factor at every round, so a fight over whether the split is 55–45 or 50–50 is a fight over five points of the company today and about three points after a typical seed. Set the figure below to your own numbers. The [dilution lesson](/library/dilution-a-cap-table-you-can-touch) carries the mechanics round by round; here the question is only what the founders’ argument is actually about.

Carta’s Founder Ownership Report puts the median founding team at 56.2 per cent of the company after a seed round, 36.1 per cent after Series A and 23 per cent after Series B. On those medians a two-founder 55–45 split is worth 31 and 25 points at seed and about 20 and 16 at Series A. The five points the founders argued about are three, then two. If the company is worth ₹100 crore at Series A, that is ₹2 crore of difference between the two founders, against ₹36 crore they hold together. Worth being fair about. Not worth a partnership. Wasserman’s wider finding sits behind this: across 212 startups, the founders who gave up more equity, to co-founders, hires and investors, built more valuable companies than those who held on.

Decide the split as if the company will be worth a thousand crore, because that is the only case in which the split matters, and in that case you will want your co-founder to still be there.

Documenting it so it survives success

A split that is fair on the day and undocumented is unfair in eighteen months, because memories of who did what diverge faster than cap tables. Four things make it durable.

Vesting for every founder, four years with a one-year cliff, monthly after. Without it the split is a gift that cannot be recovered when a founder leaves; with it the split is earned, and the discussion about fairness has a mechanism rather than a grievance. The [vesting lesson](/library/vesting-and-the-cofounder-cliff) covers the schedule and how it is done in reverse for shares that already exist.

The reasons, in writing. One paragraph in the founders’ agreement stating why the split is what it is: cash contributed and on what terms, who is full-time from which date, which asset was assigned. A founder in year three who thinks the split was unfair can be shown the paragraph they signed. Without it there is only the argument.

A revisit trigger, not a revisit habit. Name the one event that reopens the split: the part-time founder failing to go full-time by the agreed date is the usual one. Everything else is closed. A split that is renegotiated every time someone feels under-appreciated is not a split.

The split in the right documents. In an Indian private limited company the founders’ shares are issued at incorporation, so the split is written into the subscriber pages of the memorandum of association and into the share register, and changing it afterwards is a transfer of shares with its own paperwork and stamp duty. Decide before SPICe+ is filed, not after. The [incorporation lesson](/library/incorporating-in-india-pvt-ltd-llp-or-opc) has the sequence.

The decision, in one sitting

Run it after the [forty questions](/library/choosing-a-cofounder-forty-questions) and before the lawyer. Each founder writes their scores on the four factors for the other, privately. Compare. Agree how cash is treated, separately. Agree the full-time dates. Convert the scores to a ratio and round towards equal. Write the paragraph of reasons. Agree the one revisit trigger. Then put the result, the vesting and the reasons into the [co-founders’ agreement](/library/cofounders-agreement-what-it-must-contain) and do not open the question again unless the trigger fires. Ninety minutes, once, is the whole cost of a decision most teams pay for over years.

Nothing here is legal, tax or investment advice. The data describes what founders do; the method is a way of deciding what you will.