The Equal Split Is a Signal, Not a Scorecard
An equal founder-equity split is the first internal financial deal of a startup: before outside investors and employees enter the cap table, the founding team decides how much of the company each member will own. The question in this series is not whether 50/50 or 25/25/25/25 is morally fair; it is what that equal outcome can reveal about bargaining, information, relationships, and control—and what it cannot tell us about contribution, valuation, or later ownership.
The first deal is an internal allocation
At formation, founders are not dividing money that already exists. They are dividing claims on an uncertain future: future profits, future sale proceeds, voting influence, and the ability to shape decisions. An equal split means each founder receives the same percentage of the founder shares. An unequal split means the percentages differ. Neither label, by itself, explains why the team chose it.
The detailed study by Thomas Hellmann and Noam Wasserman makes this first allocation visible through a measure called the share premium. If a venture has N founders, the equal benchmark is 1/N. A founder’s share premium is the actual percentage they receive minus that benchmark. In a two-founder team, someone receiving 60% has a 10-percentage-point premium over the equal 50% benchmark, while the other founder has a 10-point shortfall. The measure does not claim to capture a person’s total worth. It records how a team priced relative difference in the first deal.
That distinction matters because the first deal can be easy to ignore precisely when it feels least urgent. A new company may have no revenue, no stable job descriptions, and no reliable forecast of who will carry the business through its next stage. Founders still have to decide who owns what. They can treat the equal benchmark as a way to avoid a difficult judgment, or they can spend time arguing over the differences they believe will matter. The percentage is therefore also evidence about how the team handled uncertainty.
The paper’s examples make the contrast concrete. At Pandora Radio, then called Savage Beast, three founders with few obvious differences in expected contributions split the equity quickly and equally. At Ockham Technologies, substantial differences in the founders’ backgrounds led to detailed negotiation and an unequal result. These examples do not establish that one approach is always better. They show why the same cap-table pattern cannot be interpreted without knowing something about the team and the process that produced it.
The evidence base itself also teaches caution. In the Hellmann–Wasserman sample, 171 of 511 teams—33.5%—split equally, while 340 split unequally. That is a meaningful share, but it is not a natural law of founding. The observations come from private technology and life-sciences ventures in the 2008 and 2009 CompStudy surveys. The annual response rate was reported as 10% to 20%, and the researchers could not observe teams before they formed. The results therefore describe the choices of teams that entered this selected sample, not every group that ever considered starting a company.
The study’s scope is also about the initial allocation. It does not track every later change in founder ownership, and its vesting measure does not contain the schedules or conditions that would tell us how a founder’s claim might be earned over time. That means “equal at the start” and “equal for the life of the company” are different statements. A team can begin with two founders at 50% each, remain equal relative to one another, and still own a smaller share of the company together after investors receive new shares and an employee pool is created. Equality between founders can persist while founder ownership of the whole company falls.
This is the denominator problem in plain form. One comparison asks whether founders received the same share as one another. Another asks what fraction of the entire, fully diluted company the founding team retains after financing and hiring. Both are legitimate questions, but they answer different questions. A later percentage cannot tell us whether the original founders negotiated carefully, avoided negotiation, or chose equality for governance reasons. It only tells us where ownership stands after additional claims have entered the system.
For a reader looking at a cap table, the first useful move is therefore not to praise or criticize the split. It is to identify the moment and the pool being measured. Is the number the founders’ initial allocation, the founders’ combined stake after a round, or one founder’s share of the remaining founder pool? Is the comparison within a two-person team or across a larger group? The answer changes what an equal percentage can reasonably signal before any judgment about fairness or business quality begins.
Equal ownership is not a contribution score
The most tempting reading of an equal split is also the least justified: everyone must have contributed the same amount. That conclusion assumes the founders could observe and agree on the relevant contributions, that they valued them on the same time horizon, and that they were willing to turn those judgments into a public internal ranking. A founding team may satisfy none of those conditions.
A quick equal split might mean the founders saw no meaningful differences. It might instead mean that differences were hard to measure, that the cost of discussing them threatened trust, or that the team preferred a simple rule to an uncertain forecast. An equal split reached after extended discussion could mean something else again: the founders considered differentiated ownership and still decided that balanced claims, control, or cohesion were worth preserving.
This is why equality should be treated as a signal, not a scorecard. It may tell us that a team accepted a common focal point, but it does not reveal whether the team’s underlying contributions were identical. It may signal a relationship in which explicit valuation feels socially costly, but it does not prove that the relationship is healthy. It may reflect a preference for balanced control, but it does not establish that titles, voting rights, pay, and future effort are balanced too.
The distinction between the first allocation and its interpretation can be mapped simply:
The study’s evidence supports this caution. In its proprietary survey of 1,476 founders across 511 private ventures, about one-third of founding teams split their equity equally. Among teams that did not split equally, several observable differences were associated with higher share premia: prior entrepreneurial experience, being the idea person, and investing more founding capital. Founder roles also mattered; CEO, chair, and CTO positions carried positive premia in the reported models. Those findings do not produce a universal allocation formula. They show that unequal teams were often pricing recognizable differences that an equal split left unexpressed.
A later cap table contains new events
The first internal split is consequential, but it is not the ownership state a founder will necessarily see after fundraising. Carta’s 2026 Founder Ownership Report describes fully diluted ownership after rounds raised from 2021 through 2025. It reports that the median founding team retains about 56% at the seed stage and 36% by Series A. At Series A, the median digital-industry founding team retains 37.5%, compared with 30.5% in physical industries. At Series B, the median AI founding team retains 27.3%, compared with 21.8% for non-AI teams.
These numbers describe dilution and business context, not the original agreement. New investors receive shares, and companies often reserve equity for employees. By Series C, Carta reports that the median employee equity pool, at 16.8%, exceeds median founder ownership, at 16.1%. The founding team’s initial equality may still matter, but it now sits inside a different ownership system shaped by capital raised, hiring, valuation, and sector.
Team composition changes the meaning of the comparison as well. Carta reports that solo founders accounted for about 36% of startups founded on its platform in 2025, up from 31% in 2024. Among startups that successfully closed venture rounds, two-founder teams were the most common, representing 36% overall and 40% in SaaS. For three-founder teams, the reported rate of equal splits rose to 27.3% in 2025 from 21% the previous year; for four-founder teams, it rose to 16.7% from 10.8%.
Those figures have different denominators and come from a different period than the Hellmann–Wasserman survey. They should not be read as one continuous dataset. They do, however, establish an important context: “equal split” is not a single social situation. Equality between two founders has different voting consequences from equality among four. A rate measured among all platform-founded companies is not interchangeable with a rate measured among companies that closed a financing round. A founder-ownership median at Series A is not evidence about how the shares were divided at birth.
The question an equal split leaves open
A careful reading begins with the actual allocation and then asks what surrounded it. How many founders were there? Were differences in idea generation, prior experience, capital, or operating roles visible? Did the team decide in a day, or after a long negotiation? Were control and compensation being balanced along with equity? What financing stage and sector explain the ownership percentage being discussed?
Until those questions are answered, an equal split supports a modest conclusion: the founders accepted equal ownership under their particular information, relationship, and bargaining conditions. It does not establish equal contribution, guarantee equal future effort, or predict a company’s quality. It does not cause the later dilution shown in a financing-stage cap table, and the later cap table cannot be used to reconstruct the original bargain without additional evidence.
The unresolved issue is therefore not whether equality is good or bad. It is why a team chose it, what differences the choice may have left unpriced, and which later events changed the value and control attached to the original shares. Understanding that requires looking at the bargaining costs behind the equal benchmark before judging the number itself.