The Startup Team Is More Than Its Founders
This series examines founding-team stability in young ventures: not only whether the original founders remain, but what happens when the broader group that formed the business changes. The central question is why continuity can become more valuable over time. The answer is not that every founder is irreplaceable or that a startup should avoid change. It is that people who build a venture together may also build relationships, routines, judgments, and organizational knowledge that are difficult to separate from the people who created them. Readers should come away able to distinguish individual talent from team-specific capability, understand what the evidence can actually establish, and judge when a personnel change risks taking accumulated organizational capital with it.
The idea can survive while the organization changes
A startup is often described as an idea attached to a founder. That description is useful at the moment of origin, but it becomes incomplete as soon as the venture begins operating. A product idea does not answer every customer question, negotiate every supplier relationship, decide which process to keep, or teach a new employee how this particular business makes decisions under pressure. Those activities are carried out by a group. In the early stages, the group is often small enough that its shared knowledge is carried informally, in conversations and habits rather than manuals.
That creates a practical puzzle. Suppose the venture keeps its name, product, and legal identity but loses one of the people who helped establish it. The visible assets may remain. Yet the firm may no longer have the same understanding of why a customer was won, which promise can safely be made, how a recurring problem is diagnosed, or which members can coordinate without explanation. A replacement may bring strong skills and still need time to reconstruct the relationships and context that the departing member carried. The question is not whether the individual was talented in isolation. It is whether the team had become part of how the organization worked.
The research subject here is therefore the founding or new-venture team, not a heroic founder considered alone. The U.S. Census study used in this series defines a founding team broadly: people with positive earnings at the firm in its first year, supplemented by owners of sole proprietorships. It separates likely key members from the remaining early team, but it does not assume that only the owner or top executive creates the venture’s capability. That choice matters. If the loss of a non-key early member also affects a firm’s trajectory, the relevant unit is a collective rather than a single celebrated individual.
The theme can be stated more precisely as a time problem. During formation, a team may be creating organization-specific capital: a shared business vision, customer and supplier relationships, reputation, norms, processes, and a business model. These are not all directly measured as one variable in the evidence. They are the kinds of assets the Census authors place in their organizational-capital model and use to interpret the persistent effects of team-member loss. If such capital is partly embodied in the people who developed it, then continuity can matter more after the venture has accumulated more history and interdependence. Losing someone later is not necessarily equivalent to losing the same number of labor hours at the beginning.
The proposed mechanism has several layers. First, information becomes contextual: a team learns not just what a customer wants, but why a particular compromise was made and which promises are safe to revisit. Second, coordination becomes complementary: members divide judgment in ways that work precisely because they have learned how the others think. Third, outside relationships may become attached to the people who built trust with customers, suppliers, and partners. A new colleague can eventually recreate parts of this capital, but the transition may carry a cost even when the replacement has strong general ability. These layers make the organizational-capital model intelligible; they are not separate variables that the studies directly measure.
This also explains why a founding team is not simply a list of early employees. The relevant question is not only who was on the payroll, but what the group learned and how much of that learning became embedded in the firm’s daily operation. A founder may hold a central vision, while an early employee may know the customer history or operating workaround that makes the vision executable. The Census study’s inclusive definition keeps that possibility visible. It also makes the claim more demanding: if the effect appeared only when a top owner left, founder importance would be an adequate explanation; effects beyond likely key members point toward a broader collective contribution.
This is why “stability” should not be reduced to keeping every person forever. It means continuity in the composition and working relationships that carry the venture’s accumulated capability. A team can change deliberately, add missing expertise, or replace a poor fit. Those are different situations from an abrupt loss that removes a person and the context they hold before the organization can transfer it. The evidence discussed later is strongest about forced separation and about the consequences of stability in particular research settings; it does not supply a universal rule against change.
What is being preserved?
Individual human capital is part of the story, but it is not the whole story. A person arrives with skills and experience. A team adds something produced through interaction: members learn one another’s strengths, develop shorthand, divide decisions, and accumulate knowledge about this firm’s customers and constraints. That shared layer may raise productivity even when no individual’s résumé changes. It may also make substitution imperfect. A new hire can possess comparable general skill without possessing the same history of decisions and relationships.
The distinction is important because a simple founder-centered account predicts that replacing a founder with an equally capable manager should largely restore performance. A team-capital account predicts a more complicated transition. The firm may retain its idea and hire competent people, but still lose speed, coordination, or relationship capital while the new arrangement forms. The effect can appear in more than one outcome: fewer employees, lower revenue, weaker productivity, or a greater chance of exit. A fall in headcount alone would be easy to explain as a vacancy. A persistent change in several outcomes would point to a deeper organizational disruption, even if the exact mix of mechanisms remains uncertain.
The descriptive evidence makes this a plausible question without settling it. In the Census database, which tracks more than six million startups and over 72 million founding-team members from 1990 to 2015, teams with higher prior earnings—a proxy for human capital—tend to be associated with better survival, scale, growth conditional on survival, and productivity. Team members also leave over time, and the remaining team tends to have higher average prior earnings because attrition is concentrated toward the lower end of the initial distribution. These patterns help show how founding teams differ as young firms develop. They do not show that stability caused the better outcomes. People with more human capital may choose better opportunities, and stronger firms may be better at retaining them.
That limitation is not a technical footnote. It changes the question the evidence must answer. Comparing a startup whose team stayed together with one whose team changed can confound team quality, business quality, industry, timing, and the firm’s ability to retain people. The studies therefore use different strategies to move beyond the easiest story. A Danish study of new ventures reports an instrumental-variable analysis using matched employer–employee data from 1981 to 1997. Its abstract says that stable new-venture teams have a 16.1 percentage-point higher likelihood of survival and a 20.4% increase in average size after ten years, with the effect of stability becoming stronger over time. Those findings directly motivate the title question, but the accessible material here is the abstract and university research record, not the article’s full methods and tables.
The U.S. study offers a more detailed test of the reverse situation: what happens when a founding-team member is separated from a young firm by a premature death? It matches firms with and without such a shock and follows their outcomes. The design is not a perfect experiment, and a death is not ordinary turnover. But it is stronger than a raw comparison of successful and unsuccessful startups because the researchers can examine whether the firms were moving similarly before the event and then trace what changes afterward. The later chapters will use that design to interpret the large and persistent effects it reports, rather than treating every association as proof.
A claim with boundaries
The most defensible opening claim is consequently narrower than “founders are everything.” Founding teams appear to matter because they combine individual capabilities with organization-specific capital. Stability may become more valuable as that capital accumulates, and separation can damage not only survival but also the performance of firms that continue operating. At the same time, the size of the effect depends on who leaves, how concentrated the team is, what outcome is measured, and how the study identifies the change. The Danish stability result and the U.S. separation result point in the same broad direction, but their percentages should not be merged: they concern different countries, periods, team definitions, designs, and time horizons.
This framing also leaves room for a genuine trade-off. A team may preserve valuable context while lacking a capability needed for the next stage. Retaining a person is not the same as retaining every asset the team has built, and adding someone is not automatically the same as destroying continuity. The practical judgment is to ask what the organization has learned together, how much of that knowledge is transferable, where the team has redundancy, and whether a proposed change is a planned capability upgrade or an avoidable loss of accumulated capital.
The next question is what, concretely, a founding team builds together—and how the evidence distinguishes that shared organizational capital from the human capital of the individuals who make up the team.