AI is shrinking the cap table: 2026's smaller teams own more each

The startup cap table is getting shorter. In 2026, more than a third of new startups have a single founder, the median seed-stage company runs on about four employees, and average headcount at later stages has fallen sharply. Fewer people are building these companies, which means fewer people are splitting the ownership, and the ones who do join are getting bigger slices.

This is one of the clearest structural shifts in startups right now, and it runs against the story you’d expect. The intuition is that AI hollows out startup teams and the leftover workers get squeezed. The data says something more interesting: teams are smaller, but the equity isn’t evaporating. It’s concentrating.

The numbers behind the shrink

The 2026 Carta data tells a consistent story of smaller teams:

• About 36% of startups founded in 2025 had a single founder, up from 31% the year before. Solo founding is now the largest single category.
• The median seed-stage company has around four employees. These are real companies raising real money, running lean.
• Average Series D headcount has fallen 29% from its 2023 peak, down to roughly 131 employees. Even Series B teams shrank over the same period.

AI is the obvious driver. When a handful of people plus good tooling can do what used to take a department, you hire the department later, or never. The result is a company with the same ambitions and a fraction of the payroll.

Fewer people, bigger slices

Here’s the part that matters for ownership. When a company is built by fewer people, the equity that would have been spread across a big early team gets divided among a small one instead.

That shows up in the grants. Carta’s data shows that median equity grants for individual contributors have actually risen over the past two years, not fallen. The early engineer or operator who joins a four-person seed company isn’t getting a smaller piece because the company is lean; they’re getting a bigger one, because there are fewer people to share it with and their individual contribution is a larger share of the whole.

It even shows up at the top of the stack. By Series C, Carta found the median employee equity pool (around 16.8%) now slightly exceeds median founder ownership (around 16.1%). The people who build the company are, collectively, holding a meaningful share, and at that stage a bit more than the founders themselves.

The catch: concentration cuts both ways

Smaller teams owning more each sounds like an unambiguous win, and for the people who are in, it often is. But concentration has a shadow side worth naming.

When four people own a company instead of forty, every ownership decision carries more weight. Get the split wrong among a handful of people and there’s no large employee base to dilute the mistake across. A solo founder who brings on a first real partner is making a decision that will define the entire cap table, because those two people are the cap table. The stakes of “who gets what” go up precisely because there are fewer whos.

Solo founding sharpens this further. More than a third of founders are starting alone, which means the first equity conversation they have, whenever they add a co-founder or a key early hire, is disproportionately important. There’s no established split to slot someone into. They’re setting the precedent from scratch.

What lean teams should take from this

If you’re building with a small team in 2026, the shift has a practical implication: get the equity split right early, because there’s less room to fix it later. A few thoughts:

• A small team is not a reason to be casual about ownership. It’s the opposite. With fewer people, each person’s share is larger and each decision matters more. The 50/50 handshake between two people who “figure it out later” is riskier when those two people own the whole thing.
• Tie ownership to contribution, not headcount. In a lean company, contributions are wildly uneven and change fast. The person carrying the product this quarter may not be next quarter. A model that adjusts ownership to reflect who’s actually building, rather than a frozen split set on day one, fits small teams especially well. That’s the core of dynamic equity.
• The first hire’s grant is a big decision now. When your team is four people, employee number one might reasonably own a real chunk. That’s not a rounding error you can wing. See our guide to equity for your first hire for how to size it.

The bigger picture: the era of the sprawling early team spreading equity thin is fading, replaced by a smaller group each holding more. That rewards founders who get ownership right from the first split, not once the team grows, because in a lot of these companies, the team never does.

Frequently asked questions

How big is the average startup team in 2026?

According to Carta, the median seed-stage company has roughly four employees, and later-stage teams have shrunk too, with average Series D headcount down about 29% from its 2023 peak. AI tooling is a major driver of leaner teams.

Are startup employees getting less equity now?

No. Despite smaller teams, Carta’s data shows median individual equity grants have risen over the past two years. With fewer people to share ownership, the individuals who do join lean teams tend to get larger grants, not smaller ones.

Why does a smaller team change how I should split equity?

With fewer people, each person’s stake is larger and each ownership decision carries more weight, with no large employee base to absorb a mistake. That makes getting the split right early more important, and makes a contribution-based model that adjusts over time a better fit than a frozen day-one split.

How common are solo founders now?

About 36% of startups founded on Carta in 2025 had a single founder, up from 31% the prior year, making solo founding the largest single category. That raises the stakes of the first equity split whenever a solo founder adds a partner or key hire.