Which AI startup founders earn the most?

Last year, AI startup founders paid themselves 8% less than their peers. This year, they paid themselves 20% more. That’s quite a reversal and surprising for a number of reasons. 

For one, AI startups are more expensive to run. They have similar economics to SaaS companies except more costs: They pay for compute and data licensing. Whereas a SaaS startup might earn 80% gross margin, AI startups are at 60% right now. 

And two, because founder salaries were down across the board. Median pay declined 43% this year and 60% of all founders paid themselves less than $100k. But AI salaries were more resistant. They declined, but by less.

In this article, we explore the AI founder pay gap and why there is such a great range in pay depending on the type of AI startup.

There are now far more AI startups

Last year, 14% of survey respondents were in AI. This year, 40% were. That’s a 287% increase. And while you might think we’re seeing a new crop of freshly funded LLM-native startups who simply get by with less, you’d be wrong: The average startup in our dataset was seven years old. Many were started long before the pandemic and didn’t become AI until last year. These AI startups also had more headcount, on average, than non-AI startups.

This means the salary drop we’re witnessing is likely less of a generational reset and more of a deliberate pay cut. Same founders, less pay.

AI salaries fell this year, but not nearly as much as the broader market. The median AI founder paid themselves $90k, a 20% premium over the $75k for all founders. 

Yet as we explore, within that median there is a surprisingly wide range.

Variation among the sub-industries

Take a moment to take in the table below. Big data AI startup founders are earning nearly double what marketing AI startup founders are earning. I see a couple of reasons for this.

1. Cash-rich customers
I would suggest that founder salary is roughly correlated with the available cash among customers in the industry they serve. Founder salaries reflect the general opportunity in that market. As that opportunity and the startup’s revenue grows, so does their pay. Our report found that founders in growth mode earn 20% more than their peers, and pay rises with employee count.

This is my hypothesis, but industries at the bottom of this pay scale have all had a rough go over the past number of years. Marketing software is overcrowded, competitive, and marketing budgets are down. Commercial real estate remains depressed and aside from bright spots in multifamily in the sunbelt region, real estate companies are cutting costs. And pure SaaS has lost some of its vitality in the rush to fund AI startups.

It is possible that this great spread reflects the spread of opportunity in those markets, at least right now.

2. Data moats
Part of what makes an AI company alluring to investors is that it has a defensible moat. The best moat right now appears to be a proprietary dataset. 

Again, this is inexact, but once again, our lowest categories may be those where software companies commonly rely on third-party datasets and the cost of building your own and maintaining it is higher—and less easily defended. Marketing, real estate, cybersecurity, and telecom are places where you will struggle to build your own, and instead, pay to license. Whereas at the top of the pay scale, in big data, legacy software, consumer goods, and industrials, it is potentially more lucrative to build a proprietary data set moat.

How can you make the case for a higher salary?

When we asked all founders how they determined their salary, most said, “It’s what the startup can afford.” But AI startup founders are most likely to say something different: “Fair market rate.” Across all, those who paid themselves based on fair market rate earned 79% more. Perhaps that’s a good place for you to start in making your salary case.

But of course, this is not just a negotiation with the board. Many founders wrote in this year to explain that they took low or no salaries to fund operations. They have greater conviction in their equity play, and are taking out less now for the hopes of a more certain payout later. It is a tradeoff. 

For more salary insights, read this year’s report.