The Founder DNA Study
The last 400 U.S. tech exits over 20 years tell a story. The unconventional finding: co-founders who knew each other before consistently exit for less.
In the venture capital world, we spend an enormous amount of time talking about “Founder-Market Fit.” We analyze domain expertise, technical chops, and past exits. But these are individual metrics. They tell us how well a single player can hit the ball, but they say nothing about how the team plays together.
At Outcast Ventures, we’ve observed that the trajectory of the most successful companies is often locked in during the formation phase. This is the stage where the founding team is still debating the fundamental physics of their problem space. It’s the moment when “who” outweighs “what.”
“The goal of co-founder selection isn’t comfort; it’s complementarity so sharp it cuts through the noise of an entire industry.”
The Dataset
We analyzed the last 400 U.S. tech exits over the past 20 years, specifically looking at unicorns and decacorns. We categorized founding teams into three buckets: “Roommates” (long-term friends), “Co-workers” (professional colleagues), and “Strangers” (met specifically to start a company).
The conventional wisdom is that you should start a company with someone you know well. Someone you trust. Someone you’ve been in the trenches with. The data suggests otherwise.
The Unconventional Finding
Teams composed of “Strangers” — people who met with the explicit intention of founding a company — outperformed “Roommates” by a factor of 2.4x in exit value. They outperformed “Co-workers” by 1.8x.
Why? Because when you found with a friend, you are optimizing for social cohesion. When you found with a stranger, you are optimizing for skill complementarity and shared ambition. The social contract is different. It’s professional first, personal second.
Why Prior Relationships Underperform
The “Roommate” trap is subtle. You avoid hard conversations because you value the friendship. You tolerate mediocrity because you don’t want to rock the boat. You share the same blind spots because you come from the same background.
In contrast, teams formed in high-density environments like Catalyst or YC are forced to have hard conversations early. They have to diligence each other. They have to prove their worth. There is no assumed trust; trust is earned through execution.
The Optimal Co-Founder Distance
We found that the optimal co-founder distance is “weak ties.” People who run in similar circles but aren’t best friends. People who respect each other’s work but haven’t spent every weekend together for the last decade.
This distance allows for objective evaluation. It allows for constructive conflict. It allows the best idea to win, not the most popular one.
Implications for Founders
If you are looking for a co-founder, stop looking at your roommates. Stop looking at the person sitting next to you at work. Look for the person who is obsessed with the same problem you are, but comes at it from a completely different angle.
This is why we built Catalyst. To engineer these collisions. To bring together the best builders who haven’t met yet. To create the “Stranger” teams that the data says will win.
