How to find a co-founder

Choosing a co-founder is closer to choosing a business partner for the next decade than to making a hire. You will share the decisions, the equity and most of the hard weeks, so how you work together matters more than any single skill either of you brings.

Start with people you have worked with

The strongest signal is having already built something together, at a job, in a lab or on a side project. You have seen how the other person handles pressure, disagreement and the dull parts of the work, and they have seen the same in you.

If nobody comes to mind, write down the best people you have worked with over the last ten years and ask each of them who they would start a company with. The answer is often someone you have never met, and the introduction arrives with a reference attached.

Where else to look

• Your extended network. Former colleagues of people you trust, people whose work you have followed, and people who have left a company you admire in the last year.
• Y Combinator’s co-founder matching. A free platform that YC opened to everyone through Startup School in July 2021. You describe yourself and what you are looking for, and it suggests profiles.
• Places where people build in public. Open source projects, research groups and hackathons, where you can see someone’s work before you meet them.
• People who know a lot of builders well. Investors, operators and talent firms. Zelcast introduces founders to people they could start a company with, when we think the two should meet.

Test the fit on real work

A few dinners will tell you whether you like someone. They will not tell you how the two of you work. Pick a real piece of the company, such as a prototype, ten customer calls or the first version of the pitch, and do it together over a few weeks. Watch how decisions get made, what happens when you disagree, and whether you both still want to keep going at the end.

Use the same weeks to talk about the things that tend to end partnerships: how much each of you needs to earn and for how long you can go without a salary, where you want to live, how hard you plan to work, what you would do if someone offered to buy the company in year two, and what each of you wants it to become.

What to agree on in the first month

• Roles. Who owns product, engineering, sales and fundraising, and who makes the call when you disagree.
• Equity. There is no standard split. Close to half of two-founder teams on Carta now split equally, up from about a third a decade ago. Whatever you choose, both of you should be able to explain it.
• Vesting. Four years with a one-year cliff is the common default, so a co-founder who leaves early does not keep a large share of the company.
• The 83(b) election. If you receive stock that vests, filing an 83(b) election with the IRS within 30 days lets you pay tax on its value now, which is usually close to nothing, rather than as it vests.
• What happens if someone leaves. Write it down while you still agree on everything.

Why it is worth the time

Noam Wasserman, who studied about ten thousand founders for The Founder’s Dilemmas, has said that people problems, tension between co-founders among them, account for roughly two thirds of failures in high-potential startups. Most of those problems are visible in a month of working together, and far harder to fix in year three.

Technical and non-technical co-founders

If you are a non-technical founder looking for an engineer, lead with the problem and the evidence you already have: customers, a waitlist, a market you know from the inside. Strong engineers hear a lot of pitches, and the ones that land show what the engineer would own and why the problem deserves years of their life.

If you are an engineer looking for someone to run the business side, look for a person who has sold something or built a market before, and test that on real work too. Enthusiasm for the idea is the easy part to find.