Starting a business with a friend: should your friend be your cofounder?

You've been talking about the idea for weeks, and one evening your friend says it: "I'm in. Let's do this together." It feels like the obvious answer. You trust them, you like spending time with them, and you wouldn't have to pitch a stranger.

This post is for the moment before you say yes. It isn't about running a business with a friend you've already started with. It's about deciding whether this particular friend should be your cofounder at all: what the research says, what tends to go well and badly, the questions to answer first, how to test it in a few weeks, and what to do if the answer is no.

The short answer

• It can work, but friendship proves less than it feels. You know how your friend behaves at dinner, on holiday and in a crisis. You may not know how they behave under a deadline, in a disagreement about money, or on month nine of slow progress.
• The research points one way. In Noam Wasserman's data, teams of friends were the least stable and teams of past coworkers the most stable.
• The problem is usually avoidance. Friends skip the hard conversations to protect the friendship, and those are the conversations a company can't skip.
• Test it like any other partnership. A few weeks of real work together tells you more than ten years of friendship about whether you can build something together.
• Have a plan for no. Decide in advance how you'll keep the friendship if the answer is that you shouldn't be cofounders.

What the research says about starting a company with friends

The best-known evidence comes from Noam Wasserman, a professor who studied nearly 10,000 founders of technology and life sciences startups for his book *The Founder's Dilemmas*. Adam Grant summarised the findings in 2016: the least stable founding teams were friends, and each friendship in a founding team increased the rate of founder turnover by 28.6%. Even teams of virtual strangers were more likely to stay together. The most stable teams were past coworkers (Next Big Idea Club).

Grant's explanation is that friends get to know each other socially, while cofounders need to have seen each other work. Prior work together answers four questions that friendship doesn't: whether your values line up at work, whether your working habits fit, what each of you is good and bad at, and how you handle conflict. He also notes that friends tend to avoid difficult subjects to protect the relationship. He quotes Peter Thiel's view that founders "should share a prehistory before they start a company together", without which they're "just rolling dice". Grant's point is that the useful prehistory is working together, not socialising together.

There's research pointing the other way, and it's worth knowing because it explains why friend teams feel so good at the start. Karen Jehn and Priti Pradhan Shah ran an experiment with 53 three-person groups, published in 1997 in the *Journal of Personality and Social Psychology*. Groups of friends performed significantly better than groups of acquaintances on both decision-making and physical tasks, "because of a greater degree of group commitment and cooperation" (University of Minnesota).

Our reading of the two together: friends are good at short, shared tasks. A hackathon weekend or a first prototype with a friend can go brilliantly. What the friendship doesn't test is the long stretch afterwards, with its unequal workloads, money worries and disagreements nobody wants to raise. That's where Wasserman's friend teams came apart.

What goes well when your cofounder is a friend

None of this means friends make bad cofounders. Some things genuinely come easier.

• Trust from day one. You don't spend months wondering whether they're honest. You already know how they treat people.
• Fast, easy communication. You share shorthand, you can disagree without ceremony, and you can call each other at 11 pm without it being strange.
• Commitment to each other, not just the idea. The Jehn and Shah groups did better because friends committed to the group and cooperated. In a hard week, that matters.
• Enjoying the work. Gallup found that employees who say they have a best friend at work are seven times as likely to be engaged in their jobs (Gallup). That's a survey of employees, not founders, and it's a correlation, but liking the person you work with every day isn't a small thing.
• Known history in a crisis. If you've seen your friend handle a breakup, a lost job or an illness in the family, you know something real about their resilience.

What goes wrong when friends become cofounders

The ways it goes wrong are predictable, which is good news, because you can check for each one before you commit.

The hard conversation never happens. You both assume you want the same thing, so nobody asks how many hours a week, how big the company should get, or what happens if one of you wants to stop. The question surfaces a year later as a fight.

You chose them because they were there. The friend who said "I'm in" first isn't necessarily the person with the skills you need. Ask yourself honestly: if a stranger had the same skills, experience and availability, would you pick them?

Your skills overlap. Friends often share a background: the same course, the same industry, the same interests. Two people who are good at the same things leave the same gaps in the company.

Unequal effort, hidden by politeness. One of you works evenings and weekends; the other has a busy month, then another. With a stranger you'd say something. With a friend you wait, and resentment builds.

Roles blur. Who decides on the product? Who has the final word with a customer? Friends often leave this vague because formal roles feel awkward between equals.

You can lose both. If the company fails or one of you leaves, it rarely ends neatly. You may lose the business and the friendship in the same month.

Our cofounder red flags guide covers warning signs that apply to anyone, friend or stranger.

What you know about a friend, and what you don't

The useful question isn't "do I know this person?" It's "do I know how this person works?" Here's how the two compare.

If most of the right-hand column is blank for your friend, you don't know yet whether they'd be a good cofounder. That's fixable. The rest of this post is about filling it in.

Questions to answer before you decide

Some of these are for you to answer alone, some for the two of you together. Write the answers down.

About your friend

1. Have I ever seen their work, not just heard about it? What did it look like? 2. What skills does the company need in the first year, and which of those do they have that I don't? 3. When they've committed to something hard before (a course, a project, a training plan), did they finish? 4. How do they react when someone disagrees with them in front of others?

About you

5. Would I choose them if they weren't my friend? 6. Could I tell them their work isn't good enough? Could I hear it from them? 7. If this fails, would I blame them? Would they blame me?

For the two of you

8. How many hours a week can each of us give, starting when, and for how long? 9. How big do we each want this to get, and how soon? 10. Who decides what, and how do we break a tie? 11. What would make either of us want to stop? 12. If we decide not to work together, how do we stay friends?

Questions 8 to 12 are the ones friends tend to skip. If either of you finds them awkward to discuss now, that's useful information: they won't get easier once there's a company. For more to cover together, see our questions to ask a potential cofounder.

How to test working together before you decide

The best way to fill in the blanks is a short, real project with a fixed end date. Our cofounder trial project guide covers how to choose the project, set the scope and end it well. With a friend, add a few rules of your own:

• Agree out loud that no is allowed. Both of you say, before you start, that either can end it at the decision date without it affecting the friendship.
• Keep work and friendship in separate places. A separate chat for the project, set working hours, and a weekly check-in that's a meeting, not a chat over drinks.
• Pick work that will cause a disagreement. Choose something with at least one real decision in it, such as what to build first or which customers to talk to. You need to see how you disagree.
• Write things down. Who agreed to do what by when. Friends rely on memory and goodwill; a trial should test whether you both follow through without either.
• Each write your verdict before you talk. At the end, write down yes, no or extend separately, then compare. It stops one of you going along with the other to be kind.

Example plan (illustrative): Three weeks. Goal: 10 interviews with small gym owners and a one-page summary of what they'd pay for. Each of us runs five interviews; we write the summary together. Check-in every Monday at 7 pm, 30 minutes, project chat only. Decision on the last Sunday: each writes yes, no or extend before we compare.

Signs it's working, and signs it isn't

Good signs: you each did what you said without reminders; you disagreed at least once and settled it; you were honest about the parts that went badly; you both want to keep going and can say why.

Warning signs: one of you did most of the work and neither mentioned it; you avoided a disagreement rather than settling it; you kept slipping from work talk into social talk because work talk felt uncomfortable; one of you is saying yes mainly to avoid disappointing the other.

How to tell a friend you shouldn't be cofounders

If the answer is no, say it soon, say it directly, and keep it about the fit rather than about them. You might write something like this:

Example message (illustrative): "I've loved working on this with you these last few weeks, and I've thought hard about it. I don't think we should be cofounders. We work at different speeds, and I think we'd end up frustrated with each other in a way that would hurt our friendship. That matters more to me than the company. I'd really like your help in other ways if you're up for it, and I'd like to keep doing Thursday football like always."

If your friend is the one who says no, take it at face value. Some people turn down a friend precisely to protect the friendship, and that is a decision worth respecting.

Other ways a friend can help without being your cofounder

Being a cofounder is one role among several. A friend who isn't the right partner can still be very useful:

• First customer or tester. If they're in your target market, their honest feedback is worth a lot.
• Advisor in their field. A friend who knows sales, law, design or your industry can answer questions without being in the company day to day.
• Introducer. Ask who they know who might be the cofounder you need.
• Sounding board. Someone outside the company who knows you well and will tell you the truth.

Looking beyond your friends

If your friend isn't the one, or you'd like to compare before you decide, the people worth meeting are the ones who match how you work, and those are hard to find through your own circle. Wasserman's most stable teams were past coworkers because they'd already seen each other work. With someone new, you need to learn that early.

Foundinity is built around that. You state what you need in a cofounder, and answer an assessment about how you work: commitment, working style, ambition, personality, goals, values, alongside your skills and interests. A multi-trait algorithm built on established psychological compatibility models ranks the people you see, and every match comes with a compatibility breakdown showing where the two of you align and where you differ.

That breakdown gives you the questions to raise in your first call, so you get to the conversations friends tend to put off at the start. A 10-day clock on each match keeps things moving, and you book the intro call in the app. Your profile is visible only to compatible members. It's free to get started; Plus is $5 a week or $15 a month. See how it works.

If you're still unsure whether you need a cofounder at all, our post on solo founder vs cofounder is a good place to start.