Start A Company With A Stranger

Alex Crompton
28 November, 2017

Conventional wisdom suggests you should start a company with one of your friends. The relationship between cofounders is crucial. Everyone knows this.

Except it’s false.


Ok, so most good companies were started by friends. But why?

It’s not because being friends improves your odds of success. It’s because the kind of people who start companies with strangers aren’t the kind of people who start good companies.

This is obvious once you think about it. Smart people have lots of options. So, if a stranger thinks starting a company with you is their best option, either they aren’t very smart, or they make bad choices.

Picking a cofounder from your friends just filters out dumb people with bad judgement. And this is important: dumb people with bad judgement tend not to make good co-founders.

That’s why good companies are started by friends: only bad founders start companies with strangers.

If pretty much every company started by strangers is bad, most good companies will be started by friends. But that doesn’t mean friendship is what makes good companies good.

It turns out what we mean by ‘start a company with your friend’ is ‘don’t start a company with strangers’.


Finding a co-founder today is a lot like dating in the ‘90s. Back then, ‘don’t date someone you meet online’ would have been good advice. Why?

Again, it’s obvious once you think about it. Attractive people have lots of options. And if you have options, you’re probably not an early adopter of dating websites.

The kind of people looking for partners online in 1999 weren’t the kind of people you wanted to date. So, in the early days of the internet, most attractive couples still met offline. ‘Don’t date someone you meet online’ was good advice.

But this is horrible advice in 2017. It turns out that meeting offline isn’t intrinsically better. Today, hundreds of millions of people meet online, and it’s great.

The pool of people you can choose from is huge, awesome, and willing. Relationships that couldn’t happen offline, happen. As surprising as it once seemed, these relationships are often extraordinary. Dating has improved.

Meeting offline isn’t what makes relationships good, just like being friends isn’t what makes companies good. What matters is whether you’re doing it with the right person.


In startups, outcomes trump friendship. It doesn’t actually matter if you’re friends at all.

If it goes well, starting a company is a unique experience. Succeeding is so hard that you can’t help but appreciate your co-founder if you do. They might start off your worst enemy but, if they’re the reason you succeed, they won’t stay that way.

The reverse is also true. If it goes badly, starting a company is horrible. Failing is so painful that you can’t help but resent your co-founder if you do. They might start off your best friend but, if you fail, they won’t stay that way.

The right person is whoever you can get farthest with, fastest. And that person might not be your friend, yet.

I’ve bet my career on this. At Entrepreneur First, we help you find a co-founder. We curate a pool of amazing people ready to start something too, and make it easy to try working with them. If it doesn’t work out, you just find someone new from the pool and start over.

I’ve seen thousands of teams start this way. Some succeeded, most failed. One surprising conclusion is that teams that are friends before they start are no more likely to succeed than teams that aren’t.

The hardest situations at EF are when good people get stuck in bad teams because they’re friends. Both the startup and the friendship end badly. Good teams become friends because they’re good. Outcomes trump friendship.


It still seems weird to start a company with someone you met at EF, just like it used to seem weird to date someone you met online.

But now online dating exists, I can’t believe people try to find a partner, drunk, in clubs and bars. And now EF exists, I can’t believe people try to find a cofounder, aimlessly, in universities and meetup groups.

Turns out, there is a better way. Bring all the good people into one place. Seems obvious in hindsight. And now we’re here, imagine going back to how things were before.

This has some interesting long run effects. Over time, what’s weird will change. Once there are good ways of meeting and filtering through people, partnering with someone the old way stops making sense.

Why would you expect to meet your perfect partner randomly, ‘in real life’? It’s absurd to think that the best possible match is someone you just happened to have bumped into already.

Starting a startup with a stranger used to mean you didn’t have many options. But, in the long run, starting a startup with your friend will be the last resort.

So, should you start a company with your friend? Well, only if you have to.

How useful was this post?

Click on a star to rate it!