Cofounder search mistakes: 8 habits that keep you from finding the right person

If you've been looking for a cofounder for months, it's tempting to conclude that the right person doesn't exist, or doesn't live near you. Sometimes that's true. More often the search itself is set up in a way that makes a good match hard to find, hard to recognise, or easy to lose.

This post is about those habits: the way people look, not the people they find. If you want warning signs in a potential cofounder, our cofounder red flags guide covers those. Here we look at eight mistakes in the search, how to tell if you're making each one, and what to do instead.

Why can't I find a cofounder? Start with what you're noticing

Each mistake leaves a mark you can see in your own search. Find the line that sounds like you and jump to that mistake.

1. Judging people on skills alone

What it looks like. You start with a role ("I need a technical cofounder") and judge every candidate on whether they can do it. The CV is right, the portfolio is good, and you move to the next step without knowing how they work, how much time they have or what they want from a company.

Why it costs you. Skills tell you who can do the job. The rest tells you who you can actually build with for years. Steve Blank, who was a cofounder of four companies, told First Round Review that selecting only for skills is "stopping short" (First Round Review). Gloria Lin, who spent about a year meeting potential cofounders before starting Siteline, described one of hers as "somebody who seemed great on paper, but we didn't mesh in real life" (First Round Review).

What to do instead. Before the next call, write down what you need beyond skills: hours a week, when you each want to go full-time, how you like to make decisions, what you'd count as success. Lin's advice is to keep the list of non-negotiables to three at most. Then ask about those things on the first call, not the fourth. Our guide to what to look for in a cofounder sets out the criteria and a one-page scorecard.

On Foundinity this part is built into the search. 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 on all of it, not on skills alone.

2. Browsing profiles instead of talking to people

What it looks like. You spend evenings reading profiles, saving the promising ones, and waiting until you find someone obviously right before you message anyone. Weeks pass. You've seen hundreds of people and spoken to three.

Why it costs you. Browsing feels like progress, but a profile can't tell you how someone handles a disagreement or a missed deadline. There's also evidence that more browsing makes people pickier rather than better at choosing. In three studies of online dating, Tila Pronk and Jaap Denissen found that people rejected more potential partners the more profiles they saw; the chance of accepting someone fell by 27% from the first profile to the last (Tilburg University). Dating isn't a cofounder search, but the habit of scrolling on in case someone better appears is the same. We wrote more about this in Tinder for cofounders?.

What to do instead. Set a weekly target for conversations, not profiles. Two or three first calls a week is a pace most people can keep alongside a job. Message people when they look promising rather than perfect; the call is where you find out.

Foundinity is set up for this. You see compatible people ranked for you rather than hundreds of random profiles, and once you match, a 10-day clock starts and there's a limited number of messages before a call is booked. If no call is scheduled in time, the match expires. The point is to get you talking.

3. Only asking your own network

What it looks like. You ask friends, former colleagues and classmates, and wait for an introduction. When none comes, you conclude there's nobody.

Why it costs you. People tend to start companies with people like themselves. A study of 816 nascent entrepreneurs in the US, using the Panel Study of Entrepreneurial Dynamics, found "very strong support" for homophily in founding teams, by gender, ethnicity and occupation, and no evidence that teams with mixed skills formed more often than chance would predict (Ruef, Aldrich and Carter, 2003). If your circle is mostly engineers, your network will mostly offer you engineers.

What to do instead. Do both. Tell your network clearly what you're looking for, as Lin did: "I told all my friends and closer connections that I was looking for co-founders and asked for recommendations and intros." Then add two places where you don't know anyone: a community in the field you're building for, an event where people build things together, or a platform where people are actively looking. Our guide on how to find a cofounder with no network covers how to start from zero.

4. Looking with no time limit

What it looks like. You're "always open to meeting people". Conversations with promising candidates drift for months, and nobody is ever a yes or a no.

Why it costs you. Without a date, there's no reason to decide, so you don't. Lin wrote that she "most appreciated when there was a clear go or no-go date". Blank's suggestion has a clock built in: start with something short, such as a hack weekend, then a provisional working period of at least 30 days before you commit.

What to do instead. Put dates on two levels:

• For the search. Review it every four weeks. How many first calls, second calls and trials did you have? If the numbers are flat, change where you're looking or what you're asking for.
• For each person. A second conversation within two weeks of the first, a trial project that runs two to six weeks, and a date on which you'll both say yes, no or extend once.

For how long each stage usually takes, see how long it takes to find a cofounder.

Don't confuse a deadline with rushing. Lin passes on advice she was given by an investor: it's common to take three to six months to hire an executive, so why give a cofounder less care? A date stops the drift; it doesn't mean picking someone by Friday.

5. Pitching the idea instead of getting to know the person

What it looks like. Your first message and first call are about the idea: the market, the plan, the product. The other person listens politely, says it sounds interesting, and you never hear from them again.

Why it costs you. A strong candidate isn't deciding whether your idea is good. They're deciding whether they want to spend years with you. Jessica Alter, writing on OnStartups in 2013, put it as "Fall in love with the person, not the idea" (OnStartups). Blank tells the story of his own failed pitch to an engineer he wanted: "I had told him about all the things he didn't care about." He won him over on the second attempt by talking about the technical problem the engineer cared about.

What to do instead. Spend most of the first call on them. Some example questions (illustrations you can adapt):

• "What made you start looking for a cofounder now?"
• "What kind of problem would you happily work on for five years?"
• "What did you like most about your last job, and what drove you mad?"

Then describe your idea briefly, as a problem you care about rather than a plan they'd be joining. If it's the idea itself you're nervous about sharing, our post on whether a cofounder will steal your idea suggests sharing it in stages.

6. Skipping the trial project

What it looks like. You've had three or four great calls. You like each other, you agree on most things, and it feels like time to make it official.

Why it costs you. Calls show you how someone talks. Work shows you how they behave when a deadline slips, when you disagree or when the dull tasks pile up. Blank's advice is to "give yourselves some time before you fully commit", starting with something short and then working together for at least 30 days. Palle Broe, writing for a16z speedrun about how 40 startups found their cofounders, adds a step many people skip: "Always, always, always make sure to speak with 2-4 of the cofounders former colleagues" (a16z speedrun).

What to do instead. Agree a small project with a clear goal and an end date: ten customer interviews and a summary, a landing page and a first test, a working prototype of one feature. Keep notes on decisions and disagreements, and check references in both directions. Our cofounder trial project guide has project ideas, a one-page brief and a sample four-week plan.

7. Waiting for a perfect idea before you look

What it looks like. You'll start looking once the idea is properly worked out, or once you've built a prototype on your own, or once you're sure it's the one.

Why it costs you. The idea you start with is rarely the one you build. Paul Graham wrote that "the initial idea is just a starting point", and that its main value is that you find your real idea while discovering what's wrong with it (Paul Graham). Waiting also costs you time with candidates. Alter argued that founders should start talking with potential partners and building side projects six months to a year before going full-time, and warned: "Waiting until the end leads to bad decisions and relationships that really don't work."

What to do instead. Start looking with a problem area rather than a finished plan: "I want to work on how small clinics handle bookings" is enough. The idea tends to get better once two people are working on it: Lin and her cofounder spent their first four months working out which direction to take before they committed. On Foundinity you can say whether you want to see people who have a specific idea, people who aren't set on one, or both.

8. Saying yes too fast out of relief

What it looks like. The search has been long and lonely. Someone finally says they're in, and you say yes before you've done the checks you'd planned.

Why it costs you. Relief is a bad reason to commit to years of work with someone. "You don't want to rush into a pairing or stop your search prematurely," Lin warns. Broe makes a related point about focus: "it's a mistake to focus on one person too early", because "spending time with 3-5 potential cofounders in parallel gives you contrast." With only one candidate, you have nothing to compare them to.

What to do instead.

• Keep two or three conversations going until a trial starts with someone.
• Before you say yes, finish the steps you set yourself: the questions, the trial, the reference calls.
• Sleep on it. A day or two between "I think this is the person" and "let's do it" costs nothing if it's right.
• Ask yourself whether you'd hire this person for a senior role on the evidence you have. If not, you don't have enough evidence yet.

The red flags guide is worth rereading at this stage; signs are easiest to explain away when you badly want a yes.

A 10-minute audit of your cofounder search

Answer these honestly. Each "no" points to one of the mistakes above.

• Can I write in two lines what I need in a cofounder beyond skills? (1)
• Did I have at least two real conversations in the last week? (2)
• Have I met someone this month from outside my usual circle? (3)
• Do I know the date I'll next review my search, and the date I'll decide on the person I'm talking to most? (4)
• On my last first call, did I spend more time listening than pitching? (5)
• Have I worked on something real with anyone I'm seriously considering? (6)
• Am I actively looking now, not waiting for the idea to be finished? (7)
• If I said yes tomorrow, would it be because of evidence rather than relief? (8)

What a healthy week of searching looks like

Here's an example week for someone looking alongside a full-time job (an illustration, not a rule):