How to Find a Co-Founder: Testing the Fit Before You Commit

On 19 June 1803, in Washington, Meriwether Lewis wrote to a former commanding officer he had not seen for seven years. President Jefferson had just given him command of an expedition that would cross the continent to the Pacific. Lewis did not want to go alone. He asked William Clark to share in its "fatigues, its dangers and its honors", adding that there was no man on earth with whom he would feel equal pleasure in sharing them.

The choice was not random. In the mid-1790s both men had served in the same army; Clark had been one of the young officer Lewis's commanders. They knew each other from the field, not from paper. With the President's knowledge, Lewis also made a promise: Clark would receive a captain's commission, and the command would be equal.

The promise was not kept. A week before departure, the commission from the War Department arrived making Clark only a lieutenant of artillery. Lewis wrote to Clark that the result was not what he had wished, or had reason to expect, and asked him to tell no one. Then he made his decision: the rank would be kept from the men, and Clark would be called "Captain" for the whole journey. The two captains led the expedition together from 14 May 1804 to 23 September 1806.

Years later, in 1811, Clark wrote that he did not think he had been very well treated, and asked that his disappointment not be mentioned. In the same lines he wrote that the two of them had been "equal in every point of view". The rank on paper was wrong; the partnership was right.

Why is finding a co-founder so hard?

People choosing what may be the longest relationship of their working lives often do it after a few coffees. That is why Gloria Lin's experience, told in First Round Review, is so useful. Lin had been the first head of product at Flipboard and the first product manager Stripe hired. She wanted to start a company but had no idea yet, so she looked for the right partner first. The search took about a year and six candidates, and she finally founded Siteline with Joel Poloney.

Her comparison is blunt: finding a co-founder feels like a more intense version of dating. From fights to marriage to divorce, the parallels keep coming. The difference is that leaving this marriage can bring the company down with it.

Five steps to finding the right co-founder

1. Tell everyone you are looking, and be careful with friends

Lin met people she had known for years, people introduced by mutual friends and people she met at events. She told her network plainly that she was looking and joined founder communities. Partnering with a close friend, she says, means going in "eyes wide open". Her source is Noam Wasserman, who wrote the book on the subject: partnerships built on social ties look stable in the first six months, the honeymoon period, but grow riskier over time.

2. Look for disagreement in the first conversation

First conversations are for finding shared interests, but Lin was really looking for the points where things would break. With one candidate it turned out that one of them wanted to bootstrap and the other wanted to raise as much money as possible. That ended it. Far better to see a gap like that over the second coffee than in the fourth month.

3. Work together for two weeks

Lin suggests timeboxing the exploration of each idea to about two weeks. For a consumer product, the cheapest possible prototype will do: a landing page or a small browser extension. For an enterprise idea, you run customer interviews. In those two weeks you test not just the idea but how you work together. According to Lin, most of her break-ups happened at exactly this stage.

4. Answer the hard questions separately

The questionnaire Lin used has 50 questions: working style, roles, fundraising and exits, personal motivation, team culture and how a split would be handled. Each founder answers alone, which takes two or three hours. Then you talk it through in person over several sessions, at least six to ten hours in all. Even a simple question reveals a lot: how many hours a week do you want to work? If one person loves investment banking hours and the other will not go past 50 hours a week, Lin says, a collision is coming.

5. Keep your non-negotiables few, and set a deadline

Lin advises keeping your list of non-negotiables to three items at most. For her, a technical background was table stakes; maturity, and someone who would make sound calls when she was not in the room, were musts. Communication style or management experience could flex for someone humble and willing to learn. When she and Joel answered the questions separately and compared, 95% of their answers were the same. But that was after four months of working together.

The one who shares the bread

The English word "companion" comes from Late Latin companio: com, "with", and panis, "bread". A companion is the person who breaks bread with you. The word first appears in the Frankish law code, the Lex Salica, most likely as a rendering of a Germanic word meaning "messmate". The Turkish word for partner, "ortak", shares its root with "orta", the middle: the owner of what stands between two people.

Both roots say the same thing. A partnership is not a title but a shared table. Lewis could not give Clark the rank, but he split everything in the middle equally: the decisions, the risk and the honour. The paper never said so, yet the men called them both by the same name.

Salt eaten together

In his Laelius, on friendship, Cicero recalls an old saying:

"Verumque illud est, quod dicitur, multos modios salis simul edendos esse, ut amicitiae munus expletum sit."

True is the saying that many measures of salt must be eaten together before the duty of friendship is fulfilled.

Before deciding on a co-founder, how much salt have you eaten together?

Once you have found your partner, the next question is how to divide the company: how co-founders should split equity is worth a read.