Every week, a founder tells me some version of the same sentence. My cofounder and I are fine. He would never cause me problems. We do not need to write anything down.
I want to take that sentence seriously, because it is almost always true. I litigated cofounder disputes for years, and here is the part nobody expects: I almost never met a villain. My files were not full of con artists and traitors. They were full of good people who had been right about each other, and wrong about the future.
So this essay is not going to tell you to trust your cofounder less. It is going to tell you what actually breaks founding teams, and why the agreement you think you do not need is really a gift to the two people you are today.
The case I saw a hundred times
Let me tell you a story. It is a composite of many real ones, which is exactly the point.
Two friends start building. Evenings, weekends, a shared repo, a product taking shape. No company yet. Fourteen months in, one of them gets an offer he cannot refuse. Rent is real, life is real, the startup pays nothing. He takes the job. There is no fight. There is a long dinner, a hug, good luck to both.
The other one keeps going. Finds a new partner. Rebuilds half the product, keeps the half that worked. Two years later, the company exists, raises a seed round, gets a little press.
And then the email arrives. The one that starts warmly and ends with a sentence like: we never really settled my part in this. I wrote most of the early code. I think we should talk about my share.
Here is what I need you to understand about that email. The person who sent it is not lying. He did write that code. His commits are in the history. His ideas are in the product. His claim is sincere, and parts of it are provable. He is not a bad person who finally revealed himself. He is the same good person, in different circumstances, looking at the same facts from a different life.
Now the company has a problem that has nothing to do with anyone's honesty. Investors' counsel finds an early contributor with no signed agreement, no assigned IP, no settled equity. The round slows down. Lawyers get involved. Two people who genuinely liked each other start writing to each other through counsel. I have watched this exact sequence eat eighteen months and six figures. Nobody lied at any point.
Conflict does not need a villain
That is the truth the trust objection misses. You are not signing an agreement because your cofounder might be dishonest. You are signing because both of you are human, and humans change with circumstances.
Ambitions change. Family situations change. Financial pressure changes. Health changes. The market changes what the project is worth, and value has a way of clarifying opinions. Memory does the rest: two honest people can sincerely remember the same handshake differently three years later, especially when one of them stayed through the desert and the other one left before the rain.
None of this is a character flaw. It is the most human thing in the world. The mistake is building something valuable on the assumption that two people will feel in five years exactly what they feel today.
What the written agreement actually does
Written down early, the answers are cheap. Who owns the work. What happens to equity when someone leaves. What is earned by time and contribution, and what is forfeited by leaving early. Who keeps the domain, the repo, the accounts. What a departing founder can expect, and what the company can expect from them.
Answered while everyone trusts each other, these questions take an afternoon and cost nothing but honesty. Answered later, through lawyers, they cost the company its momentum and the two founders their friendship.
And notice who the agreement protects in my story. The founder who stayed, obviously: his company survives due diligence, his round closes, his answer to that email is two lines and a document instead of two years of proceedings. But it protects the leaver just as much. His clean exit is recorded. What he earned, he keeps, without begging and without ambiguity. He never has to become the person who sends that email, because the answer existed before the question. A good agreement is how both founders stay good people in each other's story.
There is one more thing it protects, and I would argue it is the most important one: the thing itself. The product, the value, the months of work. A startup can survive a departure. What it struggles to survive is an unresolved departure that resurfaces at the worst possible moment, when there is finally something worth fighting over.
Trust is the window, not the reason to skip it
Here is the reframe I wish someone gave every founding team on day one. The fact that you trust each other completely is not the reason to skip the agreement. It is the only reason you can still sign one.
An agreement signed in the good times records what generous, aligned, optimistic people believe is fair. That is the version of you both that you want deciding the hard questions. Wait until circumstances change, and the same questions will be answered by two tired people with opposite interests, which is to say, they will not be answered at all. They will be argued.
You do not sign because you doubt your cofounder. You sign because neither of you can promise who you will be in three years, and because what you are building deserves better than that bet.
So sign it while it is easy. Sign it while you trust each other. That is precisely what the trust is for.