People imagine cofounder disputes start with a betrayal: stolen code, secret negotiations, a knife in the back. After years of litigating them, I can report that betrayal is the rare case. Conflicts do not start with betrayal. They start with a silence that compounds.
Five silences, to be precise. The same five triggers appeared in almost every file that crossed my desk, in every industry, at every level of founder sophistication. Here they are, each with the scene I watched replay and the sentence that would have defused it.
1. Asymmetric commitment, unspoken
The scene: one founder is full-time, the other is "transitioning out" of a job, indefinitely. Both are sincere. Neither has said out loud what they assume the other knows. Eight months later, the full-time founder is paying an invisible tax every morning, and the sentence arrives: he is not pulling his weight, and he owns half.
The defusing clause: a written time commitment with a transition trigger and consequences if it passes. Not because paper creates commitment, but because writing forces the assumption out of hiding while it is still cheap to discuss.
2. The unexamined split
The scene: equity divided in five minutes to avoid an awkward conversation, usually equally, sometimes on a napkin's logic that nobody can reconstruct later. The number is not the problem. The problem is that nobody can explain it, so as contributions diverge, each founder privately re-litigates it.
The defusing clause: a split whose reasons are written down next to the numbers. A split you can each explain, separately, in the same words, does not get re-litigated at midnight.
3. Silent drift
The scene: the most common file on my desk. Energy fades, replies slow, everyone notices, nobody speaks. By the time words are exchanged, they go through lawyers, because months of unspoken resentment have hardened into positions.
The defusing clause: vesting with a cliff, which converts "what does he deserve" into "what has he earned," plus the one non-legal tool that outperformed every clause I ever drafted: a recurring, honest founders' conversation put in the calendar before it is needed.
4. The first real money
The scene: the one outsiders find most cynical and insiders know best. The team survives the desert years fine. Then the term sheet arrives, or the acquisition interest, and every vague promise made in year one suddenly has a price. The forgotten "you'll get a few points" helper materializes. The departed cofounder's signature acquires market value. Money does not corrupt the relationship; it prices the ambiguities the relationship was carrying.
The defusing clause: all of them, honestly, which is the point of a founder agreement: it settles the promises while they are worth nothing, so that success has nothing to reprice.
5. Deadlock at a decision that matters
The scene: two equal founders, one real disagreement, no mechanism. Pivoting or persisting, taking the round or staying lean, firing the early hire or not. The company needs an answer by Friday; the structure cannot produce one. Weeks pass. The disagreement stops being about the decision and becomes about the relationship.
The defusing clause: a deadlock mechanism, agreed while you still agree on things. The next note is entirely about it.
The pattern behind the patterns
Notice what all five have in common: none requires a bad person. They require only normal humans, an unwritten assumption, and time. That is why "we trust each other" is not protection; the teams in my files trusted each other too, and the trust was real. What they lacked was not character. It was a one-sitting conversation, written down, before the silences started compounding. Nobody plans the war. They just skip the treaty.