The co-founder state-of-the-union
The number-one thing that kills startups in the first two years isn't a bad product or a slow market. It's co-founders falling out. Most of them didn't have a language or a forum for the conversation that would have saved it.

The number-one failure cause for startups in the first eighteen to twenty-four months, per a number of VC data cuts (CB Insights, Concept Ventures), is not product. It's not market. It's that the co-founders fall out.
When I say this to founders, most nod politely and think 'that won't be us.' It will be you, probably, if you don't do the maintenance. Equity is locked, the cliff just vested, the mortgage is now co-signed with the relationship, and nobody has the language or the forum to say the thing that would have saved it twelve months earlier.
Why it happens in year two specifically
Year one is survival — fundraise, first hire, first customer. The work is so crushing that nobody has the cognitive room to notice the relationship is drifting. Year two is when the first real stress tests land: a bad hire you both kind of knew was bad; a term sheet that revealed different risk appetites; a month where revenue went backwards. Those tests don't break strong relationships. They break weak ones that nobody audited.
The quarterly state-of-the-union
Once a quarter, ninety minutes, no laptops, outside the office. Same five prompts every time. You don't need a coach. You need the prompts and the discipline to sit with the answers.
- What's the one decision we made this quarter that I'd make differently now — and what does that tell us about how we decide?
- Who carried more than their share this quarter, and who carried less?
- Name one thing the other person did this quarter that you haven't acknowledged yet. Say it now.
- Name one thing the other person did this quarter that's bothering you, that you haven't raised. Raise it now.
- If we knew right now that we'd have to separate within twelve months, what would we want to have said, done, or documented before that happened?
The last one is the hardest and the most valuable. It forces you to treat the partnership as a thing that exists on a clock, not as a given. Most co-founder pairs who make it through year three say the same thing: they started running this conversation, or one like it, before they thought they needed to.
What to do if it's already broken
If the conversation has been deferred for eighteen months, don't run the prompts cold. Get a third party — a board member, a trusted advisor, a therapist who works with founders. A broken partnership in a funded company is a workout session, not a workshop. Treat it with the severity of a financial restructure, because that's what it is.
A co-founder relationship isn't a vow, it's a contract. Contracts get renegotiated. The ones that don't get renegotiated get broken.
Found this useful?
Thirty minutes. Free. No pitch deck.
If the diagnosis is clear without me, you go do it. If not, we talk about the sprint. Either way, the first call takes 30 minutes and costs nothing.
Book the callKeep reading
All posts
Busy is the enemy of progress
A good week isn't measured in standups attended or features shipped. It's measured in hard decisions made. Founders who confuse activity for momentum burn runway without ever finding out what's true.

The week you should have stopped building
Pre-Series-A companies usually have either a distribution problem or a product problem. Rarely both. The one you don't have is the one you keep working on.

Distraction has a CRM — it's called LinkedIn
Graham was right that competition doesn't kill startups — distraction does. What he didn't say: distraction in 2026 has an interface, a notification schedule, and a search function. It's called LinkedIn.