How to Align Cofounders on Company Strategy
The symptom is a decision that will not stay decided. You agree on the plan, execute for six weeks, and then the same question reopens — usually triggered by new information that both of you interpret differently. That pattern does not mean you have a communication problem. It means you are aligned at the level of the plan and misaligned one or two levels beneath it, where nobody has looked. This page is a process for aligning at the right level, and for handling the residual disagreement that no process removes.
What’s inside
Alignment is not agreement
Three levels, and which one you are misaligned at
The alignment session
Writing it so it holds
Disagree and commit, for founders
The cadence
Where Blomma fits
Alignment is not agreement
A useful distinction, because founders chase the wrong one.
Agreement means both of you think the same thing is right. Alignment means both of you will execute the same thing regardless of what you privately think is optimal. You can have alignment without agreement, and that is usually what you actually need — two founders who agree on everything are either unusually lucky or one of them is not saying much.
Chasing agreement produces two failures. It makes decisions slow, because you keep discussing until someone concedes. And it makes them fragile, because a concession made to end a conversation is not a commitment — the moment new information arrives, the person who conceded reopens it, legitimately from their point of view.
What you want instead is a process that surfaces the real disagreement, resolves it explicitly, and produces a commitment both of you will hold even when you are not persuaded. That is a different objective and it is achievable.
Three levels, and which one you are misaligned at
Strategy has three layers. Misalignment at a lower layer looks like misalignment at the top, which is why so much strategy discussion goes in circles.
Beliefs.What you each think is true about the market, the customer, and the company’s advantage. “Buyers in this segment care more about integration than price.” “Our moat is distribution, not product.” These are usually implicit, rarely stated, and they determine everything above them.
Bets.What you are choosing to do given those beliefs, and what you are choosing not to do. “We win mid-market before we go enterprise.” “We build the platform before we chase the second product.”
Plans.The specific quarter’s activity — hiring, roadmap, spend.
The diagnostic: if the same plan-level question keeps reopening, you are misaligned at the belief level. Two founders who hold different beliefs about what the customer values will disagree about pricing forever, and each disagreement will look local. Resolving it at the plan level resolves nothing, because the generator is untouched.
So test the layers. Write down the five beliefs your strategy rests on, separately, and compare. Founders reliably discover one or two genuine differences they had never articulated — and that is where the recurring argument actually lives.
The alignment session
Half a day, twice a year, with structure. Open-ended strategy conversations between founders produce agreement about abstractions and no change in behaviour.
Prepare separately.Each of you writes: the five beliefs you think the strategy rests on; the three bets you think the company is making; and the two things you would change if it were your call alone. In writing, before you meet, without conferring — conferring first produces convergence on whoever spoke first.
Compare beliefs before anything else.Spend the first half on this. Where you differ, the useful question is not who is right but what evidence would settle it. Many founder belief differences are empirical and testable, and framing them that way converts an argument into a research task.
Then bets.With beliefs on the table, bets usually resolve quickly. Where they do not, name the trade-off explicitly: this bet makes X faster and Y slower, and we are accepting that because Z. If you cannot complete that sentence, you have not made a decision.
Name the anti-bets.What you are explicitly not doing this year. Founders skip this and it is where most of the value sits — an unnamed anti-bet gets quietly attempted by one founder within a quarter.
Close each item explicitly.Decided, decided-with-dissent, or open pending evidence. The middle category is the one most needed and least used.
Writing it so it holds
The output is one page. Longer documents do not get referenced, and a strategy nobody references does not constrain behaviour.
Include the beliefs, not just the conclusions. This is the part that makes a strategy document durable: when new information arrives, you can check whether it contradicts a belief or merely a plan. Without the beliefs written down, every new data point reopens the whole strategy.
Include the anti-bets, explicitly, as a list.
Include what would change your mind. For each major bet, one sentence: the evidence that would cause you to abandon it. This is the single most useful line in a founder strategy document, because it converts “should we reconsider this?” from a fight into a check against a pre-agreed condition.
And note where one of you dissented. Recording dissent seems counterproductive and does the opposite — it means the dissenting founder does not have to keep re-raising it to feel heard, which is the actual cause of decisions unravelling.
Then send it to your leadership team. A strategy your executives have not seen does not exist operationally.
Disagree and commit, for founders
The phrase is common and the founder version needs specifics, because between cofounders there is no authority gradient to make it work.
What commitment actually requires: you execute the decision as though it were yours, including in front of the team, including when it is going badly. You do not relitigate it in hallways, with your own reports, or through slow non-execution — which is the most common and most corrosive form of founder non-commitment, because it is deniable.
What the deciding founder owes in return: the dissent recorded, the reasoning explained rather than asserted, and a genuine review point. Commitment is much easier to give when there is a date at which the question legitimately reopens.
And both of you need a deadlock rule agreed in advance — domain tiebreak, CEO decides, reversibility test, or a named third party. Deciding the rule while calm is trivial. Deciding it during a deadlock is another argument, and it is the one that damages the relationship rather than the company.
One honest note: if you find you cannot commit to a particular decision — not will not, but cannot — that is worth taking seriously. Persistent inability to commit usually indicates a belief-level or vision-level difference that the process has not surfaced.
The cadence
Alignment is maintenance, not an event.
Twice a year, the half-day session above: beliefs, bets, anti-bets, from scratch.
Quarterly, a shorter check: has anything happened that contradicts a belief or triggers one of the “what would change my mind” conditions? Thirty minutes, and it prevents the drift that makes the biannual session feel like starting over.
Fortnightly, the standing founder conversation that is explicitly not status — for things that are drifting. Most strategic misalignment is detectable weeks before it produces an argument, and this is where it gets caught.
And after any significant new information — a competitor move, a fundraise, a big customer loss — an explicit ten-minute check on whether the strategy still holds, rather than letting each of you privately update in different directions.
Where Blomma fits
The hard part is not the process. It is that stating your beliefs plainly is exposing — you might be wrong, and you are certainly going to find out you disagree with your cofounder about something you assumed was settled. Most founders would rather keep having the recurring argument than run that risk.
Blomma is an always-on AI career coach with no view on your strategy. Use it to write your five beliefs honestly before the session, including the ones you suspect your cofounder does not share. Use it to work out which level you are actually misaligned at, which is the diagnosis that determines whether the session is useful or another circular conversation. Use it to draft the “what would change my mind” lines, which founders find surprisingly hard. And use it to prepare for genuinely committing to a decision you did not want — the part of disagree-and-commit that nobody rehearses and everybody struggles with. Where it warrants a human who has done strategy at your stage, bring one in..
Cofounders do not need to agree. They need to be aligned at the level of beliefs, explicit about the bets that follow, and willing to commit to decisions they lost. Get those three and strategy stops reopening every six weeks.