A two week protocol after a cofounder conflict turns one hard conversation into something you can both observe. You leave with one or two written requests, a mid-point check at day 7, and a close at day 14. Done is not repaired. Repaired is what you can still see after a fortnight of real work.
Most cofounder fights end with a vague truce. "We'll communicate better." "I'll loop you in." The feeling in the room improves for a day. Then the old pattern returns, and you both conclude the conversation failed. Often the conversation was fine. The follow-through was empty.
Why the window is two weeks
Repair needs a container with edges. Open-ended promises dissolve under sprint pressure. A fortnight is long enough for the trigger to recur — a missed handoff, a unilateral decision, a late reply — and short enough that neither of you can postpone the review forever.
Outlander VC's field guide on cofounder conflict stresses routine touchpoints and data over interpretations: conflict lives in the story you tell about the facts, not in the facts alone (Outlander). A two-week protocol forces the story into a testable ask. Either the observable action happened, or it did not.
Couples therapist Esther Perel, working with founders through First Round, draws a related line: name whether the wound is about power or about care, then make the next move specific enough to watch (First Round Review). Vague goodwill is not a protocol. A camera-visible request is.
What you write before the clock starts
Keep the output small. One or two agreements. More than two usually fail because attention splits.
Each request should fit this shape:
- Situation. When X happens (a trigger both of you can name).
- Action. One specific behavior, owned by one person.
- Window. For the next two weeks.
- Check. How you will know it happened (a message, a shared doc, a decision logged).
Example: "When a pricing change affects the roadmap, I will message you before I tell the team, for the next two weeks." Not: "I will include you more."
Embroker's guide to cofounder dilemmas makes the same case for written plans of action — roles, load, and a conflict game plan you can reopen when emotion is high (Embroker). The two-week sheet is that plan at the smallest useful scale.
Add three standing rules for the fortnight:
- One respect boundary (no interruptions, or behavior only — no character labels).
- A circuit breaker: either of you can say "pause," with a named return time.
- A re-entry norm: clear leftover tension within 24–48 hours in a short, structured talk.
Day 0: lock the container, then stop talking about it
Run the repair conversation in a scheduled block, not in Slack after a board call. Reboot's coaching practice for cofounder conflict puts it plainly: name the conflict directly, use structure, and ask for help when you are too inside it to move (Reboot).
Confirm the written agreements. Put them where both of you will see them. Then return to building. The protocol is the follow-through, not another week of processing.
If you need a guided path into that conversation — private sense-making first, then a joint picture, then the written asks — a Conflict Session is built for exactly that sequence. The protocol is what happens after the sheet locks.
Day 7: the mid-point check
Check in separately first if the topic is still charged, then compare notes. Fifteen minutes is enough.
Ask only three questions:
- Did the request happen when the trigger showed up?
- Was the request too vague or too heavy to keep?
- What one adjustment would make the second week cleaner?
Do not reopen the entire archive. One rupture, one experiment. Outlander's fact-versus-interpretation frame helps here: stick to what occurred, not the motive you assigned (Outlander).
If both of you kept the ask, say so out loud. Positive evidence is part of the repair. If one of you missed, tighten the language. "Loop me in" becomes "send the draft before the customer call."
Day 14: close, keep, or escalate
At the end of the fortnight, you need a decision — not a mood.
| Outcome | What it looks like | Next move |
|---|---|---|
| Hold | Trigger appeared; request kept most of the time | Promote the ask to a standing rule |
| Adjust | Intent was there; wording failed | Rewrite and run one more fortnight |
| Fail | Pattern unchanged or worsened | Escalate the process, not the volume of talk |
Escalate means a fuller repair structure, a coach or mediator, or an honest look at whether the partnership should continue. Our guide on repairing a broken cofounder partnership covers the longer path. If contempt or total withdrawal is already the climate, do not invent another soft restart — name the pattern and choose a harder process.
What the protocol is not
It is not punishment. It is not a performance review. It is not a substitute for equity or role documents when those are the real dispute. Embroker notes that chronic fighting often means roles were never defined (Embroker). If the conflict is structural, the two-week ask may only buy you clarity that you need a deeper reset — decision rights, a Blueprint, or a clean exit conversation.
It is also not endless. A protocol that keeps extending without a hold or a fail decision is avoidance wearing a calendar.
How this fits the rest of the partnership
One successful fortnight does not inoculate you. Reboot argues you have to get used to working things out, with standing time for hard topics (Reboot). Pair the protocol with a recurring cofounder check-in so the next friction surfaces early. For the broader playbook, start with cofounder conflict resolution on the cofounder conflict pillar.
The two week protocol after a cofounder conflict is simple on purpose. Write the ask. Watch the behavior. Meet at day 7. Decide at day 14. That is how a conversation becomes a change you can both trust.
