Resolving co-founder conflict
1. Core idea and who it's for#
Co-founders will disagree. That is part of having more than one founder. What kills companies is not disagreement but disagreement with no way to settle it: arguments that repeat, decisions that stall, and resentment that builds until someone leaves or the team splits into camps.
Good teams treat conflict as something to plan for, like a fundraising process. They:
- Name the type of conflict. Is it about a decision, about roles, about effort, about money or about values? Each needs a different fix.
- Use a ladder. Start with a structured conversation between the founders, then a written decision process, then a trusted third party, then a coach or mediator. Only after that does it become a legal process.
- Write the rules down early. Decision rights, deadlock rules and dispute steps go in the founders' agreement (3.07) before there is anything to fight about.
- Know when to stop repairing. Some conflicts end in a breakup (3.12). Recognising that early and handling it well is also a kind of resolution.
Who it's for: founding teams of two to four, especially in deeptech. Long timelines, technical and commercial founders with different instincts, and the pressure of milestone-based funding make conflict more likely and more expensive.
The one-line rule: decide how you will decide before you need to, and escalate early, not late.
2. Evidence#
The famous "65%" figure needs care. Noam Wasserman, who studied about 10,000 founders for The Founder's Dilemmas , is widely quoted as saying 65% of high-potential startups fail because of co-founder conflict. Some summaries of the book phrase it more broadly, as problems within the management team. Either way, the lesson is the same: people problems are a leading cause of failure. Use the number as a warning, not a precise statistic.
Co-founder departures are common. Carta data shows that more than a quarter of twofounder teams lose a co-founder by year four (see 1.18 and 3.01). Planning for that is ordinary risk management.
Decision-making is a frequent trigger. Summaries of Wasserman's work report that disagreements about how decisions get made account for about 20% of co-founder conflict, second only to other causes such as roles and money. Teams that agree who owns which decision reduce this.
"Disagree and commit" works only with an owner. Founders interviewed by Stanford eCorner describe disagree-and-commit as needing one decision-maker when the founders can't agree. Initialized Capital's founder essay makes the same point: when there is an owner, the owner makes the call and everyone commits. It also warns that rapid growth can hide conflict until it breaks out.
Outside help is now normal. Accelerator StartupYard says co-founder coaching and founder therapy are no longer taboo but practical tools. It also says the fatal pattern is silence and avoidance, not conflict itself. Some lawyers and mediators offer "pre- mediation": a facilitated talk about roles, equity and commitment before a dispute starts.
Lessons from couples research apply. Kevin Hale of YC has drawn on Gottman's research on couples, which shows that how an argument is handled predicts whether a relationship lasts. Co-founders argue about the same "in-law" topics: money, time, roles and competing commitments.
3. Method#
Step 1: Diagnose the type of conflict
| Type | Sounds like | Usual fix |
|---|---|---|
| Decision | "We keep going round in circles on pricing." |
Clarify who owns the decision. Set a deadline and a tie-break rule. |
| Role | "You keep redoing my work." | Redraw the areas each founder owns (3.07, 3.08). |
| Effort | "I'm doing 70 hours and you're doing 30." |
Make expectations explicit and check commitment (3.10). Vesting is the backstop. |
| Money and equity |
"The split doesn't feel fair any more." |
Talk about the facts, and consider a measured change, such as an extra grant or changed vesting. Don't reopen everything. |
| Values and direction |
"You want a lifestyle business and I want a unicorn." |
A deep conversation, often with outside help. May end in a split. |
| Type | Sounds like | Usual fix |
|---|---|---|
| Personal | "I can't stand how you talk to me." | Feedback norms, coaching, and sometimes therapy. |
Most serious fights mix several types. Name each one separately.
Step 2: Hold a weekly founders' meeting#
Set aside 45 to 60 minutes a week with no team present. Go through:
- What's going well.
- One thing that's bothering each of us.
- Decisions that are stuck.
- A quick check on energy and commitment.
Every quarter, run a longer review of roles, goals and equity concerns. Problems raised in small doses rarely become crises.
Step 3: Climb the escalation ladder#
- Direct conversation. Use a set format: each person states the facts, what they think it means, how they feel and what they want. The other person repeats it back before replying.
- Written options. Each founder writes down the options, their preferred one and the cost of being wrong. Often the disagreement turns out to be smaller than it felt.
- Owner decides. If the decision sits in one founder's area, they decide after hearing the other, and both commit. Set a review date.
- Agreed tie-break. For shared decisions, use the rule in the founders' agreement. Options include the CEO deciding, a trusted adviser or independent director deciding, or a test (such as a customer pilot) settling it.
- Coach or facilitator. For repeated or personal conflict, bring in a co-founder coach for 3 to 6 sessions. Agree who pays (usually the company) and that the conversations are confidential.
- Formal mediation. Use a mediator when the relationship, roles or equity are at stake. Mediation is cheaper and faster than lawyers, and it is private.
- Separation or legal process. If none of the above works, move to an orderly breakup (3.12) under the shareholders' agreement. Arbitration or court is the last resort.
Step 4: Put deadlock rules in writing#
Cover these in the founders' agreement or shareholders' agreement:
- Who decides each type of decision.
- What counts as a reserved matter needing all founders.
- The tie-break for those matters.
- A requirement to try mediation before arbitration or court.
- What happens if a deadlock lasts, for example more than 60 days.
Companies with a 50/50 split and two directors need this most.
Step 5: Look after the team#
The team notices founder conflict. Agree what you will say to the team and investors. Never ask employees to take sides. Present decisions together once they're made.
Worked example (illustrative)#
The two founders of a medical-device startup, a CEO and a CTO, split 50/50, disagree about whether to file for FDA clearance on a narrow indication now or wait 9 months for a broader one. After three weeks of arguing:
- Diagnosis: a decision conflict (regulatory strategy), with a role conflict underneath. The CTO feels the CEO is overruling technical judgement.
- Written options: each writes one page on cost, timing, risk and how it affects the seed raise. The real gap turns out to be how risky the bench data is.
- Test: they pay a regulatory consultant to review the data for 2 days (cost about $4,000). The consultant backs the narrow filing, with one extra test.
- Owner and commit: under their agreement, regulatory strategy belongs to the CEO, with the CTO owning the technical file. The CEO decides on the narrow filing, and the CTO agrees and commits. They agree to look at the broader indication again after the first clearance.
- Follow-up: they start a weekly founders' meeting and book 4 coaching sessions on how they argue. Six months later, the role conflict has not come back.
4. India, US and UK differences#
The early rungs of the ladder (talking, written options, coaching, mediation) work the same everywhere. The legal backstops differ.
India#
- Mediation: the Mediation Act 2023 gives a framework for mediated settlements. For most commercial lawsuits, the Commercial Courts Act already requires preinstitution mediation before a suit can be filed, unless urgent relief is needed. Writing
- a mediation step into your shareholders' agreement fits this.
- Arbitration: shareholders' agreements usually choose arbitration under the Arbitration and Conciliation Act 1996, seated in Mumbai, Delhi or Bengaluru. Where there are foreign investors, they often choose SIAC in Singapore. Arbitration is private, but it can be slow and costly for a small company.
- Company law remedy: a shareholder can petition the NCLT for oppression and mismanagement under sections 241 to 242 of the Companies Act 2013. This usually needs 10% of shares, although the tribunal can waive that. It is a heavy, public process, used mostly when one founder has been pushed out.
- Practical point: many Indian startups have two founder-directors with equal votes. Add a chair's casting vote or an independent director, or the board can deadlock completely.
US#
- Governing law: most venture-backed companies are Delaware corporations, and founder disputes often end up in the Delaware Court of Chancery, which is experienced and fast by court standards. Section 226 of the Delaware General Corporation Law lets the court appoint a custodian when shareholders or directors are deadlocked. This is a last resort.
- Arbitration: founder agreements and employment agreements often choose AAA or JAMS arbitration. Check that the arbitration clause doesn't conflict with the choice of court in the company documents.
- Board role: once there is an investor director, the board often becomes the place where the tie-break happens. Agree in advance which founder decisions go to the board.
UK#
- Company law remedy: a shareholder can petition the court for unfair prejudice under section 994 of the Companies Act 2006. Courts can order a buyout at a fair value, so this is a strong threat in founder breakups.
- Mediation: CEDR and other mediation services are widely used. Courts expect parties to have considered mediation and can penalise unreasonable refusal through costs orders.
- Deadlock clauses: the articles and shareholders' agreement should include a deadlock process. Joint-venture-style mechanisms such as "Russian roulette" or "Texas shoot-out" buy-sell clauses are rare in startups, because they favour the founder with more cash. Avoid them unless both founders understand them.
- Directors' duties: both founders, as directors, must act in the company's interests even when they are in a dispute. Taking IP, customers or staff to a new venture can be a breach.
5. Scripts, mistakes and related playbooks#
Scripts#
Raising an issue early:
"Something's been bothering me for a couple of weeks and I'd rather raise it now while it's small. Can we spend 20 minutes on it in our Friday meeting?"
Stating facts and feelings, not blame:
"In the last three investor calls, you answered the technical questions before I'd finished. I read that as you not trusting me with the story, and I felt undermined. What I'd like is for us to agree who answers what before each call."
Moving to a decision:
"We've both made our case. This one's in your area, so it's your call. I'll commit to it fully, and let's look at it again in six weeks with data."
Suggesting outside help:
"We keep having the same argument in different forms. I think we're both too close to it. Would you be open to four sessions with a co-founder coach? I'd rather invest in this now than lose what we've built."
Telling the board or lead investor:
"We've had a real disagreement on regulatory strategy. Here's how we resolved it, and the process we've put in place so it doesn't stall us again."
Common mistakes#
- Avoiding the conversation. Silence feels safer, but resentment grows and comes out later in a worse form.
- Arguing in front of the team. Employees start choosing sides, and the conflict spreads.
- No decision owner. Without a clear owner, every disagreement becomes a negotiation between equals.
- Mixing up types of conflict. Arguing about a product decision when the real issue is effort or respect means the fix never sticks.
- Going to lawyers first. Legal letters usually end the relationship. Use mediation first unless there is urgent harm, such as IP or money being taken.
- Using equity as a weapon. Threats to block a round or a vote damage the company for both founders.
- Waiting for success to fix it. Growth often hides conflict, and the next downturn brings it back stronger.
- Reopening the equity split over every argument. A split is a long-term agreement. Change it only through a deliberate, documented process.
Related playbooks#
- 3.03 Co-founder dating: trials and questions
- 3.05 Splitting founder equity
- 3.06 Founder vesting, cliffs and acceleration
- 3.07 Founder agreements: roles, decisions, deadlocks
- 3.08 Assigning titles: CEO, CTO, CSO
- 3.12 Co-founder breakups: buybacks and good/bad leaver terms
- 3.14 Couples, friends and family as co-founders
- 3.16 How each founder's role changes as the company grows
Sources#
- CNN Money, Fighting co-founders doom startups (on Wasserman's research)
- OnStartups, Avoiding founder failure: tips from The Founder's Dilemmas
- Jewish Business News, Wasserman on causes of co-founder conflict
- Stanford eCorner, Learning to "disagree and commit"
- Initialized Capital, Co-founder conflict: why I quit my startup
- StartupYard, Founders at war?
- Denise Tambanis, Pre-mediation for startup teams
- MentorCruise, Kevin Hale on startup relationships
This playbook is general information, not legal advice. Dispute processes and remedies differ by country and depend on your documents, so speak to a lawyer early if a dispute involves equity, IP or removal. Last checked November 2026.