Co-founder dating: trial projects and the questions to ask

1. Core idea and who it's for#

Choosing a co-founder is closer to a marriage than a hire. It lasts longer than most jobs, is hard to undo, and the conflicts that matter are about values, money, control and commitment, not skills. Co-founder dating is the structured process of testing these before you commit: honest conversations about the hard topics, a real trial project , and reference checks (3.18).

The main principle is to surface disagreements while they're cheap . Find out now whether you differ on:

• ambition and outcome: a large venture-backed company, or a sustainable business; selling early, or building for the long term;
• control: who is CEO, how decisions get made, what happens in a deadlock;
• money and risk: how long you can go unpaid, family obligations, personal runway (1.05);
• commitment: full-time, from when, and what happens to any academic post (2.12, 2.13);
• working style: speed or care, how conflict and feedback are handled, communication habits;
• equity: how it's split, vesting, and leaver terms (3.05, 3.06, 3.12).

Agreeing on these in advance doesn't guarantee a good partnership. But undiscovered disagreement about them predicts the painful splits described in 3.11 and 3.12.

Who this is for: anyone with one or more candidate co-founders, whether from their own network or from the channels in 3.02, before any equity is issued.

2. Evidence: why founder teams break#

Team problems are a leading cause of failure. The most widely cited figure comes from Noam Wasserman's research for The Founder's Dilemmas , which drew on about 10,000 founders. It is commonly summarised as 65% of high-potential startups failing because of co-founder or management-team conflict (CNN Money; OnStartups). Treat the exact

number as directional, since it summarises investor views about failures. The direction is consistent across sources, though: people problems at the top sink more startups than markets or technology do.

Wasserman's other findings that bear on dating:

• Turnover rises after funding. Founding-team turnover increases sharply when the startup raises its first round, and the chance of founder-CEO succession rises with each round (OnStartups). Agreements about roles and leaving need to anticipate that moment.
• "Rich vs king." Founders usually face a trade-off between wealth and control, and rarely get both. Co-founders who want different things here clash later (David's Blueprint, on Wasserman).
• Commitment matters to investors. Paul Graham has pointed out that the YC application asks more about the founders' commitment and relationship than about their abilities (FundersClub).

Splits are common and often expensive.

• Carta data (1.18) shows that more than a quarter of two-founder teams lose a cofounder by year four.
• A 2026 study of more than 150 co-founder departures found that only 35% were clean, free separations, and that 28% cost $50,000 or more (Equity Matrix, via The Lonely Entrepreneur).
• A portfolio study by Icehouse Ventures found that 35% of 100 funded companies had a founder leave, mostly within the first two years (The Lonely Entrepreneur).

Trial projects are the accepted test. YC's long-standing advice is never to commit to a cofounder after one conversation, and to run a short, defined trial project instead (Round Funded). Structured question lists, such as the 50 questions from a First Round Review interview, are widely used for the conversation part (Inc.).

3. Method: four stages over 6–12 weeks#

Stage 1. Values conversation (2–3 sessions). Each of you answers the questions below in writing, separately. Then compare answers and discuss the differences. Bring in a neutral third party, such as a mentor or coach, if it gets stuck.

Stage 2. Trial project (4–8 weeks). Choose a task that:

• matters: it moves the company forward, such as a pilot LOI, a grant application, the first investor meetings or a prototype milestone;
• needs both of you: you have to coordinate and rely on each other's skills; has a deadline and a clear result.

Agree in writing on the basics: who owns any work produced (assign it to the future company), confidentiality, any expenses or small fee, and that vesting will be credited from the start of the trial if you go ahead.

Stage 3. Debrief. Each of you writes down: what went well, what didn't, one thing you'd want the other to change, and whether you'd sign up for seven more years. Discuss honestly. A good sign is being able to disagree openly and still make progress.

Stage 4. References and formalise. Check references, including backchannel references (3.18). Then agree on roles and titles (3.08), the equity split (3.05), vesting and leaver terms (3.06, 3.12), and a founder agreement (3.07). Sign before any significant work or money flows.

Worked example (illustrative)#

A quantum-computing postdoc is considering two candidate co-founders: a former colleague (A) and a commercial lead from a photonics company met through a venture builder (B).

Values conversation.

• A wants a lifestyle-scale business and would sell early. B wants a venture-scale company.
• The postdoc wants a venture-scale company, so the outcome conflict with A is clear early.
• B can go nine months without salary, and agrees the postdoc should be CEO for now, with a review at the Series A.
• Trial with B. Six weeks to win a paid feasibility study with a telecom company. B leads the commercial talks, the postdoc leads the technical scoping, and they meet twice a week.
• Result. They win a $40,000 study. At the debrief, B says the postdoc over-promised on timelines in one meeting. The postdoc says B committed to scope without checking. Both agree on a rule: no commitments on scope or timing without the other.
• References. Former colleagues confirm B's sales record and describe her as direct but fair in conflicts.
• Formalising. They agree a 60/40 split with B's vesting credited from the start of the trial, and sign a founder agreement.

4. India / US / UK differences#

India#

• Family expectations often shape what founders can commit to, including salary needs, timelines and marriage plans (1.17). Raise these openly in the money and commitment conversation. They're a common hidden cause of later conflict.
• Friends and batchmates from IITs and other institutes often co-found together. That gives high trust, but also a risk of avoiding hard conversations (3.14). Use the written exercise anyway.
• IP and paperwork during the trial. Make sure work done in the trial is assigned to the future company, especially if either of you is employed elsewhere (1.15) or at an institute (2.01).

US#

• Confidentiality and IP. Use a simple trial agreement or a contractor agreement with an IP assignment. Several states limit how far an employer's IP clause can reach into personal-time work (1.15), but check the candidate's own employment agreement.
• Lawyers expect founder agreements before incorporation or at it. Many startup law firms have short templates for this.
• Investors will ask how you met and how long you've worked together. A completed trial project gives you a credible answer.

UK#

• Venture builders and pre-team programmes, such as Entrepreneur First, build structured co-founder dating into their programmes. Use their frameworks if you're in one.
• Shareholders' agreements and articles formalise the outcome (3.07). Agree the commercial terms during dating, then let lawyers draft.
• EMI. If the co-founder will get options rather than founder shares, check EMI eligibility. Founders with more than 30% are excluded.

5. Scripts, mistakes, related playbooks and sources#

Scripts#

Opening the hard conversation:

"Before we go any further, I'd like us to each answer the same twenty questions in writing, about money, control, outcome and commitment, and then compare. If we're going to disagree about something big, I'd rather find out now than in two years."

Proposing the trial:

"Let's do six weeks on one real goal: a paid feasibility study with a telecom company. We'll each own a part, meet twice a week, and debrief honestly at the end. Anything we build belongs to the company we'd form, and if we go ahead, your vesting starts from today."

Ending it kindly:

"Working together these six weeks has been genuinely useful, and I've learned a lot. I don't think we want the same thing from this company. You want something that could sell in a few years, and I want to go much bigger. Neither is wrong, but we'd end up fighting about it. I'd love to keep you close as an adviser."

Common mistakes#

1. Skipping the hard topics because it feels awkward, or because you're old friends.
2. Trial projects with no real stakes, such as a business plan or a slide deck, which reveal little.
3. Committing equity before the trial ends, or with no vesting.
4. Ignoring red flags because you're desperate for a co-founder before a fundraise.
5. Leaving outcome and control unspoken. "Rich vs king" differences surface at the first big offer or the first board fight.
6. No written record of what you agreed.

Related playbooks#

• 1.05 Personal runway
• 3.01 Do you need a co-founder?
• 3.02 Where to find co-founders
• 3.05 Splitting founder equity
• 3.06 Founder vesting
• 3.07 Founder agreements
• 3.11 Resolving co-founder conflict
• 3.14 Couples, friends and family as co-founders
• 3.18 Reference checks on a co-founder

Sources#

• Fighting co-founders doom startups, CNN Money, on Noam Wasserman
• Avoiding founder failure: 26 quick tips and real data, OnStartups
• The 65% founder fallout, David's Blueprint
• Finding your co-founder, FundersClub
• The founder partnership fracture, The Lonely Entrepreneur, citing Equity Matrix 2026
• Why co-founder relationships break, The Lonely Entrepreneur
• YC co-founder matching guide 2026, Round Funded
• 50 questions to ask a potential startup co-founder, Inc.

The worked example is illustrative. Last checked November 2026.