How do I choose a business cofounder?

Choose a cofounder by examining the business you want to build together, the work each person can actually supply and how you behave when a real decision becomes difficult. Discuss goals and limits openly, then use a bounded, properly resourced piece of joint work to test important assumptions before making ownership commitments.

Enjoying somebody's company is a good beginning. So is admiring what they can do. The next question is whether the proposed relationship fits the work and life each of you is actually willing to take on.

Describe the company and the missing role

Start with the operation, not the person. What needs doing during the next stage? Which responsibilities must be carried every week? What can you do well, what can you learn and what genuinely needs another continuing owner?

A missing skill might be supplied through an employee, adviser, provider or limited collaboration. A cofounder brings a wider relationship involving continuing choices, responsibility and ownership. Decide what that relationship is supposed to add.

Do not require somebody to match your enthusiasm before showing them the actual role. “Help grow the business” can mean selling, building, handling customers, paying bills or appearing at occasional meetings. Put hours, deliverables, decision responsibility and the expected period beside the phrase.

Separate what you need now from what you may need later. A person can be excellent for a launch project and unavailable for the operating role that follows. That is useful information to discover early.

Talk about the real commitment

Discuss the kind of business each person wants, the pace, the financial exposure and the work they want to do. Include the possibility that a useful smaller company would be a success for one person while disappointing the other.

Ask what time is genuinely available after existing commitments, and what would have to change for that availability to grow. A planned resignation or possible investment is conditional. Keep it separate from hours or money already available.

You do not need somebody's entire private history to understand their limits. You need an honest account of the commitment they can make and how changes will be communicated. Treat a stated limit as a planning input, rather than a character flaw to argue away.

Y Combinator's cofounder discussion guidance recommends joint trial work and transparent reference conversations before deciding to work together. Use those ideas to gather evidence; the original example below is a separate, paid planning project, not YC's questionnaire or a guarantee of compatibility.

Give a bounded project enough resources to be informative

Consider two fictional people, A and B, exploring a booking-support business together. A would handle the prototype and operating design. B would handle the offer research and costing. They have not allocated founder equity or accepted customer orders.

They agree a four-week paid planning project. Each can supply eight hours a week, or thirty-two hours each. A's allocation is sixteen prototype hours, eight operating-plan hours and eight joint-review hours. B's is twelve offer-research hours, twelve costing hours and eight joint-review hours.

The work uses public information and invented example bookings. It does not claim that prospective customers have been interviewed, that demand exists or that the service is ready to sell. Its purpose is to examine how the pair work and what the proposed operation still requires.

Each person's time is fully funded at an all-in $50 an hour under the supplied project arrangement: sixty-four hours cost $3,200. Two professional-review hours at $150 and $100 of tools bring complete project payments to $3,600. These are fictional complete quotes for this scope, not market fees or a promised cost of forming a company.

The project has $4,000 separately available, excluding other work and household obligations. Professional review and tools use $400 at the start. Four weekly $800 payments then leave $400. All hours are available without displacing accepted obligations, and no future customer or investor receipt pays for the exercise.

Before starting a real project, establish the authorized scope, payment, ownership and permitted use of created material, confidentiality and who can commit expenditure or speak for the venture. Use the actual legal arrangement appropriate to the relationship. Calling something a trial does not settle those questions by itself.

Observe how the work reaches a decision

Choose deliverables with a clear use. Here, the prototype must show a booking progressing through the proposed service, while the operating plan identifies the paid work and dependencies behind it. The costing must trace the required resources rather than merely produce an attractive price.

At each joint review, ask each person to explain what is complete, what changed and which decision needs the other person's input. Keep unfinished items visible. A confident presentation is less useful if nobody can trace its assumptions.

Introduce an ordinary disagreement in the planning work: the prototype can offer a convenient customer option, but the cost record shows that it requires another paid handoff. The useful question is how the pair investigate that conflict.

Do they examine the required work and revise the offer, or does one dismiss the cost because the feature looks exciting? Can the other person question an estimate without the discussion becoming a referendum on loyalty? An easy project can conceal the very working difference you need to understand.

Do not create an artificial crisis or secretly test somebody with missing information. Agree that the project includes discussing tradeoffs and keep the relevant evidence available. You want to understand a working relationship, not stage a trap.

Also observe the routine parts: meeting the agreed handoff, preserving a correction, explaining an uncertainty and finishing a task after the interesting first draft. A business will need those behaviors more often than a dramatic pitch.

A project record that can change the choice

Scroll across to connect the observation with the remaining decision. Use the arrow keys when the table area is focused.

Let the calendar change the proposed role

At the end of the fictional project, B can confirm only four hours a week for the continuing role. The eight-hour trial schedule used a temporary arrangement that cannot continue. The proposed next-stage commercial role requires twelve hours a week.

That leaves an eight-hour weekly gap. The team has no funded replacement arrangement or revised operating plan. The planning project may have gone well; the next-stage role is still understaffed.

A can investigate a smaller role for B, redesign the operation, find a different continuing arrangement or postpone the next stage. None of those options exists simply because the pair want to preserve the original story.

Do not describe B as an inadequate founder because four hours fails a twelve-hour role. Nor should B promise twelve to avoid an uncomfortable conversation. The useful decision is about the actual business and the available contribution.

The origin-contribution guide distinguishes money, help and ongoing work. An introduction or a prototype deserves accurate recognition without being treated as proof that somebody will carry every future responsibility.

Use references to understand the relationship

With appropriate permission and transparency, speak with people who have directly worked with the prospective cofounder. Explain the kind of role and relationship being considered. Ask for specific circumstances rather than a broad certificate of character.

You might ask about a difficult change, a commitment that slipped or a disagreement over priorities. Establish what the reference personally observed and what belongs to another person's account. Keep the context: a strong performance in a short specialist role may not establish availability for ongoing company leadership.

Offer the same openness in return. A reference process conducted secretly can undermine the relationship it is supposed to help assess. Respect confidentiality and do not pressure someone to supply private records or information unrelated to the decision.

Keep unresolved information visible. A reference who cannot discuss a project has not confirmed it, but silence is not evidence that the project failed. The next step may be a different relevant reference or a narrower decision.

Make ownership a separate decision

After the work and conversations, decide whether the ongoing relationship makes sense before negotiating its full terms. Then establish roles, economic rights, decision authority, funding, created assets and what happens if the relationship changes or ends.

Those questions need actual agreements and appropriate professional help. A friendly equal split cannot explain who can approve borrowing or how a departure would be handled. The shareholder-decision guide shows why a percentage does not supply every operating permission.

Do not assign a dollar value to every introduction or hour simply to make a preferred split look mathematically inevitable. Past contributions, future work, risk and rights need their own descriptions. Some can be measured; others require a judgment that should be acknowledged as such.

Use the trial record to identify questions, not to pretend that four weeks has solved all of them. If the project exposed an important mismatch, preserve that finding even when the other person is talented and well liked.

Finish with a choice you can explain

A useful decision record states the business being proposed, the continuing role, actual availability and resources, evidence from the work, unresolved concerns and the next commitment. Each person should be able to explain what they are choosing without relying on the other person's optimism.

The answer may be to proceed, to keep collaborating in a limited role or to stop before ownership becomes entangled with expectations neither person can meet. An early respectful no can preserve both resources and a good professional relationship.