A cofounder trial project is a 7 to 14 day sprint in which two prospective cofounders complete real work together before discussing equity. Choose one useful outcome, define ownership, speak with customers, make at least one meaningful decision and finish with a written retrospective.
The goal is not to simulate years of company building. It is to collect direct evidence about how you communicate, decide, disagree and follow through.
Do not promise equity as a reward for completing the trial. Keep company formation, ownership, intellectual property, tax and governance decisions outside the project until you have received appropriate professional advice.
decide whether you are ready for a trial
A trial makes sense after you have discussed the problem, your motivations and the role each person might play. If you are still searching for possible partners, start with this guide to finding a cofounder.
Before you schedule the project, ask each other:
- Why do you want to build this company?
- How much time can you commit during the trial?
- What other work or personal responsibilities affect your availability?
- Which skills and decisions do you expect to own?
- What would make you stop pursuing the idea?
- What information, accounts or materials will remain private?
- Can either person end the trial without owing the other a partnership?
Listen for specific answers. “I will work whenever needed” tells you less than “I can work from 7:00 to 9:00 on four evenings and join two daytime customer calls.”
YC’s guide to finding the right cofounder encourages founders to learn about compatibility by working together. Use that guidance as a starting point, then design a test that reflects the company you may actually build.
choose one small but real outcome
Pick a project that requires cooperation without giving either person unnecessary access to sensitive systems, customer data or money.
A useful trial might produce:
- a prototype tested with prospective users
- a landing page supported by customer conversations
- a written synthesis of a specific customer problem
- a manual version of one proposed service
- a short product specification based on direct feedback
Avoid vague goals such as “explore the idea” or “see whether we get along.” You need an observable outcome and enough friction to reveal how both people work.
Write the scope in one paragraph:
From [start date] through [end date], we will produce [specific output] for [specific audience]. We will test it through [customer-contact method]. [Name] owns [area], and [name] owns [area]. We will communicate through [channel], meet at [times] and write separate retrospectives before our final discussion.
A narrow project protects the trial from becoming an unpaid, open-ended audition. It also stops you from explaining every missed commitment as a consequence of an oversized goal.
use this 10-day cofounder trial project
You can shorten this plan to seven days or extend it to 14. Keep the same sequence.
| When | Work | What you are testing |
|---|---|---|
| Day 0 | Agree on the scope, availability, access boundaries and definition of done. | Clarity, honesty and expectation setting |
| Days 1 and 2 | Divide ownership, create the first version and identify assumptions. | Initiative, ownership and communication |
| Days 3 and 4 | Put the idea in front of prospective users or customers. | Customer contact, listening and adaptability |
| Day 5 | Review evidence and make one meaningful project decision. | Judgment, disagreement and decision speed |
| Days 6 to 8 | Build or revise the agreed output. | Pace, reliability and response to feedback |
| Day 9 | Complete the deliverable and document unfinished work. | Follow-through and transparency |
| Day 10 | Write separate retrospectives, exchange them and decide what happens next. | Self-awareness, trust and conflict repair |
Do not fill every hour. Leave enough room for a missed assumption, changed plan or disagreement to emerge naturally. A perfectly scripted exercise will test your ability to follow a script, not your ability to build together.
write a lightweight working agreement
Create a one-page agreement for the trial. This is an operating document, not a substitute for a contract or legal advice.
Cover these eight areas:
Communication: Choose one main channel and define what belongs there. Decide how you will report progress, blockers and changed assumptions.
Ownership: Give one person clear responsibility for each deliverable. Ownership means moving the work forward and asking for input when needed, not controlling every decision.
Decisions: Separate reversible decisions from decisions that could create meaningful cost, risk or commitment. Agree on who decides when you cannot reach consensus during the trial.
Pace: State the hours or work blocks each person can actually protect. Do not use late-night messages as evidence of commitment.
Conflict: Agree to name disagreements directly, explain the evidence behind each position and avoid personal attacks.
Availability: Identify known unavailable periods and a response window that fits both schedules. Tell each other early when a commitment will slip.
Customer contact: Require both people to participate. One person can lead a conversation while the other takes notes, but neither should remain permanently separated from users.
Retrospective: Schedule it before work starts. Each person should write independently before discussing the relationship together.
This agreement gives you something concrete to compare with what happened. Without it, both people can leave the same week carrying different stories about what they promised.
create direct customer contact
Do not test the partnership entirely through private planning. Include a real interaction with the people who experience the problem.
Before each conversation, agree on:
- what you want to learn
- who will lead
- who will take notes
- which claims you will avoid making
- how you will protect personal or confidential information
- when you will compare interpretations
Afterward, write your observations separately before discussing them. Notice whether one person listens, changes their mind when evidence changes and distinguishes a customer’s words from their own interpretation.
work together in the same session
After you have defined the project, use one shared work block to observe how the collaboration feels in practice. A coworking cafe can work when both people agree on the venue, timing and expected level of conversation.
founderhub helps founders explore nearby builders and coworking tables on a map, join or host a table, and choose what to share when matched or when someone RSVPs. Host a 90-minute table for the trial kickoff, or join an existing work session and bring one clearly defined task for each person.
Use the final ten minutes to compare progress, blockers and the next decision. The session should produce work, not become another abstract conversation about becoming cofounders.
score the working relationship separately
Complete this scorecard alone before the retrospective. Use a score from 1 to 5:
- 1: repeated evidence of a serious mismatch
- 2: significant concern with little repair
- 3: mixed or incomplete evidence
- 4: consistently workable
- 5: strong evidence of trust and compatibility
| Area | Question |
|---|---|
| Communication | Did this person share useful context, surface blockers and respond within the agreed window? |
| Ownership | Did they move their responsibilities forward without repeated prompting? |
| Decisions | Could we make timely decisions with incomplete information? |
| Pace | Did our realistic working rhythms fit well enough to coordinate? |
| Conflict | Could we disagree directly, stay respectful and repair tension? |
| Availability | Did their actual availability match what they stated? |
| Customer contact | Did they listen carefully and respond to evidence instead of defending an assumption? |
| Reliability | Did they keep commitments or communicate changes early? |
| Judgment | Did they protect sensitive information and avoid unnecessary risk? |
| Motivation | Did their actions match the reasons they gave for building the company? |
Add one example beside every score. A number without evidence invites argument.
Do not combine the scores into a pass mark. A high total cannot cancel dishonesty, coercion or a serious values conflict. Compare your scores to find gaps in perception. If one person rates communication a five and the other rates it a two, discuss the events behind those scores.
treat these as stop signals
End or pause the trial when you see:
- dishonesty about work, availability, customers or commitments
- pressure to promise equity before both people are ready
- repeated missed commitments without early communication
- personal attacks, intimidation or retaliation during disagreement
- requests for credentials, funds or sensitive data that the task does not require
- unilateral promises made to customers or partners
- refusal to document decisions or acknowledge mistakes
- disregard for stated personal, ethical or privacy boundaries
- an attempt to claim sole ownership of shared trial work without prior agreement
You do not need to prove that someone is a bad person before you stop. A material mismatch is enough. End the project clearly, document the work completed and remove any access that was granted specifically for the trial.
run a written retrospective
Each person should answer the same questions before meeting:
- What did we complete?
- Which commitments did each person keep or miss?
- When did communication work well?
- Which decision created the most tension?
- How did we handle that tension?
- What did customer contact change?
- Did our pace and availability fit?
- What would I change in another trial?
- What concerns am I reluctant to say aloud?
- Do I want to stop, run another scoped project or explore a formal relationship?
Exchange the written answers before the discussion. Start with evidence, then address differences.
Your next step does not need to be a permanent yes or no. You can stop, run a second trial that tests an unresolved area or begin professional due diligence. If you continue, do not treat trial completion as an automatic agreement on equity.
Cooley GO’s overview of dividing founder equity provides general information from a US startup-law perspective. Use it to prepare questions, not to make legal decisions from an article. Consult qualified legal and tax professionals in the relevant jurisdiction before agreeing on equity, vesting, intellectual property, company control or other binding terms.
A short project cannot predict every future conflict. It can replace imagined compatibility with observed behavior. Define one real outcome, work closely enough to encounter decisions and uncertainty, then believe the evidence you created together.