You can like someone a lot and still be unable to work with them. A good conversation shows you the first part. A small shared project shows you the second, and it's much cheaper to find out before anyone owns part of a company.
What a trial gets you
You see how the person actually works, not how they describe working. Disagreements come up early, while they're still small. You both learn whether your pace and standards match. And either of you can walk away without any legal or financial tangle.
Designing one
Keep it small and real. Something that takes two to six weeks and produces something you can point to: a prototype of one feature, a round of customer interviews, a landing page test, a competitor teardown.
Write down the goal and the deliverable, who owns which part, how many hours each person will put in each week, and a date when you'll both decide whether to keep going. Also agree in advance on who owns whatever gets created during the trial, and put that in writing. The rules vary by place and situation, so it's worth getting legal advice.
What to watch
Pay attention to the boring stuff. How fast does the person reply? Do they do what they said they'd do? Do you agree on what "done" means? When you disagree, does it get resolved or quietly dropped? And do they still show up when the work turns tedious?
The checkpoint conversation
At the end, each of you answers four questions: What worked? What was frustrating? Would I want to do this for several years? What would need to change? Be honest. Deciding not to continue is a perfectly good outcome.
But first you need someone to try it with
That candidate might come from colleagues, communities, or introductions. Omnitwine is one way to meet relevant people. It suggests members based on your conversations with its AI and needs both people to approve before you can message. The trial project is how you test the fit once you're talking.
What a trial can't tell you
A few weeks won't show how someone behaves under long-term stress. A part-time trial won't prove full-time commitment. And working well on a small task doesn't mean you agree about the vision.
How long should it last? A few weeks is typical. Long enough to see real work, short enough to leave easily.
Should I pay the person? It depends on the scope and your situation. Just be explicit about whatever you agree.
If it goes well, the next step is The Cofounder Equity Conversation..com/blog/cofounder-equity-conversation).