Why structured questions beat chemistry
Choosing a co-founder is one of the few decisions in a young project that is genuinely hard to undo.
You can change your product, your name, even your market, but changing a partner mid-build costs momentum, trust, and often the project itself.
That is why the conversations you have before committing matter more than any pitch or resume.
This guide gives you practical questions to ask a potential co-founder.
It explains what each question tests and how to turn answers into evidence rather than vibes.
The goal is not a perfect interview script.
It is a way to learn what you actually need to know before you share ownership, time, and reputation with someone.
Enthusiasm is easy to mistake for compatibility.
Two people can be excited about the same idea and still fail together because they disagree on commitment, money, or decision-making.
Good questions surface those disagreements while they are still cheap to resolve.
Y Combinator's co-founder matching team makes a similar point.
Like a romantic partner, a co-founder is someone you will depend on and spend years with.
You probably should not marry someone after one date, and it takes more than one video call to decide whether to build a company with someone.
They encourage matched founders to meet and, when appropriate, work together on a time-boxed trial project with clear expectations and goals to vet compatibility source.
So treat the questions below as the conversation layer, and a small trial project as the evidence layer.
Together they give you a far better read than a single impressive chat.
What problem are you solving, and what is your unique insight?
Ask your potential co-founder to explain the problem in plain language, then push one level deeper: what do they know about this problem that most people do not?
Michael Seibel of Y Combinator describes a unique insight as something specific.
It usually comes from customer conversations, product analysis, or personal experience.
He notes that passion does not help here; stating a weak insight forcefully makes someone look worse, not better source.