Behavioral question

How do you know as a PM that you have enough data to make decisions?

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Whether you have a practical, non-dogmatic judgment for when data is sufficient to act versus when more research is needed.

Give your general principle first: enough data means the marginal value of more information is lower than the cost of delay.
Explain how you weigh the decision's reversibility, since a reversible decision needs far less data than a costly, hard-to-undo one.
Describe a concrete signal you look for, like a consistent pattern across multiple independent data sources or user segments.
Address what you do when data is genuinely insufficient: state the assumption explicitly and set a checkpoint to revisit after launch.
Give a specific past example where you decided you had enough, or not enough, and why.
Note that qualitative signal, like recurring user complaints, can be enough for many product decisions even without a large quantitative sample.

States a clear, generalizable principle, reversibility versus cost of delay, instead of a vague it depends.
Gives a concrete example showing the principle applied to a real decision, not just in the abstract.
Distinguishes reversible, low-stakes calls needing less data from high-stakes, hard-to-undo ones needing more.
Shows comfort acting on a stated assumption with a review checkpoint rather than waiting for perfect certainty.

Giving only a vague answer like when it feels right without any concrete decision principle.
Implying you always wait for large-sample quantitative data, ignoring that many good decisions use qualitative signal.

Can you give an example where you decided you didn't have enough data and what you did?
How do you know when you're using insufficient data as an excuse to delay a decision?