Behavioral question

Tell me about a time when you made a decision that was not data-driven. What is something that you could've changed to make that decision?

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

Self-awareness and judgment about decisions made without full data, tested against Amazon's data-driven culture.

Situation: briefly set up a real decision made on instinct or qualitative signal because data wasn't available yet.
Task: state what was at stake and why data wasn't available or sufficient at the time.
Action: describe the reasoning actually used, for example customer interviews or competitive signal instead of hard data.
Result: state the outcome honestly, good or mixed, with a number if possible, for example adoption after launch.
Reflection: name specifically what data source or experiment would have de-risked the decision, like a small pilot or proxy metric.
Tie back to a principle: default to data where available, but act decisively without it when needed.

Picks a real, specific decision with clear stakes, not a hypothetical.
Owns the outcome honestly, including if it went imperfectly, rather than making everything look like a win.
Names one concrete, feasible change, for example running a 2-week A/B test first, rather than a vague 'be more data-driven.'

Chapters of the AI PM course that teach what this question tests.