Behavioral question

Tell me about a time you had to change a KPI.

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

Judgment in metric selection, showing the candidate can recognize and correct a flawed or misleading KPI.

Set the scene: the original KPI, why it was chosen, and what went wrong, like it being easy to game or not tied to real value.
Explain the specific signal that revealed the KPI was flawed, such as the metric improving while a downstream outcome worsened.
Describe the new KPI chosen and why it better reflects the real goal.
Explain how you got stakeholder buy-in to change a metric people were already tracking and reporting on.
State the result after the change, including any resistance faced along the way.

Names the specific flawed KPI and the concrete evidence that exposed the flaw, not a vague 'it wasn't working'.
Explains the replacement metric clearly and why it better captures the real underlying goal.
Describes real stakeholder pushback and how it was addressed, showing this wasn't a trivial change.
States a concrete result after the switch, showing the new KPI drove better decisions.

Describing the change with no concrete evidence for why the old KPI was actually flawed.
Skipping the stakeholder resistance, making the change sound easier than it likely was.

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