Behavioral: Replacing a KPI That Stopped Reflecting the Real Business Goal
Quick Overview
A behavioral product management question asking for a time you had to change a key performance indicator. It tests spotting a metric that drifted from the real goal, choosing and validating a better one with guardrails, bringing affected stakeholders along and reporting a measurable result.
Behavioral: Replacing a KPI That Stopped Reflecting the Real Business Goal
Company: Uber
Role: Product Manager
Category: Behavioral & Leadership
Difficulty: medium
Interview Round: Other
The interviewer asks you to describe a time you had to change a KPI, meaning a metric your team used to judge success or set goals. Answer with one specific story from your own experience.
```hint Show the moment the number lied
The most convincing part of this story is how you realized the old metric was misleading. Be ready to name the exact data that exposed the gap between the metric and the outcome it was supposed to represent.
```
### Clarifying Questions
- Does adjusting a target (for example lowering a goal that was unrealistic) count, or should the story be about replacing the metric itself?
- Should the KPI be one my own team owned, or can it be one set by leadership or another team that I persuaded to change it?
- Is a story acceptable where the new KPI later turned out to have flaws of its own?
### What a Strong Answer Covers
- A concrete situation where the old KPI and the real goal diverged, with the evidence that exposed it.
- How the replacement was chosen: what outcome it tracks, why it is harder to game, and which guardrail metrics sit beside it.
- Stakeholder handling: whose goals, dashboards or incentives depended on the old number, and how agreement was reached.
- The transition mechanics (backfilling history, running both metrics in parallel, resetting targets) and a measurable result.
- Reflection on what the candidate learned about metric design and now does differently.
### Follow-up Questions
- Some people's quarterly goals were tied to the old KPI. How did you handle switching in the middle of a goal cycle?
- How did you confirm that the new KPI actually predicts the business outcome, rather than just sounding more sensible?
- Looking back, what signal would have told you sooner that the old KPI was wrong?
Overview: A behavioral product management question asking for a time you had to change a key performance indicator. It tests spotting a metric that drifted from the real goal, choosing and validating a better one with guardrails, bringing affected stakeholders along and reporting a measurable result.
Behavioral: Replacing a KPI That Stopped Reflecting the Real Business Goal
Uber
Jan 14, 2025
mediumProduct ManagerOtherBehavioral & Leadership
0
0
The interviewer asks you to describe a time you had to change a KPI, meaning a metric your team used to judge success or set goals. Answer with one specific story from your own experience.
Clarifying Questions Guidance
Does adjusting a target (for example lowering a goal that was unrealistic) count, or should the story be about replacing the metric itself?
Should the KPI be one my own team owned, or can it be one set by leadership or another team that I persuaded to change it?
Is a story acceptable where the new KPI later turned out to have flaws of its own?
What a Strong Answer Covers Guidance
A concrete situation where the old KPI and the real goal diverged, with the evidence that exposed it.
How the replacement was chosen: what outcome it tracks, why it is harder to game, and which guardrail metrics sit beside it.
Stakeholder handling: whose goals, dashboards or incentives depended on the old number, and how agreement was reached.
The transition mechanics (backfilling history, running both metrics in parallel, resetting targets) and a measurable result.
Reflection on what the candidate learned about metric design and now does differently.
Follow-up Questions Guidance
Some people's quarterly goals were tied to the old KPI. How did you handle switching in the middle of a goal cycle?
How did you confirm that the new KPI actually predicts the business outcome, rather than just sounding more sensible?
Looking back, what signal would have told you sooner that the old KPI was wrong?