Engineer pushes back on a data-backed product call: reaching a shared decision
Company: Perplexity
Role: Product Manager
Category: Behavioral & Leadership
Difficulty: medium
Interview Round: Other
You and an engineer disagree about a product decision. Your position rests on data; theirs rests on technical concerns. How would you resolve it while keeping the collaboration healthy?
```hint Check what each side's evidence covers
Your data and the engineer's concerns may be answering different questions about the same decision. Work out what each one can and cannot tell you before deciding who is right.
```
### Clarifying Questions
- Is the disagreement about what to build, how to build it, or when to ship it?
- What kind of data supports the product decision (an experiment result, funnel analytics or user research), and how strong is it?
- What is the engineer worried about: effort, feasibility, reliability, performance or long-term maintenance cost?
- Is there a deadline, and who makes the final call if the two of you cannot agree?
### What a Strong Answer Covers
- Understanding the engineer's objection fully, and restating it, before defending your own position.
- Treating the technical concern as evidence about cost and risk, and being honest about the limits of your data.
- Turning a win-or-lose debate into options: a smaller version, phasing, an alternative implementation, or a time-boxed spike or experiment.
- A clear decision path when agreement does not come: agreed criteria, a decision date, joint escalation, and commitment afterward.
- Actions that protect trust: credit, transparency, and reviewing the outcome against the concern the engineer raised.
### Follow-up Questions
- You overruled the engineer, and it later turned out they were right. What do you do?
- The engineer questions whether your experiment was large enough to trust. How do you respond?
- What would you do if the engineer took the disagreement to their manager without telling you?
Overview: A behavioral product manager question about resolving a disagreement with an engineer when the PM's data supports one direction and the engineer raises technical concerns. It tests listening, weighing both kinds of evidence, creating options, owning a decision and keeping the working relationship strong.