Metrics question

As a PM at Amazon, if returns and cancellations have recently peaked, how would you go about identifying the root cause?

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

Tests root-cause investigation methodology for an operational quality metric spike.

Segment by category, seller, first-party versus third-party marketplace, and reason code, since causes differ completely by segment.
Check timing correlation with a recent seller onboarding surge, a logistics or carrier disruption, or a category-specific quality issue.
Cross-reference with delivery performance data, late or damaged shipment rate, since fulfillment issues commonly drive returns.
Prioritize the segment or category contributing disproportionately to the spike, a Pareto view of which category accounts for most of the increase.
Recommend targeted action once isolated, for example seller quality enforcement if concentrated in specific third-party sellers.

Leads with reason-code and first-party-versus-third-party segmentation, the fastest way to narrow a broad problem at Amazon's scale.
Ties the investigation to concrete operational data, carrier performance and seller quality, not just demand-side guesses.
Prioritizes by Pareto contribution rather than treating every category as equally responsible.

Proposing broad fixes like improving product descriptions without first segmenting by category, seller or reason code.
Ignoring the operational and fulfillment angle entirely.

How would you isolate whether this is a seller-quality issue versus a fulfillment issue?
What immediate action would you take while root cause is still being investigated?