Metrics question

If Uber Eats orders are down how will you track 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 diagnosis for a two-sided delivery marketplace where a drop could originate from demand, supply, or restaurant partners.

Confirm scope: is the order drop global, regional, or specific to certain restaurants or cuisines, and over what time window.
Check for instrumentation issues first: a tracking bug or app release that could be undercounting orders rather than a real drop.
Segment by funnel stage: are fewer users opening the app, fewer completing checkout, or more orders being cancelled after placement.
Check supply-side causes: restaurant availability (closures, long prep times) and delivery courier availability in affected areas.
Check external causes: a competitor promotion, weather, or a local event, before concluding it's a product issue.

Moves through a clear funnel (app opens, checkout, post-order cancellation) to localize exactly where the drop happens.
Considers both demand-side (fewer users ordering) and supply-side (restaurant/courier availability) causes, reflecting the two-sided nature of Eats.
Uses an illustrative number, e.g. assume checkout conversion is flat but cancellation after order rose 5 points, pointing at fulfillment not demand.
Rules out instrumentation and external market factors before committing to a product-level root cause.

Assuming the cause is on the demand side without checking restaurant or courier supply issues.
Skipping the instrumentation check and jumping straight to a behavioral hypothesis.

How would you distinguish a local courier shortage from a broader demand problem?
What would you do in the first 24 hours versus the first week of this investigation?