Metrics question

Drivers are dropping out of a city on Lyft. How do you figure out what's going on?

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

Structured diagnosis of driver side churn in a specific market, prioritizing hypotheses by likely impact before proposing a fix.

Verify the pattern is real and city specific, ruling out a broader platform wide dip or a reporting error first.
Hypothesize pay related causes, fewer high fare trips, increased competition from another rideshare company, or rising local costs like gas and insurance.
Hypothesize platform experience causes, unsafe pickup zones, frequent app crashes, or poor pickup navigation specific to that city's road layout.
Hypothesize external causes, local regulation changes, new licensing or insurance requirements raising the cost of driving.
Prioritize investigation by ease and impact, comparing driver earnings per hour in this city against comparable cities, then following up with driver exit interviews.

Structures the investigation across pay, experience, and regulatory categories rather than guessing a single cause upfront.
Proposes a specific comparative benchmark, earnings per hour against similar sized cities, the fastest way to test the pay hypothesis with existing data.
Proposes direct driver exit interviews as the way to confirm the real reason rather than relying only on inferred data patterns.

Assuming the cause is pay related without checking if a competitor or regulation change is actually the real driver.
Not benchmarking against similar cities to confirm the drop is a real local anomaly.
No direct outreach to departing drivers to validate the hypothesis.

What would you do differently if the drop coincided with a competitor's promotional pay bonus in that city?
How would you design a driver retention incentive without overpaying drivers who would have stayed regardless?