Uber Ride Bookings Fell 10%: Diagnose the Root Cause and Plan a Response
Company: Policybazaar
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
Uber's booking rate has fallen by 10%. How would you work out why, and how would you respond?
```hint Two sides to every booking
A booking needs a rider willing to request and a driver willing to accept. Before listing causes, settle what the number measures and whether the fall is real, then find where in the request flow, and on which side of the marketplace, it shows up.
```
### Clarifying Questions
- What does "booking rate" measure here: total completed trips, ride requests, or the share of app sessions that end in a confirmed booking?
- Is the 10% a relative drop or 10 percentage points, and compared with what: last week, the same week last year, or a forecast?
- Did the drop happen suddenly or build up over several weeks?
- Is it global, or limited to certain cities, platforms or products, such as Rides versus Eats?
### What a Strong Answer Covers
- Confirming the metric definition and ruling out tracking, data-pipeline and seasonal explanations before diagnosing.
- A structured split of internal causes (releases, pricing, experiments, outages) and external ones (competitors, events, weather, regulation).
- Breaking the drop down by funnel step and by segment, including driver supply, ETAs and surge pricing.
- A prioritized order for testing hypotheses, with the data that would confirm or rule out each one.
- Fixes matched to the root cause, plus monitoring that would catch the next drop sooner.
### Follow-up Questions
- The fall is concentrated in requests that never find a driver. What do you check next?
- Only Android users on the latest app version are affected. What do you do in the first day?
- A competitor has just cut prices in your largest cities. Would you match them, and how would you decide?
Overview: A metric diagnosis question about a 10% fall in Uber's booking rate. It tests confirming the metric definition and data quality first, separating internal from external causes, breaking the drop down by funnel step, segment and driver supply, and matching fixes and monitoring to the root cause.