Product and Operations Case: Grocery OOS, Delivery Radius, and Free Delivery

Context

You are a data scientist in an onsite analytics and experimentation interview. The marketplace recently launched a new grocery vertical and is observing elevated out-of-stock (OOS) rates. Current delivery radius is capped at 3 miles. There is also interest in offering free delivery for selected restaurants to non-members.

Assume you have access to order logs, item-level pick events, substitutions, cancellations, refunds, customer support contacts, dasher time-and-distance telemetry, store metadata (inventory feeds, hours), and experiment platforms that support geo and user-level randomization.

Tasks

1. Diagnose the causes of high OOS in the grocery vertical and propose solutions. Specify hypotheses, key metrics, and analyses you would run.
2. The current delivery radius is limited to 3 miles. Should we extend the radius? Describe the analysis and metrics you would use, including how you would model operational cost versus revenue impact, and how you would test the change.
3. Should we offer free delivery for selected restaurants to non-members? How would you evaluate this? Outline experiment design, success metrics, and unit economics.