Identify Key Metrics to Address Delivery Delays
Overview
This question evaluates a data scientist's proficiency in defining primary and supporting delivery-performance metrics, performing root-cause segmentation across the end-to-end order lifecycle, and designing experiments to measure product or operational changes, within the Analytics & Experimentation domain.
DoorDash
Aug 4, 2025, 10:55 AM
Data Scientist
Technical Screen
Analytics & Experimentation
0
0
Diagnosing and Improving On‑Time Delivery in a Food‑Delivery Marketplace
Scenario
A food‑delivery marketplace is seeing growing customer complaints about orders arriving late.
Your Task (Data Scientist — Analytics & Experimentation)
- Metrics: Which metrics would you examine first, and why? Clearly define a primary on‑time metric and supporting secondary metrics.
- Root cause and segmentation: How would you identify the root causes of lateness and segment the problem (e.g., region, restaurant, courier, order attributes)?
- Experiment/product change: Propose a concrete product or operational change aimed at improving on‑time delivery, and outline an experiment design (e.g., A/B or geo holdout) to evaluate it.
Notes
- Clarify what "late" means (e.g., relative to initial customer promise vs latest ETA).
- Consider the end‑to‑end order lifecycle (prep, assignment, pickup, travel, drop‑off).
- Use segmentation (region/zone, restaurant cohorts, courier supply, time‑of‑day, order type).
- Form hypotheses and propose a measurable test with success metrics and guardrails.
Solution
Login required
Show