Design analytics for a new-market launch

Quick Overview

This question evaluates a data scientist's competency in product analytics and causal inference, focusing on metric definition, guardrail setting, and experimental rollout design for a new-market launch.

Design analytics for a new-market launch

Company: DoorDash

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: hard

Interview Round: Onsite

DoorDash is entering a new city. Define the success metrics and guardrails for three phases: pre-launch (readiness), soft launch (limited zones/hours), and full ramp. For each phase: (a) list 3–5 core metrics with precise definitions (units, numerator/denominator, time window), (b) identify leading indicators vs outcome metrics, (c) propose guardrail thresholds that would trigger a rollback or hiring freeze. Then design a geo A/B or staggered rollout to estimate causal impact on orders and ETAs, addressing: seasonality, matched-control selection, spillovers across adjacent zones, novelty effects, and minimum-detectable-effect assumptions. Specify the decision rules you’d present to leadership to proceed/hold after two weeks.

Quick Answer: This question evaluates a data scientist's competency in product analytics and causal inference, focusing on metric definition, guardrail setting, and experimental rollout design for a new-market launch.

|Home/Analytics & Experimentation/DoorDash
DoorDash logo
DoorDash
Oct 13, 2025, 9:49 PM
hardData ScientistOnsiteAnalytics & Experimentation
33
0

DoorDash New-City Launch: Metrics, Guardrails, and Causal Rollout Design

Task

Define success metrics and guardrails for three phases of a new-city launch, and design an experiment to estimate causal impact.

Phases

  1. Pre-launch (readiness)
  2. Soft launch (limited zones/hours)
  3. Full ramp

For each phase

(a) List 3–5 core metrics with precise definitions (units, numerator/denominator, time window). (b) Identify which are leading indicators vs outcome metrics. (c) Propose guardrail thresholds that would trigger a rollback or hiring freeze.

Experiment design

Design a geo A/B or staggered rollout to estimate causal impact on orders and ETAs. Address:

  • Seasonality
  • Matched-control selection
  • Spillovers across adjacent zones
  • Novelty effects
  • Minimum-detectable-effect (MDE) assumptions

Specify the decision rules you would present to leadership to proceed or hold after two weeks.

Loading comments...