Diagnose YouTube Usage Decline: Key Metrics and Segmentation
Overview
This question evaluates a data scientist's competence in product analytics, specifically funnel metric interpretation (acquisition → activation → engagement → retention), user segmentation, monitoring and experimentation design to diagnose sudden declines in platform usage.
Google
Jul 12, 2025, 6:59 PM
Data Scientist
Technical Screen
Analytics & Experimentation
73
0
Scenario
YouTube observes a sudden decline in daily active users (DAU) and total watch time across the platform.
Task
Design a systematic diagnosis plan for the drop in usage. Specify:
- Which engagement and funnel metrics you would inspect.
- How you would segment users to localize the issue.
- What further analyses or experiments you would run to confirm root causes.
Context and Assumptions
Assume you have access to event logs (impressions, clicks, plays, session starts/ends), QA/monitoring dashboards, experiment/feature rollout logs, content supply stats, and standard experimentation tools (A/B, holdouts, geo rollouts/rollbacks).
Notes
- Frame the analysis using the acquisition → activation → engagement → retention funnel.
- Segment by device/OS/app-version, geography/time zone, traffic source, and recent feature/model launches.
- Compare pre- and post-drop cohorts, and consider hold-out or rollback tests to isolate factors.
Solution
Login required
Show