Google·Data Scientist·Technical Phone Screen·Senior

Senior

Apr 2026

Summary

Google Data Scientist interview with a product analytics case centered on diagnosing a platform-wide drop in YouTube usage. One meaty question, very open-ended, and the kind where you can tell pretty quickly if you have a real framework or are just winging it.

Questions Asked (1)

YouTube is seeing a sudden drop in daily active users and total watch-time. How would you systematically diagnose what's going on? Walk through which metrics and funnel stages you'd look at, how you'd segment users to narrow down the problem, and what analyses or tests you'd run to confirm the root cause.

Product Analytics & MetricsRoot Cause AnalysisA/B Testing & Experimentation

Author's notes

This one felt manageable at first and then I realized halfway through my answer I was just listing metrics with no real structure.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by validating the data and defining the metric precisely, then systematically segment the drop by user cohorts, platforms, and funnel stages to isolate where the decline is concentrated. Use a hypothesis-driven approach to run analyses and A/B tests that confirm the root cause, and always consider both internal (product changes, bugs) and external (seasonality, competition) factors.

Pro tip: Always check for data pipeline issues or logging bugs first—many 'sudden drops' are measurement artifacts, not real user behavior changes. Also, quantify the impact size early to prioritize which segments or hypotheses to investigate.

1. Validate and Define the Metric

Confirm the drop is real by checking data pipelines, logging, and metric definitions. Clarify what 'daily active users' and 'watch-time' mean (e.g., unique users, sessions, minutes) and ensure consistency across dashboards.

2. Segment and Localize the Drop

Break down the metrics by dimensions like platform (iOS, Android, web), geography, user tenure, acquisition channel, and content type. Identify which segments are driving the decline and whether it's broad or concentrated.

3. Map to Funnel Stages

Analyze the user funnel: acquisition (new users), activation (first watch), engagement (sessions per user, watch time per session), retention (returning users), and resurrection. Determine which stage(s) show the most significant change.

4. Form and Test Hypotheses

Generate hypotheses for the root cause (e.g., product change, bug, seasonality, competitor launch) and run analyses or A/B tests to confirm. Use cohort analysis, time-series decomposition, and holdout experiments where possible.

5. Quantify Impact and Recommend Actions

Estimate the size of each contributing factor and prioritize fixes. Communicate findings with clear visualizations and propose next steps, such as rolling back a change or running a targeted experiment.

Key Points to Mention

Data validation: check for logging errors, pipeline failures, or metric definition changes before assuming a real user behavior shift.
Segmentation dimensions: platform, geography, user tenure, acquisition channel, content type, and device type to isolate the drop.
Funnel stages: acquisition, activation, engagement, retention, and resurrection—identify which stage(s) are affected.
Cohort analysis: compare behavior of new vs. existing users, and pre- vs. post-change cohorts to pinpoint when the drop started.
External factors: seasonality, holidays, competitor launches, or macroeconomic trends that could explain the decline.
Experimentation: use A/B tests or holdout groups to confirm causality, and consider quasi-experimental methods if randomization isn't possible.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.