Practice real data science interviews tuned by the interviewers who actually run them.

Create a free account and your first session is on us — then $3 per session, no subscription.

75
Session complete
Friend Content in Newsfeed · Meta · Product Sense
01Define ‘enough’84Strong
02Diagnose the gap88Strong
03Segment the analysis45Weak
04Design the experiment81Strong
Top growth area

Your one soft spot is segmentation. You reasoned well in aggregate but didn’t split the problem — break it out by user type and network size (new users vs. power users with large networks) before drawing conclusions.

Real case study interviews from

Meta
Google
Uber
Spotify
TikTok
DoorDash
Netflix
7 cases

Diagnosing Metric Changes

Investigate drops, decompose metrics, find root causes

GoogleProduct SenseHard~30 min
YouTube Watch Time Drop

YouTube's total watch time dropped 8% week-over-week. The VP of Product wants a root cause analysis before the quarterly business review. Multiple platform, regional, and content-type dimensions may be driving the decline, and early data suggests the drop is concentrated on mobile.

Metric decomposition across multiple dimensionsPlatform-specific hypothesis generationDistinguishing product vs. external causesPrioritized investigation framework
Start session →
GoogleProduct SenseHard~30 min
Google Play Privacy vs Engagement

Google Play's privacy perception score improved 8% after launching prominent privacy labels, but Play Store engagement (app installs, browse time, return visits) dropped 14% in the same period. The VP of Trust & Safety wants to understand whether the privacy initiative is hurting engagement or if this is coincidental.

Disentangling correlated metricsCausal vs. coincidental analysisUser segment-level diagnosisPrivacy-engagement tradeoff reasoning
Start session →
GoogleProduct SenseHard~30 min
Google Search Simpson's Paradox

Average searches per user declined 4% globally, but average searches per user increased in every individual country. The CEO asks why the global number is going down if every country is improving. You need to diagnose and explain this paradox clearly enough for an executive audience.

Simpson's Paradox recognitionCompositional shift analysisExecutive-level metric communicationGrowth-stage metric framing
Start session →
GoogleProduct SenseMedium~25 min
Google Maps DAU Drop

Google Maps DAU dropped 10% over the past 2 weeks. The VP of Geo wants an explanation before the exec review. The drop could stem from a recent iOS update, a product change, an external factor, or a data quality issue. You need to systematically rule out hypotheses.

Systematic hypothesis eliminationPlatform-specific diagnosisData quality vs. product vs. external cause distinctionStructured communication under time pressure
Start session →