← All questions

Data intuition & product sense

Framework → metrics → experiments → modeling cases → advanced

Why this sequence

Product-sense questions live or die on structure, not raw domain knowledge. Start with the project-intake and metric-design framework, practice big-tech product exemplars, then metric-trap cases and RCA, then applied modeling and ranking problems. Ends on Ads-team-specific depth.

37 questions · 0 mastered · 0 in review
Start here
  1. 01
    A Stakeholder Asks You to Run a Project — What Questions Do You Ask Them First?
    Q047 Medium High freq
    New
  2. 02
    If You Could Only Choose One Metric to Evaluate a New Product, Which Would It Be?
    Q045 Medium High freq
    New
  3. 03
    Favorite Google Product — How Would You Improve It and Measure Success?
    Q046 Medium High freq
    New
  4. 04
    Should YouTube Add a Bonus to Creator Ad Revenue?
    Q053 Medium High freq
    New
  5. 05
    A/B Test Metric Selection — Change a Button's Color
    Q026 Medium High freq
    New
  6. 06
    Estimate the Lifetime Value (LTV) of a User or an Ad Click
    Q042 Medium High freq
    New
  7. 07
    If You Had No Metrics for Google Docs, Which Five Would You Define First?
    Q102 Medium High freq
    New
  8. 08
    Design a Metric for Bird Species Segregation in a Forest
    Q066 Hard Medium freq
    New
  9. 09
    Car X vs. Car Y Fuel Efficiency — Which Technology Saves More Gas?
    Q048 Medium High freq
    New
  10. 10
    Compare the Profit of Two Stores Near a School Gate
    Q024 Medium High freq
    New
  11. 11
    A Metric Suddenly Changed Overnight — What Would You Do?
    Q039 Medium High freq
    New
  12. 12
    Why Do Customers Prefer the 4.9-Rated Product Over the 5.0-Rated Product?
    Q091 Medium High freq
    New
  13. 13
    Survey with 4 Options — Which Is the Most Preferred, and Is It Significantly So?
    Q050 Medium High freq
    New
  14. 14
    Design a Database Schema for a Video Company
    Q029 Medium High freq
    New
  15. 15
    Google Maps Jogging Route Recommendation — Experiment Design
    Q018 Hard High freq
    New
  16. 16
    Google Meet Enterprise Clients Complaining About Frequent Disconnections
    Q017 Hard High freq
    New
  17. 17
    You Have 1000 Features and Want to Estimate Conversion Rate — What Do You Do?
    Q059 Medium High freq
    New
  18. 18
    Predict Remaining Phone Battery Life — Case Study
    Q019 Medium High freq
    New
  19. 19
    How Would You Predict the Number of Views for a Video?
    Q009 Medium High freq
    New
  20. 20
    Spot the Modeling Mistakes
    Q021 Medium High freq
    New
  21. 21
    Bias-Variance Tradeoff — What Do You Do When Your Model Is Much Better on Training Than Test?
    Q058 Medium High freq
    New
  22. 22
    When Does K-Means Fail? What Are the Estimation Problems and How Do You Fix Them?
    Q088 Medium High freq
    New
  23. 23
    How Would You Handle a Highly Imbalanced Dataset?
    Q028 Hard High freq
    New
  24. 24
    Model the Goal-Scoring Probability at Every Location on a Football Pitch
    Q005 Hard Medium freq
    New
  25. 25
    Implement Evaluation Metrics and Bootstrap CI (PR Curve + Percentage RMSE)
    Q074 Medium High freq
    New
  26. 26
    Evaluating an Ads CTR Prediction Model — Metric Selection, Traps, and Fixes
    Q049 Hard High freq
    New
  27. 27
    Evaluate a New YouTube Auto-Playlist Ranking Algorithm
    Q082 Hard High freq
    New
  28. 28
    Sample Ratio Mismatch (SRM) — How to Detect, Diagnose, and Handle
    Q093 Medium High freq
    New
  29. 29
    Hypothesis Testing with Only 20 Samples — What Do You Do?
    Q087 Medium High freq
    New
  30. 30
    Mean vs Median — Which Is Better? Force a Choice.
    Q076 Medium High freq
    New
  31. 31
    Why Divide by n − 1 in Sample Variance? (Bessel's Correction)
    Q063 Medium High freq
    New
  32. 32
    Is the Sample Standard Deviation an Unbiased Estimator of σ?
    Q064 Hard High freq
    New
  33. 33
    Linear Regression When You Have More Features Than Observations (m > n)
    Q070 Hard High freq
    New
  34. 34
    Ad Auction Mechanics — Second-Price, First-Price, and the Role of Quality Score
    Q098 Medium High freq
    New
  35. 35
    From a 1% Query Sample, Estimate the Number of Singleton Queries
    Q065 Hard High freq
    New
  36. 36
    Estimate the Number of Distinct Queries from a 1% Sample
    Q079 Hard High freq
    New
  37. 37
    How to Allocate K Human Reviews to Estimate ML Model Accuracy
    Q010 Hard High freq
    New