Uber — Brand Science — A/B Test & Causal Inference Questions

Role context: Applied Scientist, Brand Science · Est. study time: 60 min · 5 questions

How measurement works here

Uber marketing measurement is causal and marketplace-shaped:

  • Marketplace features can't be user-randomized (treatment and control trips share the driver pool) → switchback (city × time).
  • Brand/TV can't be user-randomized → geo experiments and synthetic control across cities.
  • MMM covers the whole budget but must be calibrated by experiments (experiments are ground truth).
  • Everything is incremental gross bookings, read net of the supply constraint; few, noisy markets make inference delicate.

For the fundamentals — p-values, power, distributions, bootstrapping — see the Probability & Statistics section.

Each answer is a coaching walkthrough: a Sample answer (clarify → approach → a simulated back-and-forth → a clear call), then a Deep dive with illustrative example, then a Grading rubric.

Questions (5)

A marketplace feature can't be user-randomized — switchback designExperiment designHardBuild an MMM you can trust — calibrate it with experimentsCausal inferenceHardNo A/B for a city-level campaign — synthetic controlCausal inferenceHardDoes brand awareness actually cause trips?Causal inferenceHardYour geo test shows +6% but you have 30 markets — believe it?Decide from resultsMedium