Estimate sales impact from reviews causally
Company: Google
Role: Data Scientist
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Technical Screen
Overview: Design an analysis that estimates whether product reviews cause higher sales, or sales simply attract more reviews, using daily product data, the full review log, and a staggered widget rollout. Cover confounders, an experiment on review presentation, a fallback when none is possible, validity checks, and effect-size reporting.