Pinterest — Product Data Science — Product Case Questions

Role context: Data Scientist (Product), Pinterest · Est. study time: 60 min · 5 questions

How to approach Pinterest product cases

Every case is one move in a single chain: understand the goal → decompose into a data problem → pick the metric or method → name the bias and trade-off → land on a decision and an action. Pinterest's flavor is specific, and the strongest answers use its real vocabulary rather than generic "engagement":

• The core action is the repin (save); the honest guardrail is the hide. Read them together.
• Blended numbers hide the story. Split US vs International (ARPU differs ~10×) and core vs non-core (new/casual/resurrected) users before theorizing.
• Saves are a leading indicator, not the goal — the goal is durable retention and off-platform action, so watch whether a save bump lasts.
• Ranking/ML launches show novelty decay, so read the days-in curve, not the week-1 peak.
• The big monetization levers (ad load, pricing, brand campaigns) often can't be cleanly user-randomized, so you reason from quasi-experiments and separate incremental from gross.

These map to the four case shapes — diagnose, measure success, launch-or-not, and measure impact without a clean experiment — and the five below span all four.

Each answer is a coaching walkthrough: a Sample answer (clarify the question → lay out the approach → a simulated back-and-forth → a clear call), then a Deep dive with illustrative example with the numbers you'd actually pull, then a Grading rubric.