Design and Evaluate a Home Carousel

Pinterest is weighing a horizontally scrollable carousel placed at the top of the Home feed, in the style of Instagram Stories. The carousel could highlight recently saved Pins, recommended Pins, or topic-based collections. As a Data Scientist, describe how you would assess this feature through product, analytics, and experimentation lenses. Your response should cover:

1. Which product goal or goals you would clarify first, and how those goals would reshape your evaluation approach.
2. The primary, secondary, and guardrail metrics you would choose. Be specific about trade-offs among metrics such as module CTR, overall Home CTR, saves, downstream engagement, session depth, retention, and advertising or monetization impact.
3. The events, dimensions, and logging you would set up to support the analytics system for this feature.
4. How you would plan an A/B test or staged rollout, covering the randomization unit, exposure definition, novelty effects, power / minimum detectable effect, and likely sources of bias or confounding.
5. If the feature ships and overall Home-page CTR drops, what are the most likely explanations, and how would you tell apart a logging bug, a denominator effect, cannibalization of the existing feed, segment-specific effects, or a genuine product regression? Overview: This question tests a candidate's skill in product analytics, experiment design, metric selection, and instrumentation for measuring the effect of a new UI feature.