Diagnose Causes and Test Hypotheses for Metric Drop

Scenario

The product’s key metric suddenly drops. Stakeholders want a structured diagnosis and experiment plan.

Question

List plausible causes for the performance drop, describe analyses you would run to validate each cause, and design an A/B test to confirm the main hypothesis. Which primary and guardrail metrics would you track and why?

Hints

Think segmentation, funnel breakouts, external factors, and metric hierarchy (north-star vs. health).