I started with the funnel, which felt right, but I jumped to hypotheses way too fast before ruling out data issues or external factors.
Start by clarifying the metric definition and time frame, then segment the conversion funnel to isolate where the drop occurs. Form hypotheses across internal changes, external factors, and user behavior, and prioritize validation through data and experiments.
Pro tip: Acknowledge that 23andMe's conversion is likely tied to the kit activation funnel, not just purchase—so diagnose both purchase and activation rates. Also, consider seasonality and marketing campaign shifts, as DNA kit sales spike during holidays and promotions.
Define what 'conversion rate' means (e.g., visitor-to-purchase, purchase-to-activation) and the time period. Confirm if the drop is sudden or gradual, and if it's company-wide or segment-specific.
Break down the conversion funnel by traffic source, device, geography, user demographics, and product SKU. Compare current vs. previous periods to pinpoint where the drop is largest.
List potential causes: internal (site changes, pricing, bugs), external (competition, seasonality, PR), and behavioral (user intent shifts). Prioritize by likelihood and impact.
Use analytics, A/B tests, user surveys, and session recordings to confirm or rule out hypotheses. Check for correlations with deployments, marketing campaigns, or news events.
Based on findings, propose fixes or further investigations. Outline how to monitor the metric and prevent future drops.
AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.