You notice a sudden 15% drop in overall conversion rate across all marketing channels over the past week, but the data shows no obvious changes in traffic sources, campaign spend, or website performance. How would you investigate this ambiguity and propose a data‑driven hypothesis for the drop?

Use the RICE framework to prioritize investigation steps:

1. Reach – Identify all data sources (GA, CRM, server logs) that could influence conversion.
2. Impact – Validate data integrity and look for anomalies in key metrics.
3. Confidence – Segment by channel, device, and cohort to isolate patterns.
4. Effort – Design quick hypothesis tests (A/B or cohort analysis) and iterate. Explain each step in 30‑35 words, totaling 120‑150 words.

First, I map all potential data sources—GA, CRM, server logs—and confirm their integrity using anomaly detection tools. Next, I segment the funnel by channel, device, and cohort to isolate where the drop occurs. I apply the RICE framework: Reach (all traffic), Impact (conversion loss), Confidence (data quality), Effort (analysis time). I hypothesize that a recent mobile UI change or server latency spike caused the drop. I design a quick cohort test comparing pre‑change and post‑change users, and run an A/B test on the affected segment. I collaborate with product and engineering to roll back the UI change and monitor real‑time metrics. Finally, I report findings to stakeholders, recommending a permanent fix and setting up alerts for future anomalies. This structured approach ensures data integrity, rapid hypothesis testing, and cross‑functional alignment.