Diagnose Retail Revenue Drop and Predict Ad Impact
Overview
This question evaluates top-down funnel estimation and causal impact measurement for marketing lift as well as structured revenue decomposition and root-cause analysis in retail and consumer fintech contexts.
Scenario
You are a data scientist for a consumer fintech app preparing to run a Super Bowl TV ad and investigating a recent revenue decline.
Part 1 — Super Bowl Ad Lift Estimation
Estimate how many users will register as a result of a single Super Bowl advertisement. Build a clear, top-down funnel and justify every assumption you use. Provide a base case and a sensitivity range (e.g., low/base/high) and explain how you would validate the estimate after the event.
Hints:
- Think reach → attention → site/app visit → install → registration (and optionally KYC).
- Use realistic conversion rates and call out any tail effects (same day vs. 7–14 day follow-on).
Part 2 — Revenue Drop Diagnosis
Retail revenue has fallen. Use a structured framework to diagnose the drop and give concrete example cases that could explain it.
Hints:
- Decompose revenue. For general retail: Revenue ≈ Traffic × Conversion Rate (CVR) × Average Order Value (AOV).
- For a consumer trading/fintech app, tailor the decomposition (e.g., active users × funding/deposit rate × trades per user × average trade size × take rate).
- Show how you would quantify contributions, identify root causes, and propose next steps.
Solution
Login required