Diagnose Revenue Decline: Key Analyses and Metrics
Overview
This question evaluates a candidate's competency in revenue diagnostics, metric decomposition, cohort and segmentation analysis, and experimentation-aware causal reasoning within product analytics.
Coinbase
Aug 4, 2025, 10:55 AM
Data Scientist
Technical Screen
Analytics & Experimentation
13
0
Diagnose a Sudden Revenue Drop
Scenario
A consumer fintech/trading platform (e.g., mobile/web app with transaction-fee revenue) experiences a noticeable revenue decline over the last 7–14 days compared to the prior comparable period.
Revenue comes primarily from transaction fees on trading volume (take rate × volume), plus smaller streams such as spreads, subscriptions, staking/custody, and other fees. Seasonality (day-of-week), market conditions, and release/experiment calendars can influence user behavior.
Task
- Outline the analyses you would run to diagnose the cause of the revenue decline.
- Specify the first metrics and data cuts you would inspect and why.
- Explain how you would test potential root-cause hypotheses (both experiment-driven and organic changes).
Hints
- Consider dimensions: time, product/feature, asset/pair, geography, device/app version, user cohort/tenure, acquisition channel.
- Think segmentation, external factors (market price/volatility, payment rail/network outages, regulatory events), and experiment-vs-organic changes.
- Aim to quantify contributions via decomposition (volume vs take rate vs mix) and validate with counterfactuals.
Solution
Login required
Show