Diagnosing where prospective drivers drop out of the signup funnel and which metrics the team should track.
Background
Midterm project for Applied Analytics. As a product analyst on the driver-signup team, the goal was to get drivers who start signing up to complete their first trip. Part one answered SQL requests from a manager (ETA prediction error by city, and the share of early-2026 signups completing a trip within 168 hours). Part two analyzed a driver_signup snapshot to assess data quality, first-trip conversion, time to first trip, and the drivers of activation.
My role
Generated a synthetic dataset matching the production schema to develop and validate the part-one SQL queries.
Audited data quality and missing values in BigQuery, then built funnel statistics showing the share of drivers at each signup stage.
Segmented signups by city, channel, and OS, analyzed the distribution of days to first trip, and defined the team's primary KPI plus supporting metrics.
Built a multi-page Looker dashboard and a slide deck presenting findings and product recommendations.
Deliverable
Slide deck (PDF)
The slide deck opens best in your device's PDF viewer.
Results
A good north-star metric for a funnel must be both measurable from the available data and directly movable by the team.
Validating queries on synthetic data first made it safe to iterate quickly without production access.