Lyft Product Design: Arrival-Time Estimates in the Driver Experience
Company: Lyft
Role: Product Manager
Category: Product Design & Strategy
Difficulty: medium
Interview Round: Other
You are a product manager at Lyft. Design the ETA (estimated time of arrival) system that Lyft drivers rely on: decide what it should achieve, which drivers and which moments in a shift it should serve first, what you would build, and how you would know it is working.
```hint Find the decisions
List the points in a driver's shift where a time estimate changes what the driver does next, and ask what a wrong estimate costs the driver at each one.
```
### Clarifying Questions
- Does "for drivers" cover only the estimates drivers see, such as the time to reach a rider and the trip length when a request arrives, or also the driver arrival time riders see, which comes from the same predictions?
- Are we redesigning the estimates Lyft already shows, or building the capability from scratch?
- Which outcome matters most here: driver earnings and retention, rider wait times, fewer cancellations, or overall marketplace efficiency?
- Is the prediction model itself in scope, or only the product experience built on top of it?
### What a Strong Answer Covers
- A stated goal for driver-facing ETAs and why their accuracy matters to drivers, riders and Lyft's marketplace
- The moments in a shift where an estimate drives a decision, with driver segments compared and one chosen for a reason
- Pain points for that segment ranked by what they cost, and solutions prioritized with explicit criteria
- How the estimates are produced and how their accuracy is judged, at the level a PM owns
- A primary metric with guardrails for riders and the marketplace, plus risks such as driver distraction and knock-on effects on dispatch
### Follow-up Questions
- If more honest pickup estimates lead drivers to decline more long pickups, what happens to riders in low-supply areas, and how would you respond?
- How would you test a new ETA model when its estimates also influence which driver gets matched to which rider?
- How would the design change at airports or stadiums, where the last stretch to the rider is hardest to predict?
Overview: A Lyft product-design case in which a PM candidate designs the arrival-time estimates drivers depend on during a shift. It tests goal setting, picking the driver segment and decision moments that matter, prioritizing solutions, judging prediction accuracy, and choosing metrics with rider and marketplace guardrails.