Uber Instant Pay: Defining Success Metrics for On-Demand Driver Cash-Outs
Company: Uber
Role: Product Manager
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Other
Uber offers drivers a feature called Instant Pay: instead of waiting for the regular weekly payout, a driver can send the earnings already made to their own card or account at any time. You are the product manager who owns Instant Pay. How would you measure whether it is a success?
Define what success means for this feature, choose the metric you would treat as primary, and add the supporting and guardrail metrics you would watch.
```hint Start from why it was built
Drivers were already being paid every week. Ask what changes for a driver, and for Uber's supply of drivers, when money can be withdrawn at any time, and let that goal pick your primary metric rather than how often the cash-out button is pressed.
```
### Clarifying Questions
- Does a cash-out cost the driver anything, and does the fee depend on where the money goes, for example a partner debit card versus any other card or account?
- Is Instant Pay open to every driver, and to delivery couriers, in all markets, or is it launching in one region first?
- Is the goal mainly to keep drivers on Uber and driving more, to attract new drivers, or to earn fee revenue?
- Are there limits on how often or how much a driver can cash out in a day?
### What a Strong Answer Covers
- A stated goal that links drivers' need for fast access to their earnings with Uber's need for a reliable supply of drivers.
- A driver journey from setup to first and repeat cash-outs, with a metric for each stage.
- A primary metric tied to driver retention or hours online, and a reason for preferring it over counts of cash-outs.
- Guardrails covering failed or slow payouts, fraud and losses on money paid out early, fee complaints and support contacts.
- A way to separate the feature's real effect from the fact that the most committed drivers are the ones most likely to use it.
### Follow-up Questions
- Drivers who use Instant Pay churn less than drivers who do not. Why does that not prove the feature works, and how would you measure its true effect?
- Cash-outs per driver climb quickly, but hours online stay flat. How do you read that result?
- If cash-outs to outside cards carried a fee, how would you weigh the fee revenue against driver goodwill?
Overview: A product metrics question about Uber's Instant Pay feature, which lets drivers cash out their earnings on demand instead of waiting for a weekly payout. It tests whether a candidate can tie the feature to a clear goal, choose a primary success metric over simple usage counts, set guardrails and separate causal impact from self-selection.