How to Target Coupon Users

A ride-hailing company like Lyft is planning a coupon promotion to grow commuter trips, but the available coupon spend is constrained. How would you examine the rider population to decide which users should be offered the coupon?

In your response, cover:

• How you would set the business objective, for instance lifting incremental rides, revenue, or long-term rider retention.
• Which rider groups you would look at, such as regular commuters, dormant riders, cost-conscious riders, new users, or riders who are likely to ride even without an incentive.
• Which historical data you would draw on, including trip frequency, usage during commute hours, origin-destination routes, fare levels, past coupon redemption, rider tenure, and retention.
• How you would separate riders who are likely to ride because of the coupon from those who would have taken the trip anyway.
• Which analytical or modeling method you would apply, for example exploratory segmentation, propensity modeling, or uplift modeling.
• How you would assess whether the targeting approach is effective, covering experiment design, success metrics, cannibalization risk, and return on investment.
• Which trade-offs or risks you would monitor, including selection bias, subsidy abuse, unfair targeting, or short-term gains that fade.

Assume the company can randomly assign coupons to some eligible riders if measurement requires it.

Overview: This question tests a data scientist's ability in causal inference, uplift and propensity modeling, user segmentation, and experimental design for targeted marketing campaigns.