Design Experiment to Measure Airport Surge-Pricing Impact
Overview
This question evaluates expertise in experimental design and causal inference for two-sided marketplaces, including competencies in randomization/unit selection, handling interference and spillovers, clustering, timing windows, and choice of primary and secondary metrics.
Experiment Design: Causal Impact of Airport Surge-Pricing Push Notifications on Driver Supply
Context
You operate a two-sided ride-hailing marketplace. A new push notification is sent to eligible drivers when the airport is in surge, aiming to attract more drivers to the airport. Drivers within and around the airport can see and respond to the push at overlapping times, so interference (spillovers) between treated and untreated drivers is plausible.
Task
Design an experiment to measure whether the push notification causally increases driver supply at the airport while handling potential interference.
Please address:
- Experimental design: randomization unit(s), holdouts, timing windows, and any clustering.
- Primary and secondary success metrics to track.
- How you will detect and account for spillovers on untreated drivers.
- How you will identify causal impact in the presence of interference (e.g., geographic clustering, holdout zones, difference-in-differences, network interference adjustments).
Solution
Login required