Improve Estimated Time of Arrival for Uber Riders

##### Scenario Uber ride-hailing: understanding and improving the Estimated Time of Arrival (ETA) shown to riders. ##### Question List driver-, rider-, and platform-level factors that can affect displayed ETA. From a product-management perspective, why is ETA an important metric? You observe a positive correlation between ETA length and ride conversion rate; provide possible internal and external explanations. Design an experiment to measure the causal impact of ETA on rider conversion. How would you choose the switchback window length in that experiment and what are the trade-offs of shorter versus longer windows? ##### Hints Think causal vs. correlational, switchback vs. A/B, geographic or temporal randomization, metric sensitivity, supply–demand dynamics.