How would you defend your choice of success metrics for Uber Eats if a panelist argues that your primary metric negatively impacts courier retention?
