You're working on a product team that wants to launch a new interface change. The feature is meant to improve the user experience, but there is debate about whether it actually helps users or just changes behavior in ways that look positive on the surface. You've been asked to define how to test it before rolling it out broadly.
How would you design an A/B test to measure the impact of a new UI feature?
