Scenario
A consumer messaging app is considering launching an in-app Video Call feature. You have access to full historical user and call data (e.g., messaging, audio call usage, device/network attributes, and user engagement/retention).
Questions
-
Using the available data, how would you determine whether users need an in-app Video Call feature before launch?
-
After launch, what success metrics and guard-rail metrics would you track to evaluate feature quality?
-
Describe step-by-step how you would design and run an A/B test for this feature.
-
What are the engineering pros and cons of enforcing a hard hang-off (automatic call termination) policy, and how might that affect experiment results?
Hint: Think funnel, engagement, retention, power-user segmentation, experiment unit, sample size, run duration, and potential user experience trade-offs.