Redesign Experiment: Energy-Usage Dashboard
Context
A redesigned energy-usage dashboard is being launched in a consumer app. The goal is to measure whether the redesign increases user engagement with the dashboard.
Assume:
-
Unit of randomization is the user (logged-in ID).
-
Only eligible, active users who can see the dashboard are included.
-
Traffic and seasonality vary by day of week.
Task
Design an A/B test to determine whether the new dashboard increases user engagement.
Address the following:
-
State clear hypotheses and define the unit of randomization and exposure.
-
Specify primary and secondary metrics, plus guardrail metrics.
-
Describe the statistical methods for analysis (p-values, confidence intervals, variance reduction).
-
Compute/justify sample size and determine test duration.
-
Include practical checks (randomization, instrumentation, novelty/learning effects).