Scenario

You work on a marketing campaign for a consumer product and want to test whether a new campaign variant improves downstream behavior versus the current version. Leadership wants a clear recommendation, but you are concerned about defining the right success metric and avoiding a misleading readout.

Question

How would you approach A/B testing for this marketing campaign? Walk through how you would define success, size the test, and decide whether to ship based on the results.