Your question is Compare Marketing Message Effectiveness. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You work on a consumer apparel site and are testing two different marketing messages in a campaign. One message may drive more clicks or purchases, but you also want to avoid harming downstream behavior or creating biased results.
How would you compare the effectiveness of the two messages in an A/B test? Explain how you would define success, choose the randomization unit, and decide whether the result is strong enough to ship.