Marketing ran an A/B test on a credit line increase email and saw no overall effect. Then they split the results into 20 customer segments and found the 25-to-30 age group has p = 0.03. They want to roll it out to just that segment. What's your take?