Scenario

You work on an e-commerce product where an A/B test of a checkout or merchandising change shows higher conversion in treatment, but lower average order value. The team is encouraged by the conversion lift, but there is concern that the test may be driving smaller baskets and hurting overall business value.

Question

What would you recommend if an A/B test shows a positive impact on conversion but a negative impact on average order value? How would you decide whether to ship, iterate, or stop the change?