Your question is Conflicting Metric Impacts. 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).
How do you handle cases where an experiment shows a positive impact on one metric but a negative impact on another?
Treat this as a practical experimentation question. Explain how you would establish metric hierarchy, assess statistical and practical significance, investigate trade-offs, and make a ship, iterate, or do not ship recommendation.