Your question is Support-Process Test with Retention Guardrails. 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 customer support experience where a new support process is being considered for rollout. Early feedback suggests it may improve satisfaction, but the team is worried it could create friction that hurts retention.
How would you design and analyze an A/B test to determine whether the new support process improves satisfaction without negatively affecting retention?