Do not roll back or redesign on day one. Confirm the measurement is real, then segment by user type, device, and tenure, because a novelty dip among existing users looks identical to a broken design in the aggregate. Watch session recordings and support volume for the specific failure. Decide within an agreed window whether it recovers, needs a fix, or needs reverting.
Why interviewers ask this
This is a composure and rigor test. Interviewers have seen designers panic-revert good work after two days, and seen others defend a failing design for a month. They want a diagnostic sequence, awareness of the learning curve effect on existing users, and a pre-agreed decision rule so the outcome is not decided by whoever is loudest in the Monday meeting.
How to structure your answer
- Verify the tracking before believing the drop.
- Segment by new versus existing users, device, and plan.
- Watch recordings of failed sessions to find the exact moment.
- Separate a temporary learning dip from a real regression.
- Hold to the decision rule and timeline you agreed before launch.
Example answer
First thing, I check that we are measuring the same thing we were measuring last week. Redesigns change the markup and the event triggers, and a broken event looks exactly like a catastrophe. That has caught me twice. Then I segment, and the split I care about most is new versus existing users. If new users are fine and existing users dropped, that is very likely relearning and it usually recovers over two to three weeks. If new users dropped too, the design is genuinely worse and I want to know where. So I watch twenty recordings of sessions that failed the task, which is unpleasant and extremely fast at finding the truth. On one launch it was a single control that no longer read as clickable because we had made it too quiet, a one-day fix rather than a rollback. The important part is agreeing the decision rule before launch: what size of drop triggers a revert and how long we wait. Deciding that while everyone is anxious and the number is red is how good work gets thrown away.
Walking into this interview soon? GhostPilot listens to your live call, spots the question the moment it is asked, and puts a structured answer on your screen in real time. Try it on your next mock, or grab a $29 Session Pass, no subscription, for the real thing.
See how it worksFollow-up questions to expect
- How long do you give a learning curve before calling it a regression?
- Who makes the call on rolling back?
- How do you keep stakeholders calm during that week?
Related ux designer questions
Your interviewer will ask their own version of this. Paste your actual job description into the free Question Predictor and get the 20 questions that role is most likely to ask, with what each one is really probing.
Predict my questions