Handle it in two phases. Immediately: reproduce it, assess impact and scope, support the mitigation (rollback or flag off), and communicate. Afterward: work out why testing missed it, whether it was a missing case, a gap between environments, or an untested integration, then add a regression test at the lowest useful layer and fix the class of gap, not just the instance.
Why interviewers ask this
Escaped defects are inevitable, so interviewers care about your response rather than your record. They want containment before analysis, and a blameless root cause that produces a change in the process rather than a promise to be more careful. Adding the regression test at the cheapest layer that would have caught it, and asking what class of bug this represents, separates a senior tester from a diligent one.
How to structure your answer
- Separate immediate containment from later analysis.
- Reproduce and quantify impact before anything else.
- Support the fastest safe mitigation.
- Do a blameless analysis of why it was not caught.
- Add a regression test and fix the class, not just the case.
Example answer
First phase is containment and I am not analyzing anything yet. Reproduce it, pin down the exact conditions, and quantify: how many users, which paths, is data being corrupted or just displayed wrong, because that changes the urgency completely. Then support the fastest safe mitigation, usually flagging it off or rolling back, and make sure whoever handles customers knows what to tell them. The analysis comes after it stops hurting. The question I ask is not who missed it, it is what property of our process let it through. Usually it is one of a few things: nobody thought of the case, which is a test design gap; the case existed but the environment differed, which is a data or configuration gap; or it lived in an integration nobody owned. The last one I hit was a currency rounding bug that only appeared for a locale missing from our test data, so the real fix was not one test case, it was adding locale coverage to the data strategy. Then I add a regression test at the lowest layer that would have caught it, and I check whether the same class of gap exists elsewhere.
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 would you present that analysis without anyone feeling blamed?
- What if the same area produces escaped defects repeatedly?
- How do you decide whether the regression test belongs at unit or API level?
Related qa engineer 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