"Tell me about a failure"
The disguised-strength answer fails this question. So does the catastrophe. There is a narrow, honest middle.
This question is asked in every panel and answered badly in most of them. Candidates offer a failure that is secretly a success — "I worked too hard and burned out" — or a failure so trivial it insults the question, or one so catastrophic it becomes the only thing the panel remembers. What the interviewer wants is narrow: evidence that you can see your own mistakes clearly, and that seeing them changed something.
Choose the right failure
- It must be genuinely yours. A story where the vendor failed, or the client was impossible, or leadership deprioritised it, is not a failure story — it is a blame story, and it is heard as one.
- It must have had real consequences — a missed deadline, a bad hire, money lost, a product nobody used. Small enough to survive, large enough to matter.
- It should be far enough back that you have done something with the lesson, and recent enough to still be about who you are now.
- Avoid anything that maps directly onto the core requirement of this job. A failure of judgement is fine; a failure at the exact thing they are hiring for needs a very strong recovery attached.
The shape of the answer
Three beats, and the third is the one that is graded. What happened, in two sentences without hedging. What your part in it was, stated plainly — this is the beat candidates skip. What you changed afterwards, concretely enough that it is checkable.
"I ran a migration that we cut over on a Thursday night with a rollback plan I hadn't tested. It failed halfway, and the rollback failed too, so we were down for about five hours and I spent the night restoring from backup. The cause wasn't the migration — it was that I'd treated the rollback as paperwork rather than as code. Since then I don't approve a cutover plan, mine or anyone's, unless the rollback has been run end to end in staging. I've since caught two broken rollbacks that way."
What makes that work
It names the consequence without minimising it, locates the cause in a decision rather than in circumstances, and ends on a change that is specific and testable — you could ask a former colleague whether he actually does that now, and the answer would be yes. Compare it to "I learned the importance of testing", which is what the same story sounds like when the third beat is skipped.
Do not over-apologise
Tell it evenly. Excessive self-criticism reads as poor judgement in the other direction, and panels notice candidates who seem unable to let a mistake be finished. You made the mistake, you understood it, you changed what you do. That is the whole answer.
Everyone has failed. The question is only whether you can describe yours without a lawyer.