Interview Prep
How to answer 'tell me about a time you failed.'
The question everyone dreads. Here's what interviewers actually want to hear — and why 'I worked too hard' is the worst possible answer.
June 2026 · 7 min read
Every interviewer asks this question. Almost everyone answers it wrong. They either pick a failure that doesn't look like a failure, or they pick something real and turn the interview into a therapy session. The right answer is a real failure, told with ownership, followed by what you actually learned.
The Framework
The Framework
The 80/20 rule: spend 20% on what went wrong, 80% on what came next.
What they're really asking: This isn't a confession. They want to see how you process failure and what you do with it. The failure itself is just context.
The Framework
Don't pick a fake failure.
What they're really asking: Interviewers have heard "I care too much" a thousand times. The candidate who says "I lost a client because I miscalculated the timeline" is infinitely more credible than one who says "I sometimes struggle to delegate."
The Framework
Take full responsibility. All of it.
What they're really asking: The moment you say "but the engineering team also…" the answer is over. Ownership means ownership — even in a failure story.
For Product Managers
For Product Managers
Tell me about a time a product you owned didn't land the way you expected.
What they're really asking: Did you see the signals? What would you do differently — in how you defined success, validated the idea, or communicated the risk?
For Product Managers
Describe a time you misread what users actually wanted.
What they're really asking: Every PM gets this wrong at some point. They want to see the specific mechanism — what signal led you astray, and what you now use instead.
For Software Engineers
For Software Engineers
Tell me about a production incident you were responsible for.
What they're really asking: The failure isn't disqualifying. Your response to it is. Did you communicate early? Fix it fast? Write a proper postmortem?
For Software Engineers
Describe a technical decision you made that you later regretted.
What they're really asking: Hindsight is fine. What do you know now that you didn't then? How did you update your decision-making process?
For Data Scientists
For Data Scientists
Tell me about a model or analysis that led to a bad business outcome.
What they're really asking: Data science decisions have real consequences. They want to see that you understand impact and have built better habits as a result.
For Data Scientists
Describe a time you communicated findings that turned out to be wrong.
What they're really asking: How quickly did you catch it? How did you communicate the correction? Did you own it or obscure it?
For Consultants
For Consultants
Tell me about a client engagement that didn't go as planned.
What they're really asking: Client work goes wrong. They want to see your judgment — in the moment and in retrospect.
For Consultants
Describe a time a recommendation you made turned out to be wrong.
What they're really asking: Was your framework wrong? Was your data wrong? Both are valid failures — but they require different lessons.
For Everyone
For Everyone
Tell me about a time you underestimated how long something would take.
What they're really asking: Estimation is a skill. They're checking your self-awareness and planning habits, not your calendar.
For Everyone
Tell me about a conflict with a colleague or manager that you handled badly at first.
What they're really asking: Failure isn't always a project. Sometimes it's how you treated someone. The candidate who can admit interpersonal failure often stands out.
Frequently asked questions
What's a good failure story to use?
A real one. Ideally one where you had clear responsibility, the stakes were meaningful, and you learned something specific — not just "I work harder now."
Is it okay to use a personal failure?
Generally no — keep it work-related unless the interviewer explicitly invites personal examples.
How long should my answer be?
2-3 minutes max. Situation in one sentence, your role in one sentence, what went wrong in two sentences, what you did next in three sentences, what you learned in one.
Do Amazon interviewers specifically expect failure stories?
Yes. Explicitly. Several Leadership Principles — Ownership, Learn & Be Curious, Earn Trust — are tested specifically with failure-related questions.
The best failure stories are the ones that sound like they actually happened to a real person. Not catastrophic, not trivial — a real moment where something went wrong, you owned it, and you came out with something specific you now do differently. Bar Raiser builds failure stories from your actual experience, so you're telling your story with your specifics — not borrowing someone else's.
Want answers that actually sound like you?
Bar Raiser drafts your interview answers from your real resume and experience — so they come out in your voice, not a template.