How to Answer 'Tell Me About a Time You Failed' (With Examples)

Bottom line: Every interviewer who asks this question is testing the same three things: do you take accountability, can you learn from setbacks, and are you self-aware enough to reflect honestly? A well-chosen failure story answered with those three elements will always score high — regardless of how significant the failure was.

What the interviewer is actually testing

This is not a trick question. Interviewers don't expect candidates with perfect records — they know careers involve setbacks. What they're looking for is specific: can you own a genuine failure, describe your response to it accurately, and articulate what you actually changed as a result?

The three things this question assesses:

Accountability: Did you take personal responsibility, or did your story subtly blame circumstances, systems, or other people?
Resilience: How did you respond? Did you manage the recovery effectively, or did the failure derail you?
Self-awareness: Do you understand clearly what went wrong and why? Have you actually changed something as a result?

Candidates who score lowest give answers that aren't really failures ("I work too hard"), blame external factors, or show no evidence of change. Candidates who score highest give genuine failures with clear ownership, managed recoveries, and tangible learning.

How to choose the right failure

The failure you choose should be:

Real: Interviewers probe follow-up questions. A vague or constructed story collapses under questioning.
Proportionate: Choose something that shows genuine professional stakes — not a trivial mistake, not a catastrophic ethical failure. Aim for something that cost time, money, a relationship, or a result.
Early enough in your career that the learning is established: a failure from five years ago that you've demonstrably grown from is stronger than a failure from last month with no resolution.
Something you had genuine control over: If the failure was entirely outside your influence, there's no real accountability to demonstrate.

What to avoid

Non-failure answers: "I failed to get the promotion because I cared too much about quality." The interviewer will know you're dodging.
Financial failures with external clients: Stories where your mistake directly cost a client money or damaged a commercial relationship require careful framing — they can raise more questions than they answer.
Failures involving other people's behaviour: "The project failed because my colleague didn't deliver" is not a failure story — it's a blame story.
Failures with no resolution: If the situation is still unresolved, ongoing, or you genuinely haven't changed anything, choose a different example.

The framework: four parts, not three

Standard STAR works here — but add a fourth element explicitly:

Situation and Task: What were you trying to deliver and what went wrong?
Action: How did you respond when you realised you had failed?
Result: What was the outcome? (It won't be entirely positive — and that's fine.)
Learning: What specifically changed in how you work as a result? This is the most important part, and most candidates skip it or make it vague.

Five worked examples by career level

Example 1: Early career — missed deadline

Situation: In my second year of work, I was responsible for producing the monthly data report for our sales team. I underestimated how long a data cleansing step would take and missed the submission deadline by two days.

Action: I told my manager as soon as I knew I wasn't going to make it — not at the deadline, but 24 hours before. I worked late the following two evenings to complete it, communicated the delay to the report users directly, and delivered it with a note explaining the issue.

Result: The delay affected one sales team planning meeting. My manager was frustrated but appreciated the early communication. No data decisions were made on incomplete information.

Learning: I now build a 20% buffer into any estimate involving data preparation, and I flag risk to my manager when I'm more than 30% through a task and behind pace — not when I've already missed the deadline.

Example 2: Mid-career — failed project launch

Situation: I led the rollout of a new internal process across four offices. I planned the training sessions, wrote the documentation, and set the go-live date. The process went live on schedule — and within a week, two offices had partially reverted to the old method because the training hadn't addressed their specific workflow variations.

Action: I went back to both offices personally, ran listening sessions with the team leaders, identified the three workflow gaps I'd missed, and redesigned the training for those contexts. I also paused the rollout in the remaining two offices until the revised training was ready.

Result: The full rollout completed six weeks later than planned. Both initially struggling offices achieved the target process adherence within a month. The delay was frustrating, but we avoided a partial implementation that would have been much harder to unpick later.

Learning: I now run a discovery session with each end-user group before designing any process training. I assumed consistency across offices — I should have verified it.

Example 3: Management — hiring mistake

Situation: I hired a team member who I assessed as technically strong but who struggled significantly with the collaborative elements of the role. Within three months it was clear the hire was a poor fit, and the team was absorbing the friction. The decision was mine — I had overweighted technical ability in the interview and underweighted the behavioural indicators I saw but rationalised away.

Action: I managed the performance situation directly and honestly, worked with HR on a structured support plan, and ultimately the team member moved to a more individual-contributor-focused role in another department where they performed well. I also reviewed my interview scorecard and rewrote the behavioural assessment criteria.

Result: The team member found a better-fitting role. The team recovered quickly once the dynamic changed. The revised scorecard has been used for all subsequent hires in my team.

Learning: I learned to trust my observations of behavioural signals in interviews — not just technical performance. I was seeing the indicators of a mismatch and talking myself out of them. I no longer do that.

Example 4: Sales / commercial — lost deal

Situation: I was leading a pitch for a significant new contract — worth around £180,000 annually. We lost on price in the final round to a competitor. In the debrief, the client told us our proposal had been their preferred option commercially — but our pricing was 12% above the competitor's.

Action: I requested a full debrief with the client, which they agreed to. I used that meeting to understand exactly where our pricing model had diverged from their expectations and what they valued most in the solution.

Result: We lost the contract. The client remained warm and came back to us 18 months later when the incumbent underdelivered. We won the renewal at a price that reflected the learning.

Learning: I now run a pricing alignment conversation with procurement stakeholders at least four weeks before final submission — not to negotiate early, but to calibrate before we anchor on a number in the proposal.

Example 5: Technical / analytical — wrong recommendation

Situation: I built an analysis to support a marketing budget decision, recommending a 30% increase in spend on one channel based on attribution data I had modelled. The recommendation was implemented and the results were significantly below expectation. Post-implementation analysis showed my attribution model had over-credited last-touch conversions and underweighted upper-funnel activity.

Action: I flagged the issue internally as soon as the first month's results came in, rebuilt the model with a corrected multi-touch attribution approach, and presented the revised analysis to the marketing director with a clear explanation of the error and how the new model differed.

Result: The channel spend was reduced back to its original level the following quarter. The revised model became the team's standard attribution framework. The budget error cost approximately £40,000 in suboptimal spend.

Learning: I now require a second analyst to review attribution model assumptions before any model outputs influence budget decisions above a specific threshold.

Frequently asked questions

What if my failure was very recent?

You can use a recent failure if the learning is clearly established. The question is whether you can show that something has genuinely changed — not just that you feel bad about it. If the situation is still too raw to articulate a clear learning, choose an older example.

How long should my answer be?

Two to three minutes. The Learning element should take at least 30 seconds — it's where most candidates cut short. Don't rush the "what I changed" part — that's what the interviewer most wants to hear.

Can I use a personal failure rather than a professional one?

Only if you're early career with limited professional experience. For most experienced candidates, a personal failure in a work interview can feel mismatched and hard to link back to the role. Stick to professional or academic examples where possible.

What if the failure was partly someone else's fault?

You can acknowledge shared context briefly — but your answer must focus on your decisions, your response, and your learning. The moment your answer shifts to what others did wrong, you've stopped answering the question.