What STAR stands for

STAR is a four-part structure for answering a question about something you did.

Situation. Where you were and what was going on, in one or two sentences. A date, a role, a company size. Enough for the listener to picture it and no more.

Task. What you specifically were responsible for. Not what the team was doing. The word your interviewer is listening for here is I.

Action. What you actually did, step by step, in the first person. This is the longest part and should be roughly half the answer.

Result. What changed, with a number attached. A percentage, a timeframe, a headcount, a currency figure. If you can only remember one rule from this page, make it this one, because the Result is where most otherwise-good answers quietly stop being checkable.

The whole answer runs ninety seconds to two minutes out loud, which is about 150 to 220 words. If yours is over three minutes, the Situation is too long. That is the most common structural failure in a behavioral round and also the easiest to fix, because the cut is almost always in the first thirty seconds rather than spread through the answer.

A STAR answer, start to finish

Here is one, annotated by part. The question was: tell me about a time you had to fix something that was already broken.

Situation. I joined a fifteen-person support team where first-response time had drifted to about eleven hours against a four-hour target, and it had been missed for two quarters. Task. I was asked to get it back under four hours without adding headcount. Action. I pulled three months of tickets and found that sixty percent of the delay was sitting in one queue, unassigned overnight, because routing rules had never been updated after a reorganisation. I rewrote the routing so overnight tickets went to the team in the other timezone, and I set up a daily check on the unassigned count. Then I sat with the two agents who owned that queue and rebuilt their morning triage. Result. First response went from eleven hours to three and a half over six weeks, and it held for the next two quarters.

That is a competent STAR answer. It has a number, it uses I where it should, and the Action is genuinely half of it. Read it again and notice how much work the Situation is doing in two sentences: a team size, a metric, a target, and how long it had been missed. Four facts, no scene-setting, no explanation of what a support team does. That is the correct ratio and almost nobody gets there without cutting a first draft in half.

It also breaks under one question, which we will get to.

The same answer, weaker

For contrast, here is what most candidates actually say to that question before they have practised it.

I joined a support team where response times were really poor and morale was suffering because of it. My task was to turn that around. We looked at the data, we identified some bottlenecks in how tickets were being routed, and we worked together to redesign the process. I also spent time coaching the team on triage. As a result things improved significantly and the team was in a much better place by the end of the quarter.

Same project. Same person. It fails on three counts, and they are the three that account for most weak STAR answers.

The Situation has no numbers, so really poor could mean anything from six hours to six days. The Action is entirely plural, we looked, we identified, we worked together, which leaves the interviewer unable to tell whether this person led the analysis or attended the meetings. And the Result is improved significantly, which is a claim rather than a measurement.

The gap between the two versions is not writing ability. Both were produced by the same person about the same project. The difference is that the first one was prepared by somebody who assumed a follow-up was coming.

Where STAR answers break

The format has a structural weakness and it lives in the Action step.

Action asks what you did, and what you did almost always happened alongside other people. So the answer drifts into the plural without anybody deciding to be dishonest. We found the routing problem. We rebuilt the triage. Within two sentences the candidate has described a team's work in the first person plural and attached their own name to the Result.

Which produces the question that ends most STAR answers:

Which part of that was actually you, and not your team?

Notice what the format has done. STAR trained the candidate to compress the Action into a tidy sequence, and compression is exactly what removes the boundary between their work and everybody else's. The better and more rehearsed the STAR answer, the smoother that boundary gets, and the more obvious it is to somebody who asks about it for a living.

There is a second break, and it is quieter. STAR has no step for what went wrong. Situation, Task, Action, Result describes a thing that worked, start to finish, with the awkward middle removed. Interviewers know that real projects contain at least one decision that turned out badly, because they have run projects. So an answer with no error anywhere in it reads as edited rather than accurate, and the follow-up goes looking for the edit. Candidates who have never considered this find themselves improvising a weakness under pressure, which produces either a fake one that fools nobody or a real one they have not thought through.

The fix for both is the same and it is uncomfortable: put the boundary and the error into the Action yourself, before anyone asks. In the example above, that means saying the routing analysis was mine and the triage rebuild was mostly the two agents, and that I spent the first week looking at agent productivity before I checked the routing, which cost about five days.

Say it out loud and the follow-up has nowhere to go, because you have already answered it. Leave it out and the follow-up becomes the interview.

There is a reason this feels wrong to do. Every piece of advice about interviewing tells you to present your strongest case, and volunteering the boundary between your work and your team's feels like giving away credit you earned. In practice it does the opposite. An interviewer who has to extract that boundary discounts everything around it, because they now know you were willing to let it stay blurred. An interviewer who is handed it takes the rest of your answer at face value.

The R most people get wrong

Result without a number is not a Result. It is a claim, and claims are the raw material follow-up questions are made from.

The team was much more efficient. The client was really happy. Things improved a lot. Every one of these is a sentence an interviewer has heard several hundred times and none of them can be checked, so each one converts into a follow-up asking for the number you have just demonstrated you do not have.

If you genuinely cannot use exact figures because they are confidential, use direction and proportion. Response time roughly halved. We went from missing the target every week to missing it twice a quarter. Both are specific enough to be interrogated, which is the point.

The other Result failure is subtler and worth knowing about. Some candidates attach a real number to an outcome they did not cause. Revenue grew forty percent that year is true, sits at the end of a story about a project, and implies a connection that will not survive one question about what else was happening in the business. Use the number closest to the thing you actually changed, even when it is smaller and less impressive, because a small number you can defend beats a large one you cannot.

20 STAR interview questions, with answers

Grouped by what the interviewer is testing rather than by topic, because interviewers think in competencies and the same story often answers three of these. Each entry gives the shape of a strong answer, and the linked pages carry a full worked example, the follow-up that breaks it, and a repaired version that survives.

Failure and mistakes

The most common opening in a behavioral round, and the one candidates sand down the most.

  1. Tell me about a time you failed. Pick a failure with a countable cost and name your own decision inside it. Full answer: tell me about a time you failed.
  2. Tell me about a mistake you made at work. The mistake matters less than the gap between when you knew and when you told someone. Full answer: a mistake you made at work.
  3. Tell me about a time you received negative feedback. Quote the feedback exactly. Softened paraphrases are the tell. Full answer: receiving negative feedback.
  4. What is the biggest lesson you have learned professionally? Attach it to one incident with a date, otherwise it is a value statement. The follow-up is always what changed in how you work, phrased so a former colleague could confirm it.

Conflict and disagreement

  • Tell me about a time you had a conflict with a coworker. Describe what the other person wanted and what they thought you did wrong. Full answer: conflict with a coworker.
  • Tell me about a time you disagreed with your manager. Choose one you lost and executed anyway. Full answer: disagreeing with your manager.
  • How do you handle working with someone difficult? Say what you tried before you concluded they were the problem, and what they would say about you. The version where the other person is simply difficult tells an interviewer you have only heard one account of it.
  • Tell me about a time you had to give someone difficult feedback. The useful detail is how they reacted and what you did in the following weeks. Delivering it once is the easy half, and interviewers know that.

Leadership and initiative

  • Tell me about a time you showed leadership. Name the decision your team disagreed with. Full answer: showing leadership.
  • Tell me about a time you took initiative. Account for the hours it came out of, and whether anyone had told you not to. Full answer: taking initiative.
  • Tell me about a time you influenced someone without authority. Name what you gave up to get it. Full answer: influence without authority.
  • Tell me about a time you went above and beyond. Say what you deprioritised. Full answer: going above and beyond.

Pressure and priorities

  • Tell me about a time you worked under pressure. Name the source. Pressure with no face is just a busy fortnight. Full answer: working under pressure.
  • Tell me about a time you juggled competing priorities. Say which one you dropped and whether its owner agreed. Full answer: competing priorities.
  • Tell me about a time you missed a deadline. The date you knew matters more than the date you missed. Full answer: missing a deadline.
  • Tell me about a time you had to learn something quickly. Say what you deliberately chose not to learn. Full answer: learning something quickly.

Decisions and data

  • Tell me about a difficult decision you made. Difficult means somebody paid, not that the analysis was complex. Full answer: a difficult decision.
  • Tell me about a decision you made with incomplete information. Give your confidence at the time as a number. Full answer: deciding with incomplete information.
  • Tell me about a time you used data to make a decision. Name what the data told you that your instinct did not. Full answer: using data to decide.
  • Tell me about a time you changed your mind about something important. Say what specifically changed it, and how long the gap was between changing your mind privately and saying so out loud. That gap is what the question is measuring.

STAR at Amazon

Amazon runs the most structured version of this format in the industry. Every behavioral question maps to one of sixteen leadership principles, interviewers take near-verbatim notes, and a bar raiser sits in to hold the standard across the loop.

The practical difference is depth. Most interviewers ask one follow-up. Amazon interviewers ask four, and they are trained to keep going until the plural pronouns stop. A STAR answer that would pass elsewhere gets taken apart at Amazon specifically because the Action step has not separated your work from your team's.

The second practical difference is that your answer has to land against a named principle, and you rarely get told which one. An answer about rescuing a launch might be scored on Ownership, on Bias for Action, or on Deliver Results depending on which the interviewer was assigned, and the details that matter differ in each case. Preparing for Amazon means having stories that are specific enough to be re-pointed at whichever principle the question turns out to be testing.

The full mapping of principles to questions, with worked answers, is on Amazon's leadership principles.