Amazon Behavioral Interview Questions and Answers: 60

Every Amazon interview, from the phone screen to the last round of the loop, carries behavioral questions. They are not a warm-up before the technical part. Each one is graded against a specific leadership principle, written up, and read in the debrief where the hiring decision is made. A strong coding round does not rescue a weak answer about a missed deadline.
This post lists sixty Amazon behavioral interview questions sorted by principle, so you can see what each interviewer is listening for and build stories that cover several principles at once. After the list come six worked answers, the difference between the phone screen and the loop, the questions that come up most, and the answer shape Amazon expects.
How Amazon behavioral interviews work
Interviewers at Amazon do not pick questions at random. Before the loop, each interviewer is assigned two or three leadership principles to assess (the exact pattern varies by team, but two or three is what candidates and former interviewers consistently describe). They open with a question aimed at one principle, then follow up until they can write a verdict: what you personally did, what the result was, and what you would change.
Three things follow from that:
- Follow-ups are the real interview. The opening question is generic; the fourth question is where you are judged.
- Your stories are shared. Interviewers write up what you said, and the debrief compares notes. Telling the same story twice is fine; telling it differently is not.
- One interviewer, the Bar Raiser, comes from outside the hiring team to keep the standard above the current average. Their follow-ups go deeper.
For the full loop, from the online assessment to the offer, read the Amazon interview process guide. For what each principle means, the Amazon leadership principles guide has all sixteen in one table.
60 Amazon behavioral interview questions by leadership principle
Each principle below lists three or four questions and one line on what the interviewer is listening for. The full guide for each principle has the meaning, workplace examples, and a sample STAR answer.
Customer Obsession
Listening for: whether you start from the customer's problem or from your own solution, and whether you can name a time the customer's interest cost you something.
- Tell me about a time you went out of your way to help a customer.
- Describe a decision where the customer's interest and the business's interest pulled in different directions.
- Tell me about a time you had to say no to a customer request. How did you decide?
- How do you find out what your customers actually need, rather than what they say they want?
Full guide: Customer Obsession interview questions
Ownership
Listening for: whether you acted beyond your job description and stayed with a problem after it stopped being convenient.
- Tell me about a time you took on something that was not your responsibility.
- Describe a project where you made a long-term decision at a short-term cost.
- Tell me about a time you saw a problem nobody owned. What did you do?
- Give me an example of a commitment you made that you could not keep.
Full guide: Ownership interview questions
Invent and Simplify
Listening for: a real simplification with a measurable effect, not an abstract fondness for innovation.
- Tell me about a process you simplified. What did you remove and what happened?
- Describe the most innovative thing you have built. Why was it needed?
- Tell me about a time an idea of yours was misunderstood or rejected at first.
- When did you borrow an idea from outside your team or company?
Full guide: Invent and Simplify interview questions
Are Right, A Lot
Listening for: judgment under uncertainty, and whether you actively look for evidence that you are wrong.
- Tell me about a decision you made with incomplete information. How did you decide?
- Describe a time you were wrong. How did you find out and what did you do?
- Tell me about a time you changed your mind because of someone else's data.
- How do you know when a decision is good enough to act on?
Full guide: Are Right, A Lot interview questions
Learn and Be Curious
Listening for: concrete evidence that you learn on your own initiative and apply it.
- Tell me about something you taught yourself recently. Why, and what did you do with it?
- Describe a time curiosity led you somewhere useful for your team.
- What is something you learned from a failure that changed how you work?
Full guide: Learn and Be Curious interview questions
Hire and Develop the Best
Listening for: whether you take other people's growth seriously, and whether you can give hard feedback.
- Tell me about someone you helped grow. What did you do specifically?
- Describe a time you gave difficult feedback to a colleague or a report.
- Tell me about a hiring decision you got right, or wrong.
- How do you decide who gets a stretch assignment?
Full guide: Hire and Develop the Best interview questions
Insist on the Highest Standards
Listening for: a standard you held when it would have been easier not to, and how you handled a defect.
- Tell me about a time you refused to ship something that was not good enough.
- Describe a time you raised the quality bar for your team.
- Tell me about a defect that reached customers. What did you change so it stayed fixed?
- When did your standards create friction with someone, and what happened?
Full guide: Insist on the Highest Standards interview questions
Think Big
Listening for: a goal that was larger than the obvious next step, and what you did to make it concrete.
- Tell me about the most ambitious proposal you have made.
- Describe a time you convinced others to pursue a bigger goal than they had planned.
- Tell me about a time thinking small limited a project. What would you do differently?
Full guide: Think Big interview questions
Bias for Action
Listening for: speed with judgment. You moved before you had everything, and you knew which decisions were reversible.
- Tell me about a time you made a decision quickly without all the information.
- Describe a time you took a calculated risk. What was the calculation?
- Tell me about a time waiting for more data would have been the wrong choice.
- When did acting fast go wrong, and what did you learn?
Full guide: Bias for Action interview questions
Frugality
Listening for: results from constraint, not cost-cutting for its own sake.
- Tell me about a time you delivered with fewer resources than you wanted.
- Describe a time a constraint led to a better solution.
- Tell me about a time you decided not to spend money or add people. What happened?
- How do you tell the difference between frugal and cheap?
Full guide: Frugality interview questions
Earn Trust
Listening for: candor, especially about your own mistakes, and whether people chose to rely on you.
- Tell me about a time you had to rebuild trust with someone.
- Describe a time you admitted a mistake in front of your team.
- Tell me about a time you gave feedback that was hard to hear.
- Tell me about a time you told a senior person something they did not want to hear.
Full guide: Earn Trust interview questions
Dive Deep
Listening for: whether you go to the data and the details yourself when a metric and a story disagree.
- Tell me about a time a metric and an anecdote disagreed. What did you do?
- Describe the most complex problem you had to understand from the ground up.
- Tell me about a time you found a root cause others had missed.
- How do you know when you have gone deep enough?
Full guide: Dive Deep interview questions
Have Backbone; Disagree and Commit
Listening for: a respectful, evidence-based disagreement, and real commitment after you lost.
- Tell me about a time you disagreed with your manager. What happened?
- Describe a decision you disagreed with but still had to carry out.
- Tell me about a time you pushed back and were right. Now a time you were wrong.
- How do you disagree with someone more senior than you?
Full guide: Have Backbone; Disagree and Commit interview questions
Deliver Results
Listening for: an outcome you can measure, delivered despite something going wrong.
- Tell me about a time you delivered under a tight deadline. What did you trade off?
- Describe a project that went off track. How did you bring it back?
- Tell me about a goal you missed. What did you do next?
- What is the result you are proudest of, and what was your part in it?
Full guide: Deliver Results interview questions
Strive to be Earth's Best Employer
Listening for: whether you make the people around you safer, more capable, and more likely to stay.
- Tell me about a time you improved the working environment for your team.
- Describe a time you helped someone through a difficult period at work.
- Tell me about a time you noticed a colleague was struggling before they said so.
Full guide: Strive to be Earth's Best Employer interview questions
Success and Scale Bring Broad Responsibility
Listening for: whether you think about the second-order effects of your work on customers, communities, and people outside the team.
- Tell me about a time you considered the wider impact of a decision beyond your team.
- Describe a time you left something better than you found it.
- Tell me about a decision that was right for your team and wrong for someone else. How did you handle it?
Full guide: Success and Scale Bring Broad Responsibility interview questions
Six worked answers
A question list only gets you halfway; most people want to see what a finished answer sounds like. Below are six, one each for Customer Obsession, Earn Trust, Deliver Results, Dive Deep, Have Backbone and Bias for Action. Every one is invented for this post: the people, systems and numbers are made up, and none is a real candidate's story. They are also different from the sixteen stories in the leadership principles examples post, so the two can be read side by side.
Read them for proportion and for the follow-ups listed underneath, which is where an Amazon interviewer would go next. Do not borrow one. The second follow-up asks for a detail only the person who was there can supply.
Customer Obsession
Question. Tell me about a time you had to say no to a customer request. How did you decide?
Example answer (invented). "I was the engineer responsible for the public API of a scheduling product. We had announced that version 1 would be switched off in March, and our largest customer asked us to keep it for another year. That was the request I turned down, because keeping two versions was slowing every release. Before I answered, though, I pulled the call logs. About 400 small integrators were still on version 1, and most had not opened our emails. So the decision I took to my manager was no to the one-year extension and yes to a different thing: I built a thin compatibility layer that translated the six most used version 1 calls to version 2, which took me three weeks, and I moved the switch-off by eight weeks. I phoned the large customer myself to explain. On the new date, 23 integrations broke instead of the roughly 400 we would have broken in March, and support had a list of all 23 in advance."
Follow-ups to expect. How did you pick the six calls? What did the eight-week delay cost the roadmap? What did the large customer say, and did they stay?
Earn Trust
Question. Describe a time you admitted a mistake in front of your team.
Example answer (invented). "I changed a pricing rule for annual plans, and my change rounded a discount the wrong way. For two days, 312 accounts were charged between one and four dollars too much. Nobody had noticed; I found it while checking an unrelated report. My task was simple to state: fix it, and decide how loudly. I posted in the team channel that the bug was mine, with the number of accounts and the total, which was a little under 700 dollars. I reverted the rule the same hour, wrote the refund script, and asked a colleague to review it line by line because I no longer trusted my own reading of that code. Then I wrote a one-page incident note for finance and support with the customer wording I proposed. Refunds went out the next morning with an apology. I also added a test that replays a day of real invoices against any pricing change, and it has since caught two errors from other people before release, including one from my manager."
Follow-ups to expect. Why did review not catch it the first time? Would you have reported it if the amount had been ten dollars? How did finance react?
Deliver Results
Question. Tell me about a time you delivered under a tight deadline. What did you trade off?
Example answer (invented). "I led the mobile team for a retailer's app, and we had a fixed date: the build had to be in the stores before the November sales freeze. Three weeks out, the new version of our payment provider's library started crashing on older Android phones, which were about a fifth of our users. Waiting for the provider's fix would have meant missing the freeze. I split the problem. I asked one engineer to build a fallback that sent those phones to the web checkout inside the app, and I cut two features from the release, saved baskets and a new product filter, after agreeing with the product manager that neither affected sales week. I tested the fallback myself on six old devices we borrowed from the support team. We submitted two days before the freeze. Over sales week the crash-free rate was 99.6 percent and the fallback handled about 18 percent of orders. The two features shipped in January. The trade-off was real: the web checkout converted a few points worse than the native one, and I told the business that before we chose it."
Follow-ups to expect. Who decided which features to cut? What would you have done if the fallback had failed testing? How did you know the conversion gap?
Dive Deep
Question. Tell me about a time you found a root cause others had missed.
Example answer (invented). "Our cloud bill had gone up 40 percent in a quarter while traffic was flat. The explanation going around was that we had more customers on the larger plan, and finance had accepted it. I owned two of the services, so I wanted to see it for myself. I broke the bill down by service and then by line item, which nobody had done because the dashboard only showed the total. Compute was flat. Storage for logs had grown ninefold. I traced it to a debug flag that had been switched on during an incident in the spring and never switched off, on one service that logged every request body. I turned it off in staging first, checked that our alerts did not depend on those logs, then turned it off in production and set a 30-day retention on the old data. The next month's bill was 31 percent lower. I then added a weekly cost report per service to our team review, so the total is no longer the only number anyone sees."
Follow-ups to expect. Why had nobody looked at the line items? How did you confirm the alerts were safe? What else did the per-service report turn up?
Have Backbone; Disagree and Commit
Question. Describe a decision you disagreed with but still had to carry out.
Example answer (invented). "Our director wanted to replace our in-house search with a hosted vendor. I was the senior engineer on search and I disagreed. I wrote three pages: my latency measurements from a trial account, a cost projection at our volume in two years, and what we would lose in ranking control. I presented it at the design review and said plainly that I thought we would regret it. The director's reasons were that two of the four people who understood our search had left and that we could not hire fast enough. The decision was to go with the vendor. I told the review I still expected the cost problem, and that I would lead the migration. I did, and I wrote the exit criteria into the contract review: a latency ceiling and a price per query at which we would reopen the decision. The migration took ten weeks. I was wrong on latency, which came out better than ours. I was right on cost: at eighteen months we hit the price I had projected, and because the threshold was written down, the renegotiation took one meeting and cut the rate by about a quarter."
Follow-ups to expect. What would have changed your mind before the decision? Did you ever tell your team you disagreed? What did you learn from being wrong on latency?
Bias for Action
Question. Tell me about a time waiting for more data would have been the wrong choice.
Example answer (invented). "I was a data engineer, and the scheduler all our pipelines ran on announced that the version we used would lose support in 30 days. There were two realistic replacements. The usual process for a tooling change was a six-week evaluation with a written comparison, which we did not have time for. I asked myself what would make the choice hard to undo, and the answer was not much: our jobs called the scheduler through one internal wrapper. So I proposed a three-day test and my manager agreed. I ported our two heaviest pipelines to each candidate, ran them against a week of real data, and compared run time, failure handling and how many lines I had to change. One was clearly simpler to operate. I wrote a half-page note with the result and the risk I had not tested, which was backfills, and we started migrating the following Monday. All 140 jobs were moved in three weeks, inside the deadline. The untested risk did bite: one backfill job had to be rewritten a month later, which cost me two days."
Follow-ups to expect. How did you know the wrapper made it reversible? What would have made you ask for an extension instead? Who did you tell about the backfill risk, and when?
Notice what the six have in common. The situation takes two or three sentences. The longest part is what "I" did, in order. Every result has a number, and four of the six include something that went wrong or a prediction that missed. That last part is not decoration: an answer with no cost in it gets a follow-up asking for one.
Phone screen questions versus loop questions
Amazon does not publish which questions are asked at which stage, and interviewers choose their own. What can be said is about proportion and depth, and it comes from Amazon's candidate pages and from what candidates report.
On the phone screen. As of November 2026, Amazon's prep page for senior software engineers describes a 60-minute screen in which half the time goes to the leadership principles and the other half to coding and system design. Candidates at other levels report less, sometimes two questions at the end. With that little time, the questions candidates report are the broad ones that can be graded from a single story:
- a result you delivered under pressure or a deadline you missed (Deliver Results, Ownership);
- a problem where the first explanation was wrong (Dive Deep);
- a decision you changed because of a customer (Customer Obsession);
- a disagreement with a manager or a peer (Earn Trust, Have Backbone);
- something you had to learn quickly (Learn and Be Curious).
Three stories that between them cover a delivery, a disagreement and a failure are enough for this stage. The phone screen guide covers the format by role and a one-week plan.
In the loop. The same page describes five interviews of 55 minutes. Candidates and former interviewers consistently say each interviewer is assigned two or three principles, so across a loop most of the sixteen get asked about at least once, and the same principle can come up twice from different angles. Three things change compared with the screen:
- More principles get a turn. With five interviewers covering two or three each, there is room for the ones a half-hour slot has no time for, such as Frugality, Invent and Simplify or Think Big.
- Scope questions come in for senior and manager candidates: Hire and Develop the Best, Strive to be Earth's Best Employer, and versions of the ordinary questions that ask about a team or an organization instead of a task.
- Follow-ups go further, most of all from the Bar Raiser. Expect a longer chain on one story, and be ready with a second example when the first one was thin.
The practical difference is the size of the set. Three stories carry a screen. A loop needs the eight to ten described below, because interviewers compare notes afterwards and four people hearing the same project is noticed.
The questions that come up most often
Across roles, a few questions appear in almost every loop because they cover several principles at once. Prepare these first:
- Tell me about a time you failed. (Ownership, Earn Trust, Learn and Be Curious)
- Tell me about a conflict with a colleague or manager and how you resolved it. (Have Backbone, Earn Trust)
- Tell me about a time you missed a deadline. (Deliver Results, Ownership)
- Tell me about a decision you made with incomplete information. (Bias for Action, Are Right, A Lot)
- Tell me about a time you had to choose between what the customer wanted and what the business wanted. (Customer Obsession)
- Tell me about something you simplified. (Invent and Simplify, Frugality)
- Tell me about a time you had to learn something quickly. (Learn and Be Curious)
- Tell me about a time you disagreed with a decision and had to carry it out anyway. (Have Backbone; Disagree and Commit)
- Tell me about the hardest feedback you have given or received. (Earn Trust, Hire and Develop the Best)
- Tell me about your most significant professional achievement. (Deliver Results, Think Big)
Each maps to more than one principle. That is why a small set of well-prepared stories goes further than a long list of thin ones.
How to answer in the STAR shape Amazon expects
Amazon interviewers are trained to collect specifics, and the STAR method is the shortest route to giving them. The shape that works in an Amazon room:
- Situation in two sentences. Enough context to understand the stakes, no more.
- Task as your task, not the team's.
- Action as the longest part, in the first person. "I decided", "I wrote", "I called". The word "we" in the action section is a common reason a story fails at the debrief, because the interviewer cannot tell what you did.
- Result with a number where one exists: latency, revenue, time saved, defect rate, days. If the result was bad, say so and go straight to what you changed.
Then stop. Leave room for the follow-up, because that is where the interviewer forms their verdict. Ninety seconds plus three strong follow-ups beats a four-minute monologue. Follow-ups are also the one part you cannot rehearse alone, which is the reason to run a few stories past a colleague or one of the coaches with Amazon on their history before the real thing.
How many stories you need
Sixty questions does not mean sixty stories. Candidates who do well on an Amazon loop usually walk in with eight to ten stories, each mapped to two or three principles, and a note on which follow-ups each story can survive.
Build the set this way:
- Pick the moments from your last five years with a clear beginning, a decision you owned, and an outcome you can describe honestly.
- Mark every principle each story can serve. Pulling a release when the error rate rose can serve Insist on the Highest Standards, Customer Obsession, and Have Backbone.
- Make at least two of them failures. Interviewers ask for one directly, and a candidate who cannot produce one looks either unreflective or inexperienced.
- Write each story in STAR form, then cut the Situation by half.
The story bank guide walks through the method in detail.
By role
The principles are the same for every role, but the weight shifts. These are patterns reported by candidates, not rules; any principle can appear in any round.
- Software engineers: Dive Deep, Insist on the Highest Standards, Deliver Results, Ownership and Bias for Action, with technical specifics inside the behavioral answer. See the Amazon software engineer interview guide.
- Engineering managers: Hire and Develop the Best, Earn Trust, Have Backbone, Deliver Results, and Think Big. People decisions, not just projects. The engineering manager interview guide covers the wider loop.
- Product and program managers: Customer Obsession, Invent and Simplify, Are Right, A Lot, Dive Deep, and Deliver Results. Data and trade-offs carry the answer.
- Operations and non-technical roles: Customer Obsession, Deliver Results, Ownership, Bias for Action, and Frugality, often through questions about process and people under pressure.
Common mistakes
- Answering with a team story. If the interviewer cannot tell what you did, the answer scores as if you did nothing.
- No numbers in the result. "It went well" is not a result at Amazon.
- A story that stops at the result. Interviewers want what you learned and what you would do differently.
- Reusing one story for everything. The debrief will notice that four interviewers heard the same project.
- Naming the principle instead of showing it. The specifics do the work.
- Having no failure story ready, or offering a fake one ("I work too hard").
Where coaching fits
A list of questions and six sample answers will get your stories written. Whether they hold up is decided by follow-ups, and that takes a second person. Nine of the 40 coaches on techinterview.coach have Amazon on their history as of November 2026, and a mock round with one of them can cover a screen or a loop round in the real format, with written feedback on what an interviewer could have recorded for each principle; they are listed on the Amazon interview coaches page. To rehearse alone first, the practice interview runs a behavioral round with an AI interviewer and sends a recap. It is free; sign in to start.
Related guides
- Amazon leadership principles: all sixteen in one table, and how each stage of the process grades them.
- Amazon leadership principles examples: sixteen more worked STAR stories, one per principle.
- Amazon phone screen interview: the first live stage and the three stories to bring to it.
- The Amazon Bar Raiser interview: the interviewer whose follow-ups go furthest.
- Amazon interview debrief: what happens to your answers after the loop.
- Building your story bank: the method for getting from sixty questions to ten stories.