general

Amazon Customer Obsession: Meaning, Examples and Questions

Customer Obsession is the first of Amazon's sixteen leadership principles, and the one the company talks about most. It is also the one candidates most often answer badly, because it sounds like something everybody already does. Every company says it cares about customers. Amazon's interviewers are looking for something narrower: evidence that you have changed a decision, a product or your own plan because of what a customer needed, even when that was inconvenient.

Interviewers test it in nearly every loop, for engineers as much as for product or sales roles. An engineer who has never thought about the person on the other end of the API is a risk in a company that starts every planning document with a press release written from the customer's point of view.

This guide covers what the principle actually says, what it looks like in practice, fourteen interview questions, a worked STAR answer, the follow-ups interviewers use, and the mistakes that cost candidates the round.

The official wording from Amazon

Leaders start with the customer and work backwards.

That is the opening line of the principle on the Amazon leadership principles page. The rest says that leaders work hard to earn and keep customer trust, and that while they pay attention to competitors, their obsession is with customers.

Read it as a sequence. First the customer: who they are and what they need. Then backwards: what has to be true for them to get it, and only then what you build. The phrase "work backwards" is also the name of Amazon's internal planning method, where a team writes the press release and FAQ for a product before writing any code. When an interviewer asks about customer obsession, they are asking whether you naturally think in that order.

What Customer Obsession means in practice

The principle is easier to recognize as behavior than as a definition. People who have it tend to do these things:

They can name their customer. Not "users" in general, but the specific person or team whose day changes when the work ships. For an internal platform engineer that might be the thirty developers who deploy through the pipeline.
They know the difference between what customers ask for and what they need. The request is a faster horse; the need is to get somewhere sooner. They dig for the need before committing to the request.
They say no on the customer's behalf. A feature that one loud account wants but that would make the product worse for everyone else gets turned down, with the reasoning written up.
They go and look. They read support tickets, sit in on sales calls, watch session recordings, or walk the warehouse floor rather than relying on a summary.
They measure trust, not just usage. Retention, repeat purchase, complaint rate and time to resolution matter more to them than sign-ups.

Customer Obsession examples at work

Interviewers hear a lot of stories about "always putting the customer first". The ones that score are specific. Here is the shape of four that work.

A backend engineer deprecating an API. The plan was a 30-day notice and a hard cutoff. She pulled usage logs, found that three of the largest customers still called the old endpoint from batch jobs that ran monthly, and that a 30-day window meant some of them would see one failure before they ever saw the notice. She extended the window to 90 days, wrote a migration guide with their exact call patterns, and emailed the three accounts directly. Nobody broke.

A product manager killing a popular request. The sales team wanted a bulk export feature that two big prospects had asked for. He looked at the onboarding funnel and found that 40 percent of new accounts never completed setup. He argued, with the numbers, that fixing setup would retain more revenue than the export would win, and got the quarter re-planned. The export shipped two quarters later, smaller.

A data analyst who distrusted a dashboard. The NPS score was up, but the free-text comments were getting angrier. She read 300 comments by hand, found that the survey was now only reaching customers who had completed a purchase, and that the people abandoning checkout were invisible. The survey was fixed and the real score was eight points lower.

A support lead changing a refund flow. Refund requests took four days because each one needed a manager's approval. He found that 92 percent were approved anyway, got approval for an automatic refund under a threshold, and cut the median time to under an hour. Complaints about refunds dropped by more than half.

Note what the stories have in common: a specific customer, a specific thing the person found out, a decision that cost something, and a measured result.

Customer Obsession interview questions

The questions below are drawn from real Amazon loops. The first group asks about the principle directly; the second group tests it without naming it.

Direct questions

Tell me about a time you went above and beyond for a customer. Why did you do it?
Give me an example of a change you made in your team or organization to meet a customer's needs. What was the result?
Tell me about a time you had to decide between what a customer asked for and what they actually needed.
When is it right to push back on or say no to a customer request? Give me an example.
Tell me about a time you used customer data or feedback to change a product or a technical decision.
Describe a time you advocated for the customer against an internal stakeholder. Who won?
Tell me about a time you received customer feedback that was hard to hear. What did you do with it?
Do you have an example of improving a customer experience that nobody had complained about?

Questions that test it indirectly

Who is your customer in your current role, and how do you know what they want?
Most of us have felt frustrated or impatient with a customer at some point. Tell me about a time you felt that way and how you handled it.
When you have many customers with competing needs, how do you prioritize? Walk me through a real case.
How would you define good customer service? Give me an example of delivering it.
Tell me about a time you anticipated a customer need before the customer raised it.
How would you rate yourself out of ten on customer obsession, and what is the evidence for that number?

A sample STAR answer

Question: give me an example of a change you made to meet a customer's needs.

Situation. I was a backend engineer on the reporting service of a B2B analytics product. Our largest accounts, the ones with over a million events a day, kept hitting a timeout when they exported a full month of data. Support had a workaround: ask the customer to export one week at a time. Nobody had treated it as a bug because the small accounts, which were most of our users, never saw it.

Task. I owned the export endpoint, and I decided this was mine to fix even though it was not on the roadmap. The customers hitting it were the ones paying the most, and the workaround made us look like we did not understand their scale.

Action. I pulled the logs for every export over the previous quarter and found that 6 percent of all exports failed, but 80 percent of exports from the top twenty accounts did. I wrote a one-page proposal for asynchronous exports: the request returns immediately, the file is built in a background job, and the customer gets an email with a link. I got it prioritized by showing the numbers to my manager and to the account team, built it in three weeks, and migrated the top accounts first so they would see the fix before anyone else.

Result. Export failures for large accounts went from 80 percent to under 1 percent. Support tickets mentioning exports dropped by about 70 percent the following quarter, and two of the top accounts mentioned the change by name in their renewal calls.

Why this answer works

It names a specific customer group and the specific thing that was hurting them, and it shows the candidate going to the logs rather than accepting the workaround.
The actions are personal. "I pulled the logs", "I wrote the proposal", "I migrated the top accounts first". The interviewer can tell what this person did.
The result is measured twice: the failure rate and the support volume, with a third signal from the renewals that ties it back to customer trust.

What the interviewer probes next

A good first answer earns harder follow-ups. Expect some of these and have the answers ready.

How did you know the customers cared? Interviewers want evidence you spoke to or observed customers, not that you assumed. Have the ticket, the call or the data point ready.
What did it cost? Customer obsession that cost nothing is not a decision. Be ready to say what was delayed, what was dropped, or who was unhappy internally.
What was the number before and after? If you do not have one, say what you would measure and why you did not have it at the time.
What would you do differently? A candidate who says "nothing" has not thought about it. Pick one real thing: an earlier conversation, a smaller first version, a metric you should have set up first.
Tell me about a time you got this wrong. Interviewers often ask for the opposite story. Have one where you built what was asked instead of what was needed, and what you learned.

Common mistakes

Treating customer obsession as saying yes. The principle is about serving needs, and sometimes that means refusing a request. A story with no refusal or trade-off in it is weak.
A vague customer. "Users", "the business" or "stakeholders" are not customers. Name the person or the segment.
Describing the company's culture instead of your actions. "At my company we were very customer focused" earns nothing. The interviewer is grading you.
No result. If you cannot say what changed for the customer, the story is incomplete. Even a qualitative result ("the account renewed and expanded") is better than none.
Telling it in "we". Amazon interviewers score what you did. Say "I" wherever it is true.

Related leadership principles

Customer Obsession rarely comes up alone. Three principles share most of its stories:

Earn Trust: the second half of the official wording is about earning and keeping customer trust, and interviewers often grade the two together.
Dive Deep: finding out what the customer actually needs usually means going to the raw data or the raw conversation.
Ownership: the export story above is also an ownership story, because nobody assigned it.

The full list with questions for every principle is in the Amazon leadership principles guide.

Get matched with a coach who has run Amazon loops, or prepare with the STAR method guide, the story bank guide and the Amazon leadership principles guide.