In this article
Before you start
This page is one step of the behavioral interview guide. It assumes nothing beyond the two ideas below.
Why interviewers ask this
Candidates report this question in behavioral rounds, and the reason is practical. Every manager eventually makes a call that a good employee questions, so the answer shows what you do next: stay silent, argue forever, or speak up and then work well.
Two bad readings are waiting for you. If you say you never disagree, the interviewer may hear a person who hides problems. If you describe a fight, they may hear a person who cannot lose gracefully.
A weak answer blames the manager or ends with "I was proven right." A strong answer shows a specific disagreement, the evidence you brought, a respectful way of raising it, and a sincere commitment to the final decision. Amazon, for one, writes this expectation into its published leadership principles; whether any other employer scores it the same way is not public, so treat that as one example, not a rule. The Amazon Leadership Principles page explains the rest of that list.
Disagreeing with a manager's decision from first principles
- Bounded disagreement
- A bounded disagreement is a disagreement over one decision with real but limited stakes, backed by evidence you can state briefly, so it makes a story that is small enough to tell and serious enough to matter.
In plain words
Picture a copilot and a captain. The copilot sees a risk in the weather and says so clearly. The captain weighs it and chooses the route. The copilot then flies that route with full effort. That is the whole idea.
The analogy breaks in two places. At work your manager may know things you do not, such as budgets or promises to customers, so your disagreement is part objection and part question. And if the risk is a safety or ethics problem, you escalate it instead of flying the route.
In one precise sentence: you voice a reasoned objection to a decision, accept that someone else holds the decision, and then act as if the plan were yours.
Step 1: Pick a real, bounded disagreement
Step one picks a real, bounded disagreement. Choose a decision that affected your work, where you had evidence and the stakes were real but not catastrophic. Good cases are a deadline, a technical approach, a priority call or a customer promise.
Avoid three kinds of case: a personal clash, a decision that was plainly unethical (that is a different conversation), and a disagreement where you simply disliked the plan. The interviewer needs a story with a reason you can state in one sentence. If you feel you have never disagreed with a manager, use a smaller real case, or a disagreement with a senior peer, rather than claiming there was none.
Input: your memory of your recent work. Output: one case with a decision, your view, and your evidence. You now have the raw material; the next step is how you raised it.
Step 2: State your view with evidence and an alternative
Step two states your view privately, early and with evidence, and proposes an alternative. Amazon's text says leaders are "obligated to respectfully challenge decisions when they disagree, even when doing so is uncomfortable or exhausting" (Amazon, accessed 2026-10-26). An Amazon Web Services strategy post on disagreeing upward gives a similar method: respect the decision-maker's concerns and responsibilities, and propose an alternative that meets their underlying need (Amazon Web Services (AWS) Enterprise Strategy Blog, accessed 2026-10-26).
In the story, say what you showed (a sample, a number from your own work, a customer case), what you proposed, and that you asked for the manager's reasons. Input: the decision. Output: your view heard and weighed. Skills in this step overlap with influencing without authority.
Step 3: Commit sincerely and say what you changed
Step three commits fully and names what you changed. Once the decision is made, you work for it. Bezos described the pattern as a candid expression of a view, a chance for the team to weigh it, and "a quick, sincere commitment to go their way" (2016 letter, published April 2017). He also notes the habit runs in both directions: the boss commits too.
Commitment is not silent surrender and not a signal that you were wrong. It means your effort matches the decision. Say what you did to make the chosen plan work, say what happened, and finish with one concrete change in how you work now. That closing change is the rubric line from the opening. You now have the full shape of the answer.
Step-by-step description
- Choose a real, bounded disagreement.
- State your view with evidence and an alternative.
- Commit fully and name what you changed.
- Step one picks a real, bounded disagreement.
- Step two states your view with evidence and an alternative.
- Step three commits fully and names what you changed.
Knowing it is half the job. Saying it is the other half.
Interview Clozo is an AI interview copilot for real interview rounds: it hears the question and puts a strong answer on your screen, and you say it in your own words. See the features or how it works.
Start Free Trial →The Staff-level answer
What was a situation where you didn't agree with your manager's decision?
On a reporting project, my manager decided to ship a new dashboard before we had validated the numbers against finance. I disagreed because a wrong total would cost us trust that is hard to rebuild. I asked for a short slot, showed a few mismatched sample rows, and proposed shipping to one internal team first. She weighed it and kept the original date, since a customer had been promised. I said I still thought it was risky, then committed fully: I added automated checks and a clear notice on the dashboard. A couple of mismatches surfaced after launch and we fixed them quickly. Since then I bring evidence and an alternative with my concern earlier.
This story is an illustrative composite, not a real case. What makes it Staff-level: you were overruled for a good reason, you stayed useful, and the closing sentence shows a change in your own behavior rather than a claim that you were right.
How long should the answer be?
A 60 to 90 second answer holds about 130 to 225 spoken words.1 That assumes an illustrative pace of 130 to 150 words a minute, an estimate and not a measured figure. At 130 words a minute a 60-second clip is 130 words; at 150 words a minute a 90-second clip is 225. The model answer above is about 110 words, roughly 45 to 50 seconds at those paces, so it sits at the short end; add a sentence of context or outcome if the interviewer gives you room.1
Text description of the diagram
- At 130 words a minute a 60 second clip is 130 words.
- At 150 words a minute a 90 second clip is 225 words.
| Part of the answer | Illustrative share | Words at 210 total |
|---|---|---|
| Context and decision | 20% | 42 |
| Your view and evidence | 30% | 63 |
| Decision and commitment | 30% | 63 |
| Outcome and change | 20% | 42 |
The shares are an illustrative budget, not a standard.1 If you run long, cut context first.
The same story, weak version
An illustrative weak answer: "My manager changed the deadline and I disagreed, but I did not want conflict, so I said nothing and did the work. It went fine." It names no evidence, no alternative and no lesson, and it suggests you hide problems. A second weak form is the opposite: a story that ends with "and I was proven right", which turns a question about teamwork into a victory lap.
The rubric
| Dimension | Weak answer | Strong answer | Staff answer |
|---|---|---|---|
| Choice of case | Personal clash or trivial point | Real work decision with stakes | Real decision where the manager had context you lacked |
| Way of disagreeing | Silence, or going around the manager | Raised respectfully with reasons | Evidence, an alternative, and a question about their constraints |
| Commitment | Resentful or unstated | Worked on the chosen plan | Visibly improved the plan you disagreed with |
| Outcome | "I was right" or no result | Honest result | Honest result, including what stayed uncertain |
| Closing change | None | A lesson | A specific change in how you work now |
How interviewers actually score this differs by company and is not public; read the table as advice on what a strong answer shows.
Three variants of this question
Tell me about a time you disagreed with a decision at work.
"Tell me about a time you disagreed with your boss" is the same question with another word for the manager, and "a decision at work" drops the manager entirely. Use the same story: the case, your evidence, the commitment and the change. If the decision-maker was a senior colleague, say so and keep the structure.
What do you do when you disagree and your manager overrules you?
You voice the view once, with evidence, then commit and make the decision work. Bezos calls "you've worn me down" an awful way to decide and prefers quick escalation of real misalignment. My own advice is not to keep reopening a closed decision.
What if the decision turned out to be wrong?
Say what you did when the evidence arrived: report it plainly, help fix it, and do not say "I told you so". If the decision was a two-way door, mention that it could be reversed cheaply, which is why committing was reasonable.
Common mistakes
- "My manager made a bad call and I told everyone." (wrong) It shows poor judgment about where to raise a concern. (right) "I raised it with my manager first, with evidence."
- "I never really disagree with my manager." (wrong) It sounds like you hide problems, and some employers write candor into their values, as in the Amazon text. (right) "I disagree when I have evidence, and I say so early."
- "I did what she said, but I knew it would fail." (wrong) It shows a resentful commitment. (right) "I disagreed, then I worked to make the chosen plan succeed."
- "And in the end, I was proven right." (wrong) It makes the story about winning. (right) "The result was mixed, and here is what I took from it."
- Raising a disagreement over a minor preference. (wrong) A small point shows weak judgment about what matters. (right) Pick a decision with a clear cost if it goes wrong.
Follow-up questions
What would you do differently next time?
Name one concrete change, such as bringing evidence to the first conversation or asking about the manager's constraints before arguing. A vague "communicate better" is weak; practice this kind of specific line with the STAR method for teamwork questions.
What if your manager would not discuss it at all?
Ask for a short, specific slot, frame it as understanding their constraints, and bring a written alternative. If a real misalignment remains, Bezos advises recognizing it early and escalating it, which concerns deep misalignment between teams and is not a license to go around a manager over one decision.
How do you disagree without damaging the relationship?
Raise it privately, lead with the shared goal, show evidence, and propose an alternative that meets the manager's underlying need. Then commit visibly. The interpersonal communication page covers how to describe this skill.
Did the outcome prove you right?
Answer honestly. If you were wrong, say so and what you learned; if you were right, credit the manager's choices that helped. A neighboring question about your limits is covered in greatest weakness.
- Choose a real, bounded work disagreement with evidence and a stated reason.
- Challenge respectfully, early, with an alternative, and ask about the manager's constraints.
- Once decided, commit sincerely and make the plan work.
- Finish with one concrete change in how you work, not a claim to be right.
- Keep the clip to 60 to 90 seconds, about 130 to 225 words; a shorter answer of about 110 words is fine.
Sources
- Amazon Leadership Principles (Have Backbone; Disagree and Commit) — Amazon (undated page, accessed 2026-10-26), 26 Oct 2026
- 2016 Letter to Amazon Shareholders (High-Velocity Decision Making) — Amazon (Jeff Bezos), 17 Apr 2017
- Guts, Part Three: Having Backbone - Disagreeing and Committing — Amazon Web Services Enterprise Strategy Blog (undated page, accessed 2026-10-26), 26 Oct 2026
