In this article
Before you start
This page is one part of the behavioral interview guide. It assumes only the two ideas below.
Why interviewers ask this
A decision is the point where a person's judgment becomes visible. The United States Office of Personnel Management describes a structured interview as a method that measures job-related competencies by asking about behavior in past experiences, in hypothetical situations, or both. Decision prompts fit both forms: "tell me about a time you decided" is the past form, and "what would you do if" is the hypothetical one.
A weak answer reports the outcome and says "we decided". A strong answer shows that you owned the call, saw at least two real options, chose by stated criteria and learned something. Companies with written principles publish what they value in decisions. The Amazon Leadership Principles, on Amazon's own page, include Bias for Action, which says speed matters and many decisions are reversible, and Are Right, A Lot, the principle about decision judgment. What a given company scores is its own; treat this page as preparation, not as a scoring sheet.
Decision making questions from first principles
- Decision making question
- A decision making question is an interview prompt that asks a candidate to explain a choice between two or more options, either one made in the past or one posed as a hypothetical, so the interviewer can judge how the candidate gathers evidence, weighs trade-offs and owns the outcome.
- Decision criterion
- A decision criterion is a named factor, such as customer impact, risk or speed, that you use to compare options. Naming your criteria shows that the choice followed a reason and was not a guess, and it gives the interviewer something specific to question.
- Two-way door decision
- A two-way door decision is a choice you can undo cheaply, so it can be made fast with limited evidence. A one-way door decision is hard or impossible to undo and deserves slow, careful work. The terms come from an Amazon.com shareholder letter.
In plain words
Picture choosing a route to work. You know two roads, you care about arriving on time and avoiding stress, and you pick one. A good decision story says which roads, what you cared about, why you chose, and what you would check next time. The analogy breaks in one place. At work other people live with your choice, so the story must also say who you talked to and what you gave up.
In one precise sentence, a good decision answer states the stakes, compares real options against named criteria, commits to one, and reports the result and the lesson.
The structure that carries this is STAR, an interviewing method for behavioral and situational questions. Decision questions put almost all their weight on the action step. In that step, say what you did, why you did it and which alternatives you rejected.
Step 1: Say what was at stake and what you knew
Start with the situation in two sentences: the goal, the deadline, and the one fact that made the choice hard. Then say how reversible the choice was. The Amazon.com shareholder letter separates one-way doors, which must be made carefully and slowly, from two-way doors, which can be made quickly and reopened if they go wrong. Saying which kind you faced explains your pace without a defense.
Input: the messy situation. Output: a stake in one sentence and a reversibility label. For example, "the launch fed a customer renewal, and a feature flag let me undo it in minutes." You now have the frame that justifies everything later, which is why interviewers hear this step as judgment and not background.
Step 2: Name the options and what would change your mind
Say at least two real options, then the criteria you used and your pick. Add one sentence that defines what would have changed the pick. A way to make that sentence exact is a weighted score. A weighted score is a total that multiplies each option's rating on every criterion by that criterion's importance and adds the results. The flip point is the weight at which the winner changes.
Here is an illustrative case: shipping a feature now behind a flag, or delaying two weeks for a full fix. With illustrative weights of 0.4 for customer impact, 0.35 for risk and 0.25 for speed, shipping now scores 3.9 and delaying scores 4.25. The pick reverses if the weight on speed rises above about 0.33, with the other two weights shrinking in proportion.1 "If speed had carried about a third of the whole decision instead of a quarter, I would have shipped" is a sayable flip point.
| Criterion (weight, illustrative) | Ship now behind a flag | Delay for full fix |
|---|---|---|
| Customer impact (0.4) | 4 | 5 |
| Risk (0.35) | 3 | 5 |
| Speed (0.25) | 5 | 2 |
| Weighted score | 3.9 | 4.25 |
You need not recite numbers in the room. The point is that your reasoning could be written as a small table, so say the options, the one or two criteria that decided it, and the threshold.
Text description of the diagram
- Illustrative weighted score for delaying for a full fix is 4.25.
- Illustrative weighted score for shipping now behind a flag is 3.9.
- The pick reverses when speed is weighted at about 0.33.
Step 3: Close with the result and what you changed
State the decision in one sentence, then the result, then the lesson. Wikipedia's summary of the method (revision 1377620587) lists the questions an interviewer wants closed: what the outcome was, whether you met your objective, what you learned and whether you have used it since. If the result was bad, say so plainly and move to what you changed. A decision can be good and the outcome unlucky, and a candidate who separates the two sounds experienced.
Input: the choice you made. Output: a result with a measure if you have one, plus one changed habit. For example, "I now write the alert threshold down before launch." As a rule of thumb, keep the whole story short enough to say in about a minute and a half.
Step-by-step description
- Say what was at stake and how reversible it was.
- Name options, criteria and the flip point.
- Give the result and what you changed.
- The answer starts with the stakes and how reversible the choice was.
- The middle step names the options and what would change your mind.
- The answer ends with the decision, the result and the lesson.
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
The answer below is an illustrative example, not a real person's story. Replace the details with your own and keep the shape.
Tell me about a decision you made with incomplete information.
Our team could ship a reporting feature on schedule with a known slow query, or delay it to fix the query. I owned the call. I compared customer impact, risk and speed. Delaying scored slightly higher at first, but the renewal date made speed about a third of the decision, which tipped it to shipping, and a feature flag kept the choice reversible, so I shipped to a small group. I set a latency alert as the signal and agreed to pause if it fired twice. It never fired, and we fixed the query next sprint. If the renewal had not depended on the date, I would have delayed. Next time I will write the threshold down before launch.
What makes it Staff-level: it names the owner, shows that reversibility set the pace, gives a pre-agreed trigger for changing course, and ends with a changed habit.
The rubric
| Dimension | Weak answer | Strong answer | Staff answer |
|---|---|---|---|
| Ownership | Says "we decided"; the candidate's role is unclear | Says "I decided" and what the candidate personally did | Owns the call and names who was consulted and who disagreed |
| Options | Only one path is described | At least two real options are named | Also says which option was rejected and the cost of rejecting it |
| Criteria | No reason beyond "it felt right" | Names the criteria used | Shows how criteria were weighed and what would flip the pick |
| Handling uncertainty | Waits for full data or guesses | States what was known and unknown | Links pace to reversibility and sets a trigger to change course |
| Result and learning | Outcome only | Outcome and a lesson | Separates decision quality from luck and shows a changed habit |
Three variants of this question
Tell me about a difficult decision you had to make
Say the stakes, what made the choice hard and the trade-off you accepted, then the result. State whether it was a two-way door you could undo cheaply, because that explains why you acted on limited evidence and which safeguard, such as a small rollout, you set.
Tell me about a decision you made that was unpopular
Name who disagreed and why their view was reasonable. Then explain your criteria, how you communicated the choice, and what happened. The influencing without authority page covers the persuasion half.
What is your decision-making process?
A process question is the hypothetical form of the decision prompt. Give the four moves in order: frame the stakes and reversibility, list options, compare on criteria, set a signal to revisit. Then attach a short real example so the process is not abstract.
Common mistakes
- "We decided to go with the second plan." (wrong) The interviewer cannot tell what you did. (right) "I recommended the second plan and made the final call."
- "I had no real alternatives." (wrong) It suggests you did not look. (right) "I considered two other options and rejected them because of cost and risk."
- "I waited until I had all the data." (wrong) It signals slow judgment on a reversible choice. (right) "The choice was reversible, so I acted on what I knew and set a check."
- "It worked out, so it was the right call." (wrong) It confuses luck with judgment. (right) "The result was good, but the reason is that the downside was small and I was watching a signal."
Follow-up questions
What would you do differently?
Name one specific change, such as writing a trigger down earlier or consulting a person sooner. A reflection with no concrete change reads as a polished non-answer.
How do you answer a question about a decision that turned out wrong?
Say the call, what you knew at the time and how you found out. Judge the decision by what was known then, not only by the result, and finish with the one habit you changed.
What if the priorities had been different?
Give the flip point from Step 2: the shift in priorities at which you would have chosen the other option.
How do you answer when you disagreed with your manager?
Say what you recommended and why, how you raised it, and what you did once the call was made. The Google behavioral questions page covers related prompts reported by candidates.
What is the difference between a behavioral and a situational decision question?
A behavioral question asks what you did; a situational question asks what you would do. The Office of Personnel Management describes structured interviews as using either or both. The common interview questions, human resources (HR) interview questions and greatest weakness pages show neighboring prompts.
- A decision answer names the options, the criteria, the pick and the result, and says "I" for your part.
- Reversibility explains your pace: two-way doors can be decided fast, one-way doors slowly.
- Give a flip point: what would have changed your mind.
- Separate decision quality from outcome luck, and end with a changed habit.
Sources
- Structured Interviews — United States Office of Personnel Management, 26 Oct 2026
- Situation, task, action, result (revision oldid 1377620587) — Wikipedia (secondary, used for the definition only), 26 Oct 2026
- Letter to shareholders, Exhibit 99.1 (one-way and two-way doors) — Amazon.com, filed with the Securities and Exchange Commission, 1 Apr 2016
- Amazon Leadership Principles — Amazon.jobs (vendor documentation), 26 Oct 2026
