Amazon STAR method: structure answers that show evidence.

The STAR framework

Before STAR: give the interviewer CIR

CIR means Company, Industry and Role. Give the minimum context needed to understand where you worked, what the business did and what you were responsible for at the time. Candidates often omit this because it feels obvious to them. It is not obvious to an interviewer meeting them for the first time. If CIR is missing, the interviewer may spend the first minutes reconstructing the candidate's job instead of evaluating the example.

CIR should be brief: usually one or two sentences, adjusted to what the interviewer already knows. It is orientation, not a second introduction.

Use 15–15–50–20 as a diagnostic guide to the proportion of time spent in each component, not as a stopwatch. Keep Situation and Task concise enough that the interviewer reaches the evidence quickly. Action deserves roughly half the answer because it shows how you operated. Result needs enough space to prove impact and reflect on what followed.

Situation and Task: set the stakes

Explain where you were, what needed to happen and why it mattered. Include scale only when it helps: customer impact, team size, timeline, budget, volume or risk. Avoid a long company history. State your responsibility clearly.

Action: reveal your judgement

The most common weak Action section is a shopping list: “I analysed the data, spoke to stakeholders, created a plan and implemented it.” It tells the interviewer what happened, but not how the candidate thought. The stronger version reveals the decision chain:

• How did you diagnose the real problem?
• Which constraints or objectives shaped the decision?
• What alternatives did you seriously consider?
• Why did you reject those alternatives?
• Why was the chosen solution the best available option at that moment?
• How did you implement it, respond to resistance and adjust when new evidence appeared?

Do not give a chronological list of meetings. Select actions showing judgement. Use “I” for your decisions and “we” for genuinely collective work.

Result: use evidence and learning

Quantify impact where reliable evidence exists, but do not reduce “quantified” to money or percentages. Results can be evidenced through:

• Revenue, cost, productivity, volume, quality or error-rate changes;
• Time saved, delivery time, speed to decision or time to recovery;
• Customer feedback, satisfaction, complaints or adoption;
• Stakeholder acceptance, behaviour change or repeated use of the solution;
• Risk reduced, incidents avoided, controls improved or reliability increased;
• Team capability, ownership or knowledge that remained after the project.

State the baseline and time period where possible. If the result is qualitative, describe the observable evidence rather than calling it “successful.” A setback can still be strong if you own it and show what changed afterward. Never invent precision.

Prepare for follow-ups

• What was your specific contribution?
• Why did you choose that approach?
• What alternatives did you consider?
• How did you measure success?
• What resistance did you face?
• What would you do differently?

Common STAR mistakes

• Too much context before the action;
• Obscuring individual ownership;
• Listing activity without judgement;
• Giving a result with no evidence;
• Memorising wording so tightly that the question is not answered.

A practical way to edit a draft answer

First, underline every sentence that describes something you personally decided, changed, analysed or influenced. If little remains, the Action section is probably reporting the project rather than your contribution. Next, circle every number or factual result you could defend in a follow-up. Remove decorative metrics that do not help the interviewer understand the outcome.

Finally, read only the Situation and Task. Could they be shortened without losing the stakes? The time you recover can be used to explain a difficult choice or a meaningful trade-off. This is the kind of editing I use in coaching: not making the story sound grander, but making the candidate's evidence easier to evaluate.

STAR is a preparation tool, not a performance

The interviewer does not need to hear the labels “Situation, Task, Action, Result.” The structure should be felt rather than announced. Natural delivery comes from knowing the experience well, listening to the exact question and being willing to pause when a follow-up changes the direction of the conversation.

Run a Leadership Principles coherence check

A story can answer one Leadership Principle well and still create concern elsewhere. Before using it, ask whether the way you delivered the result conflicts with ownership, trust, customer focus, judgement or another relevant principle. The objective is not to force every principle into every answer; it is to avoid demonstrating one at the expense of another.

For example, an escalation used to deliver a result is not automatically negative. But if the candidate escalated too early, bypassed direct discussion or shifted ownership upward, it may create an Earn Trust or Ownership concern. A credible answer explains what the candidate tried first, why escalation became proportionate, how they preserved the working relationship and what they still owned after escalating.

Frequently asked questions