Free tool · 8 checks · 12 common questions · No AI, no upload · Updated 16 Nov 2026

STAR method answer builder

The STAR method answers “Tell me about a time…” with one real story: Situation, Task, Action, Result. Write yours in four boxes and see, as you type, how long it runs out loud, whether the Action gets most of the time, and whether you say “I”, give a number and say what you learned. Save 3 to 5 stories and see which questions they cover.

TL;DR
  • Keep it to 1 to 2 minutes (about 150 to 300 words at 150 a minute). Amazon says its answers typically last 2 to 3.
  • Spend about 60% on the Action and say “I”, not “we”, for what you did.
  • End with a number and a lesson. Then say it out loud; a story that reads well can still run long when spoken.
The builder · about 10 minutes a story

Write your story, part by part

Pick the question, write each part, and watch the live report. Not sure where to start? Load a sample and edit it.

Amazon principles this question often maps to (our mapping): Hire and Develop the Best and Think Big. Tip: the story is saved under this question unless you tick others below. Model answers for this question

Try a sample:
0 words · about 0:00
0 words · about 0:00
0 words · about 0:00
0 words · about 0:00
Which questions could this story answer?

One good story often answers two or three. Tick them to fill your coverage map.

0:00 · start writingSee report ↓

Your story bank · saved on this device

Save 3 to 5 stories and most questions are covered

No saved stories yet

Write one above (or load a sample) and press “Save to story bank”. Stories stay in this browser only; nothing is uploaded.

Coverage map
  • Leadership: no story yetNo story yet
  • Conflict: no story yetNo story yet
  • Failure: no story yetNo story yet
  • Disagreeing with a manager: no story yetNo story yet
  • Deadline or pressure: no story yetNo story yet
  • Prioritising: no story yetNo story yet
  • Customer or user: no story yetNo story yet
  • Ownership: no story yetNo story yet
  • Decision with data: no story yetNo story yet
  • Ambiguity: no story yetNo story yet
  • Influence: no story yetNo story yet
  • Learning fast: no story yetNo story yet
Answer length · 3 sources opened 16 Nov 2026

Aim for 1 to 2 minutes; Amazon allows 2 to 3

Career services agree on one to two minutes. Amazon, which calls STAR the cornerstone of its behavioural interviews, says its answers typically run 2 to 3.

Recommended length of one STAR answer, in minutes
Every source but Amazon tops out at 2 minutes
  • UMBC Career Center1–2 min
  • University of Reading≤ 1.5–2 min
  • Amazon interview guide2–3 min
Source: UMBC Career Center, STAR method, University of Reading Careers (updated 1 Aug 2023), Amazon, Amazon interview guide (8 Apr 2025) · Nov 2026
View as table
SourceWhat it saysMinutes
UMBC Career Center“Answers can take one to two minutes”1–2
University of Reading Careers“shouldn’t be longer than 1 ½ – 2 minutes”up to 1.5–2
Amazon interview guide“typically lasting 2-3 minutes each”2–3
Words in 1 minute
150
At a calm 150 words a minute (our default)
Words in 2 minutes
300
The top of most guidance

Your own pace matters more than any average. In 2,438 recorded US phone conversations, both speakers together averaged 196 words a minute, pauses included, and single calls ran from 111 to 291. That is casual talk, not an interview target, and it shows how much people differ. So the builder asks you to read your answer aloud and times it: a 250-word answer takes about 1:28 at 170 words a minute and 1:55 at 130.

Worked example · one fictional story, written twice

Same story, 1 of 8 checks vs 8 of 8

A data analyst finds a bug in the refund numbers the night before a board review. Both versions are run through the same code as the builder.

Weak version · 133 words · 0:53 at 150 wpm
Action is 30% of the answer; 1 of 8 checks pass
Read the weak version
Situation
So basically I was working at a startup, it was a food delivery company, quite a big one actually, and we had a lot of dashboards for different teams like finance, ops and marketing, and there was this board review coming up, which was kind of a big deal for everyone because the founders were presenting to investors and the finance team was really stressed about it.
Task
We were supposed to make sure all the numbers were right before the review.
Action
We found that there was an issue with the refund numbers, so we worked on fixing the query and we checked it with the payments data. The team stayed late and we managed to sort it out before the morning.
Result
It went well and the finance head was happy with the numbers.
Source: Auto Interview AI STAR builder, run on a fictional story · Nov 2026
View as table
PartWordsShare
Situation6750%
Task1411%
Action4030%
Result129%
Strong version · 176 words · 1:10 at 150 wpm
Action is 55% of the answer; 8 of 8 checks pass
Read the strong version
Situation
In my first year as a data analyst at a food delivery startup, the finance head needed refund numbers for a board review the next morning.
Task
I owned the refund dashboard, and at 6 pm I found it was double counting partial refunds, so every city looked worse than it was.
Action
I first checked ten orders by hand against the payments export to prove the bug, which took twenty minutes. Then I rewrote the refund query to count each refund once by refund ID, and I compared the new totals with the bank settlement file for the last month. They matched within 0.4%. I told the finance head at 8 pm what was wrong, what I had changed and how I had checked it, so she could decide whether to use the new numbers. I also added a test that fails if a refund ID appears twice.
Result
The board saw refunds of 3.1%, not 4.6%, and kept two cities it had planned to cut. I learned to check every money number against a second source first.
Source: Auto Interview AI STAR builder, run on a fictional story · Nov 2026
View as table
PartWordsShare
Situation2615%
Task2514%
Action9655%
Result2916%
CheckWeakStrong
All four parts are therePass. Situation, Task, Action and Result each have something.Pass. Situation, Task, Action and Result each have something.
Length: 1 to 2 minutesFix. About 0:53 out loud. Add 18 words of what you did and why.Pass. About 1:10 out loud at 150 words a minute.
Action is the biggest part (aim for about 60%)Fix first. Action is 30% of your words (word share stands in for time). MIT CAPD suggests about 60%.Pass. Action is 55% of your words (word share stands in for time). MIT CAPD suggests about 60%.
Short set-up (Situation + Task about 30%)Fix first. Situation and Task are 61% of your words. Long set-ups use up the time interviewers want for your actions.Pass. Situation and Task are 29% of your words. Long set-ups use up the time interviewers want for your actions.
You say what YOU did ("I" at least as often as "we")Fix first. In Action: "I/my/me" 0 times, "we/our/us/the team" 5 times.Pass. In Action: "I/my/me" 7 times, "we/our/us/the team" 0 times.
The result has a numberFix. No number in the Result. Add one you can defend: time saved, %, ₹, users, rank, deadline met.Pass. Your Result includes a number or amount.
You close with what you learnedFix. Add one line on what you learned. Microsoft asks for it (its STAR(R) adds Reflection).Pass. Your Result says what you learned or would do next time.
Few hedge wordsFix. Found: "basically", "kind of", "actually".Pass. No "basically", "kind of", "tried to", "helped with" and the like.
Who asks for STAR · 6 employers and public bodies, quoted

Big employers name STAR in their own interview advice

These are their words, from their careers or guidance pages. Structured interviews (set questions, answers scored against set benchmarks) ranked top for predicting job performance in Sackett and colleagues’ 2022 re-analysis.

WhoWhat they say
Amazon“The cornerstone of these interviews is the STAR method: situation, task, action, and result.”
Microsoft“We recommend you use the STAR(R) model when answering questions.”
Meta“Use the S.T.A.R. (Situation, Task, Actions, Results) method …”
Deloitte (US)“Structure your responses using the STAR method (situation, task, action, result) …”
UK Civil Service“consider using the STAR method which allows you to set the scene, show what and how you did and the overall outcome.”
US Office of Personnel Management (for interviewers)“Behavioral description interview questions ask applicants to describe a past behavior … based on the behavioral consistency principle that past behavior is the best predictor of future behavior.”
How a behavioural round usually goes
  1. 1
    The question

    “Tell me about a time…”, often tied to a value (Amazon has 16 Leadership Principles).

  2. 2
    Your story

    One real example in STAR order, 1 to 2 minutes.

  3. 3
    Follow-ups

    “What would you do differently?” “What was your part?” Have the details ready.

  4. 4
    Scoring

    In structured interviews, answers are scored against set benchmarks (US OPM).

Why employers bother: in Sackett and colleagues' 2022 review of hiring research, “structured interviews emerged as the top-ranked selection procedure” for predicting job performance. Amazon lists 16 Leadership Principles; the builder shows which ones each question tends to test.

FAQ · 9 questions

Quick answers

What is the STAR method?

STAR is a way to answer behavioural questions ("Tell me about a time…") with one real story in four parts: the Situation, your Task, the Actions you took, and the Result. Amazon, Microsoft, Meta, Deloitte, the UK Civil Service and the US Office of Personnel Management all name it in their own interview guidance.

How long should a STAR answer be?

Most career services say one to two minutes (UMBC: "one to two minutes"; University of Reading: no longer than 1½ to 2 minutes). Amazon is the exception: its own guide says STAR responses typically last 2 to 3 minutes. At 150 words a minute, 1 to 2 minutes is about 150 to 300 words.

How much of the answer should be the Action?

MIT’s career office suggests about 60% Action, 20% Situation, 10% Task and 10% Result. The builder shows your split live. If your Situation and Task are more than about a third of the answer, cut the set-up, not the actions.

Should I say "I" or "we" in a STAR answer?

Say "I" for what you did. MIT CAPD says "I" statements show ownership and that "we" makes it hard to tell what you contributed; Amazon’s guide gives the same advice. You can name the team for context, but every action step should be yours.

What if my result has no number?

Use the most concrete outcome you can defend: time saved, a deadline met, an error rate, a grade, a rank, people who used it, or a decision that changed. If there was no measurable result, say what you learned and what you would do next time; Microsoft’s version (STAR(R)) adds that Reflection step on purpose.

How many STAR stories should I prepare?

MIT CAPD suggests 3 to 5. One strong story usually answers two or three questions (a deadline story can also show ownership and a decision with data). Tag your stories in the builder and the coverage map shows which of the 12 common questions still have no story.

What is the difference between STAR, CAR, SOAR and STAR(R)?

They are the same idea with different labels. CAR (Context, Action, Result) folds Situation and Task together. SOAR swaps Task for Objective. STAR(R), which Microsoft recommends, adds a fifth step, Reflection: what you learned. The builder checks for that learning line in your Result.

Can freshers use the STAR method without work experience?

Yes. College projects, internships, hackathons, a club or fest you ran, a part-time job or a family business all count. The fresher sample in the builder uses a final-year project. What matters is that the story is real and the actions are yours.

Is my answer saved or sent anywhere?

Your draft and story bank are saved only in this browser, so they are still there when you come back. Nothing is uploaded and no AI reads your text; every check is a simple rule you can see on this page.

Sources (opened 16 Nov 2026): MIT CAPD, The STAR method for behavioral interviews (updated 30 Aug 2023); UMBC Career Center, STAR method; University of Reading Careers (updated 1 Aug 2023); Amazon, Amazon interview guide (8 Apr 2025); Amazon, Leadership Principles; Microsoft Careers, Interview tips; Meta Careers, Preparing for your software engineering interview (6 Oct 2022); Deloitte US, Behavioral interview tips; UK Civil Service, Success Profiles: Civil Service Behaviours (updated 29 Jan 2025); US OPM, How do I create structured interview questions?; US OPM, Structured Interviews (PDF); Sackett, Zhang, Berry & Lievens (2022), Journal of Applied Psychology 107(11); Yuan, Liberman & Cieri, Interspeech 2006 (speaking rate). Samples are fictional. Not affiliated with any employer named here.

Free to start · runs in your browser · by Tough Tongue AI

A story that reads well can still stumble out loud.

A voice interviewer asks you behavioural questions, follows up on your answers (“what was your part?”), and tells you what was strong and what was thin.

  • Follow-up questions
  • Out loud, like the real room
  • No install needed
Start a free mock interviewOut loud, with a voice interviewer · by Tough Tongue AI