The STAR Method: How to Answer Behavioural Interview Questions

Behavioural questions all open the same way — “tell me about a time you…” — and that predictability is a gift. Unlike a curveball technical question, you know the shape in advance. The only people who fumble them are the ones who answer with a vague vibe instead of an actual story. STAR is the frame that turns a vibe into evidence.

The anatomy of STAR

Four parts, in order, with wildly different budgets of your time:

• Situation — set the scene in one sentence. Where, when, what was at stake.
• Task — what you specifically had to do. One sentence.
• Action — the steps you took. This is 60% of the answer and the part most people rush.
• Result — how it turned out, ideally with a number or a clear before/after.

If you remember one thing: Situation and Task are the trailer, Action is the film, Result is the closing card. People invert this constantly — long setup, no substance.

A worked example, annotated

“Tell me about a time you handled a conflict on your team.”

Situation: “Two senior engineers disagreed on whether to rewrite a service mid-quarter.” Task: “As tech lead I had to get us to one decision without losing either of them.” Action: “I had each write a one-page case, ran a 30-minute review focused only on risk, found the rewrite could be phased, and got both to commit to the first phase before we touched code.” Result: “We shipped it over six weeks with zero downtime, and both engineers stayed bought in — one of them later led phase two.”

The Action does the work. “We talked it out and it was fine” is not an answer; it's a summary of an answer you didn't give.

Level changes the scope, not the structure

Weak answer → better answer

Why it works: the better version says “I” and names specific actions, so the interviewer can score you. The weak one hides behind “we” and adjectives — you can't tell what the candidate actually did, which is the only thing being assessed.

Tailoring the Result to your field

• Technical / data roles: reach for a number — latency, error rate, hours saved, uptime.
• Non-technical roles: use a clear before/after when you don't have a metric — “the handover doc that stopped three repeat escalations a month.”
• Career-switchers: pull stories from wherever the skill was really tested — a volunteer role, a side project, your last field. Transferable evidence still counts as evidence.

One story, many questions

You don't need a fresh story per question. A single strong story usually answers several prompts depending on which beat you emphasise — a project rescue is a “deadline” story, a “conflict” story, and a “leadership” story with three different framings. Building that flexible set is the core of preparing for behavioural questions.

Prompt this answers: [conflict / failure / deadline / leadership / ambiguity]
S — [one sentence: setting + stakes]
T — [one sentence: what YOU had to do]
A — [three to four things you personally did]
R — [outcome + number or before/after]
Also usable for: [other prompts this story can flex to]

STAR handles the “time you succeeded” prompts and the harder “time you failed” ones alike. And the reason a story that reads well still collapses in the room is the same reason your opener does: nerves speed you up, so practise these out loud, not just on paper.