Interviews
STAR Interview Method: Practical Examples
STAR is a structure for answering behavioural questions: Situation, Task, Action, Result. The hard part is not the structure — it is having real stories with defensible results to put into it.
Want this as a repeatable system?
The free Starter Guide turns this into six modules you can apply today.
What STAR really is
STAR is a memory aid, not a script. It gives you four slots to fill so that you do not ramble or skip the part that proves you can do the job.
- Situation — brief context. Two or three sentences maximum, chosen to make the rest of the answer relevant.
- Task — what you specifically were responsible for, including the constraint you were working under.
- Action — what you personally did. This is the longest section and it must be in the first person.
- Result — the outcome, plus what you would do differently.
Three worked examples
Generic
We had an issue with a client who was unhappy about the timeline. I arranged a meeting and we worked through the problems together to find a solution everyone was happy with.
Evidence-based
Situation: a key client refused to sign off a deliverable we had already built, two weeks before launch. Task: I was the project lead and accountable for getting that sign-off without conceding scope. Action: I stopped escalating by email, went to their office, and discovered their objection was about a compliance sign-off nobody had mentioned, not about the work itself. I brought our compliance lead into the next call, restructured the deliverable list around their approval criteria, and gave them a two-item review instead of the full document. Result: signed off nine days later with no scope change. I would now raise approval criteria at kick-off rather than at sign-off.
Generic
Unfortunately the project ran late because of changing requirements from the client and I did my best to manage the situation.
Evidence-based
Situation: I committed to a six-week data migration while also supporting a live release. Task: I owned both, and the release support was not optional. Action: I missed the first weekly checkpoint by three days, told my manager before she asked, and proposed cutting two low-value data fixes to protect the migration date. She agreed, and I sent the revised scope to the client the same day. Result: migration landed on the original date with the reduced scope, and the two cut fixes were picked up in the next quarter. What I learned: a late warning is much cheaper than a late deadline.
Generic
I am a very organised and detail-oriented person who always tries to improve my processes and deliver high quality work to the team.
Evidence-based
Situation: I used to write long handover documents for every project I finished. Task: I was spending most of my last week on each role on documentation nobody read. Action: I asked three colleagues what they actually wanted to know when taking over my work, and replaced the document with a one-page summary plus a short recorded walkthrough. Result: handover time dropped from most of a day to about two hours, and two of the three people said it was the first handover documentation they had used end to end.
Build a story bank before you need one
Interview preparation is much easier when your real experiences are already organised. Do this once and you can answer almost any behavioural question by selecting a story that fits.
1List ten real moments
Things that changed, things you fixed, things you pushed back on, things you built, and one thing you got wrong. Ten is enough for almost any interview.
2Tag each against likely competencies
Typical tags: conflict, leadership, failure, ambiguity, prioritisation, influence without authority, customer focus, learning quickly.
3Write the Result line first
If you cannot state an outcome honestly, the story is too thin for STAR. Either find a better example or answer the question with something smaller and real.
4Rehearse aloud, twice, then stop
Out loud, at a normal speaking pace. Rehearsing in your head is a surprisingly poor substitute.
Using AI to prepare, not to fabricate
AI is good at generating likely questions and at pressuring-testing your answers. It must never be the source of the story itself.
That last prompt is the valuable one. Vague answers — “I led the team”, “it went well”, “I improved things” — are what follow-up questions are designed to expose, and an assistant will find them faster and more bluntly than an interviewer will.