STAR and CARL: structuring your story

A good answer to “tell me about a time” is mostly about what you did — so give that beat the most air.

The idea

Behavioural questions ask for a story, and a story needs a shape. STAR gives you four beats — set the Situation, name your Task, walk through the Action you took, and land the Result. The trap is spending forty seconds painting the scene and ten on what you actually did. The interviewer is listening for your judgement, and judgement lives in the Action.

For a story that didn’t go to plan, reach for CARL — Context, Action, Result, Learning. It trades some of the result for the thing a failure story is really for: what you took away and would do differently now.

How it works

1. 1Open, don’t narrate. One or two sentences of Situation — where, when, what was at stake. If it runs past ~15 seconds, you’re scene-setting.
2. 2Name the Task. Your specific responsibility or goal in that moment. Keep it tight — it frames why the Action matters.
3. 3Spend most of your time on Action. First-person, step by step: what you decided and did, and the trade-offs you weighed. This is the point of the story.
4. 4Land the Result. The outcome, ideally with a number or a concrete change. Never let it be a thin “so it went well.”

Budgeting a 90-second answer, a healthy split looks like:

STAR (a story that worked)
  Situation   15s   ██████
  Task        10s   ████
  Action      50s   ████████████████████   <- the bulk
  Result      15s   ██████
  ---------------------------------
  total       90s

CARL (a story that didn't)
  Context     15s   ██████
  Action      45s   ██████████████████
  Result      10s   ████                  <- often partial
  Learning    20s   ████████              <- carries the weight
  ---------------------------------
  total       90s

Notice the trade when you switch frames: the Result shrinks and a Learning beat takes its place. A failure story earns its keep in what you learned, not in a tidy ending.

When to use it

Reach for	When the question is…	The trade-off
STAR	“A time you led / delivered / persuaded / solved” — a story with a clean outcome.	Tempts you to over-set the scene; guard the Action budget.
CARL	“A failure / conflict / mistake / something you’d do differently.”	Needs an honest, specific Learning — not a humble-brag ending.

Both are containers, not scripts. Say the beats in order, but never announce them out loud (“so, my situation was…”). The structure should be felt, not heard.

Watch out for

• The runaway opening. Half your answer spent on background before you’ve done anything. If the interviewer’s eyes drift, you over-set the scene.
• “We” instead of “I”. In the Action beat, the interviewer needs your decisions. A team’s work with no visible you is a “so what was your part?” follow-up waiting to happen.
• A result with no evidence. “It went really well” lands as nothing. Give a number, a before/after, or a credible proxy (“the on-call pages dropped from nightly to about one a week”).
• A failure story with no learning. If your CARL “mistake” is really a success in disguise, it reads as evasive. Name a real cost and a real change.
• Reciting the acronym. STAR is your prep scaffold, not your delivery. Land it as a story.

Worked example

“Tell me about a time you handled a tight deadline.” A tuned STAR answer: Situation (~12s): “Two days before a launch, we found the checkout flow double-charged on retries.” Task (~8s): “I owned payments, so the fix and the call on whether to delay were mine.” Action (~50s): “I reproduced it, traced it to a missing idempotency key, wrote the guard and a backfill for the affected charges, paired with QA on the edge cases, and flagged the two customers already hit so support could refund them proactively…” Result (~15s): “We shipped on time, zero double-charges since, and the idempotency pattern became our default for new endpoints.” The scene is one breath; the Action is the story.

Check yourself

Prep Room · practise the delivery out loud — the structure only helps once it’s second nature.