STAR answers that survive the second and third question

How to build a STAR answer with the right amount in each part, and how to handle the probing questions that come after the story is finished.

The short answer: STAR — situation, task, action, result — is a structure for answering a behavioural question, and most people get the proportions wrong. Situation and task together should take about a fifth of the answer; action should take well over half; result closes it. The common failure is spending ninety seconds setting the scene and thirty on what you actually did, which is exactly backwards, because the interviewer is scoring your actions and cannot score context. The second failure is preparing only the story and not the probes: the follow-up questions about what you would change, who disagreed and how you knew are where a rehearsed narrative comes apart and a real memory holds.

Key points

Proportions matter more than the acronym: roughly 20% situation and task, 60% action, 20% result.
Say "I" rather than "we" in the action section — an interviewer scoring you cannot give credit to a team.
Every story needs a result, and a result that is a number, an outcome or a decision rather than "it went well".
Prepare the three standard probes for each story: what you would do differently, who pushed back, and how you knew.

The proportions, which are the whole technique

Everyone knows the acronym. Almost nobody gets the balance right, and the balance is what determines whether an answer can be scored at all.

Where the time in a two-minute answer should go
Part	Share	What it must contain
Situation	~15%	Enough context to make the difficulty legible. No more.
Task	~5%	What you specifically were responsible for.
Action	~60%	What you did, in sequence, in the first person.
Result	~20%	What changed, ideally measurable, plus the reflection.

Roughly 20 seconds of setup, 70 seconds of action, 25 of result.

The reason is mechanical. An interviewer is marking against behaviours — did the candidate take ownership, did they consider alternatives, did they involve the right people — and every one of those is evidenced in the action section. Context that is not needed to make the actions legible earns nothing and consumes the time the actions needed.

Preparing the probes, not just the story

A prepared story and a remembered one sound identical for about ninety seconds. They diverge at the first follow-up, which is why interviewers ask one.

"What would you do differently?" The most common probe and the most commonly fumbled. "Nothing, it went well" reads as either dishonest or incurious. Have a genuine answer: something you would sequence differently, someone you would have involved earlier.
"Who disagreed, and what did they say?" Tests whether the situation was real. Every non-trivial piece of work has someone who saw it differently, and a story with no dissent in it sounds constructed.
"How did you know that was the problem?" Tests reasoning rather than outcome. Be able to say what evidence you had and what you were still uncertain about.

Adding three bullet points per story to your evidence bank covers this permanently. It takes a few minutes per story and it is the difference between an answer that survives scrutiny and one that only survives being unchallenged.

A worked answer, timed

"Tell me about a time you had to influence someone without authority"

Situation (15s): Our month-end pack depended on a file from the operations team that arrived late roughly half the time.
Task (5s): I owned the pack but had no authority over the team producing the input.
Action (70s): I tracked the arrival times for two months so the conversation was about data rather than blame; found the delay was a dependency on a report they also received late; proposed we both take the upstream extract directly; built the intermediate step myself so the change cost their team nothing; ran it in parallel for one cycle before switching.
Result (25s): The input arrived on the first working day every month afterwards. What I would do differently: I spent three weeks assuming it was a priority problem before I measured it, and the measurement took an afternoon.

The action section carries the evidence — measuring before escalating, finding the upstream cause, absorbing the cost of the change, running a parallel period. Those are the behaviours being marked, and every one of them is visible.

Note that the reflection is folded into the result rather than waiting for the probe. Volunteering it costs nothing, sounds like self-awareness rather than defensiveness, and frees the follow-up for something more interesting.

Frequently asked questions

How long should a STAR answer be?

Ninety seconds to two minutes spoken. Under a minute usually means the action section is too thin to score. Over three means the interviewer has stopped tracking the thread and will have to interrupt, which costs you the ending. Time it out loud when you rehearse, because the written version is always shorter than the spoken one.

What if the honest result was that it did not work?

That is a perfectly good answer, and for a failure question it is the only honest one. The result section becomes what happened plus what you did about it plus what changed afterwards. Interviewers are scoring self-awareness on those questions, and a failure story that ends with a genuine lesson scores better than a success story with no reflection in it.

Can I use the same story for more than one question?

Yes, and you should — one substantial project can answer conflict, pressure, influence and analysis depending on which part you foreground. What you cannot do is tell the same story twice in one interview with the same emphasis. Have a second story ready for your strongest theme so you are never repeating yourself in the room.