Behavioural questions and the shape of a good answer

Behavioural interviewing rests on a defensible premise: what someone did before predicts what they will do next better than what they say they would do. Hence "tell me about a time when you had to deal with a difficult colleague" rather than "how would you deal with a difficult colleague".

In a structured process these answers are scored. The interviewer usually has a form with named criteria and a scale, and they are listening for specific evidence against those criteria. Candidates who understand this stop trying to tell a good story and start supplying evidence. Because applications and interviews rely heavily on personal accounts, self-reporting bias is relevant to understanding how reported evidence can be distorted.

What the scorer is looking for

• A real, specific situation — one event, at a stated time, not a general description of how you usually operate.
• Your action, distinguished from the team's. The most common scoring failure is an answer entirely in "we".
• Reasoning. Why you chose that action over the alternative, which is the part that separates judgement from habit.
• An outcome, including a partial or negative one.
• Reflection, where the question asks for it.

STAR, and where it stops helping

Situation, Task, Action, Result is the standard scaffold and it is a good one for structure. Two adjustments make it work far better in practice. SHRM publishes broader guidance on hiring and workplace practice.

First, compress Situation and Task hard. Candidates routinely spend ninety seconds on context and thirty on what they did, which is exactly backwards — the scoring is almost entirely in the Action. Two sentences of setup is usually enough.

Second, add reasoning inside the Action. The bare STAR answer describes what you did; the strong answer describes the choice. "I could have escalated it to the site manager, but that would have taken three days we did not have, so I went to the supplier directly and told the site manager afterwards." That sentence is what a scorer marks highly, and STAR does not prompt for it.

Time allocation

For a two-minute answer, roughly: fifteen seconds situation, ten seconds task, seventy seconds action and reasoning, twenty seconds result, five seconds reflection. Practise with a timer once — most people are shocked at how long their setup runs.

Building a story bank

Do not prepare an answer per question; there are too many questions. Prepare eight to ten situations and map them to themes, because the same event answers several questions when framed differently.

Cover these themes:

• A conflict with a colleague or manager.
• A failure, with real consequences, that was at least partly yours.
• A time you influenced someone with no authority over them.
• A decision made with incomplete information.
• A deadline or resource crisis.
• A time you changed your mind after being challenged.
• A time you improved something nobody asked you to improve.
• A difficult customer, client, patient, or service user.
• A time you had to deliver unwelcome news.
• Your largest or most complex piece of work.

Write each as five or six bullet points, never as a script. Note for each which themes it can serve — a good story usually serves three.

The failure question

"Tell me about a time you failed" is where the most candidates damage themselves, by choosing a disguised strength — the perfectionism answer, the working-too-hard answer, the failure that was really someone else's fault. Interviewers recognise all of these instantly and score them as evasion, which is worse than the failure would have been.

Pick something genuinely yours, with a real consequence, from far enough back that you can discuss it evenly. Structure: what happened, what your part in it was, what the cost was, what you changed, and evidence that the change held. That last element is what converts the answer from confession to competence.

Avoid failures that touch the core requirement of the job. A safety failure in a safety-critical role, or a confidentiality failure in a role built on confidentiality, is a bad choice however well you handled it.

When you do not have an example

It happens — you are asked about something you have not done. Do not invent one; invented examples collapse under follow-up questions, and there will be follow-up questions.

Two workable moves. Offer the nearest genuine analogue and name it as such: "I have not managed a formal grievance, but I have taken a capability process to conclusion, which is the closest thing — can I use that?" Or answer hypothetically while flagging it: "I have not faced that exactly. Here is how I would approach it, based on the closest thing I have done." Both are scored better than a fabrication and much better than a long silence.

Handling the follow-ups

A good interviewer probes: what did the other person say, what would you do differently, what did your manager think, what happened after. This is not scepticism — it is how they distinguish a real memory from a rehearsed narrative. Real memories have texture and can be interrogated in any direction; rehearsed ones cannot.

This is the strongest reason to prepare bullet points rather than scripts. You cannot answer an unexpected follow-up from a memorised paragraph, and the moment where you visibly search for the script is the moment the answer stops being believed.