Soft skills in interviews

What the interviewer is actually listening for

Behind every behavioural question sits one assumption: the best available prediction of what you will do is what you have already done. Once you see that, the format stops being a performance and becomes an evidence hearing. The assessment is not about how well you speak. It is about whether anything you say can be checked.

Whatever the wording, the interviewer is listening for four things:

• A specific incident, not a policy. "I usually try to hear everyone out first" describes an intention. It also says you could not retrieve a real moment, which answers a question the interviewer did not have to ask.
• "I" kept separate from "we". If the whole story runs in the first person plural, your contribution is unmeasurable. Teams deliver; the interviewer is hiring one person out of that team and has to know which part was yours.
• Difficulty that is visible. If the situation was not hard, no skill was tested. A story where everything went smoothly proves only that nothing was demanded of you.
• A result, and what you took from it. What changed, and what you have done differently since. A story with no ending reads as a story you never thought about again.

Most interviewers are working from a grid while you talk: a column for the skill, a column for the evidence, a rating. That is why the follow-up questions feel oddly literal — "what exactly did you say to her?", "who else was in the room?", "how did you know it had worked?". They are not suspicion. They are somebody trying to fill in a box you left empty.

It also explains why polished answers can score badly. A candidate who talks fluently about valuing teamwork and a candidate who describes forty seconds of what they said to a colleague on a Tuesday afternoon are not competing on eloquence. The second one is handing over evidence; the first is handing over an opinion, and the interviewer already has one of those.

The STAR method, without sounding like a manual

STAR is not a script to recite. It is a way of distributing ninety seconds so that the part carrying the evidence gets the airtime, instead of the part that is easiest to talk about.

1. Situation — one sentence

   Only as much as the other person needs in order to understand why this was hard. Not the history of the organisation, not the reporting lines, not how you came to be in the role.

2. Task — one sentence

   What you had to achieve, and what the constraint was: time, resources, people, or the fact that nobody had authority over anybody.

3. Action — three to five sentences

   70% of your time goes here. Verbs in the first person singular. What you said, to whom, in what order — and what you decided not to do.

4. Result — one to two sentences

   A number if one exists, an observable consequence if it does not. Then one sentence on what you kept from it.

The second thing to get right is what you prepare. Do not prepare fifteen answers; prepare five or six incidents. One difficult afternoon where a deadline moved can serve adaptability, prioritisation, conflict management or communication, depending on which part of it you put in the Action. Preparing incidents means you are never caught without material, and it keeps you from reciting.

Rehearse each one out loud, once, with a timer. Rehearsing in your head hides the problem: the Situation you think takes ten seconds takes ninety when it leaves your mouth. And never announce the framework. "So, situation: we had a client…" tells the interviewer you are running a template, and from that point they listen to the template rather than to you. The structure should be invisible; only the discipline should show.

15 questions you will hear, and what each one tests

The wording changes between organisations. The skill underneath does not. Below are fifteen questions you are likely to meet, and what each one is scoring.

Two things follow from reading the right-hand column rather than the left. First, the list of skills is much shorter than the list of questions, which is why five or six well-chosen incidents cover almost all of it. Map your incidents onto skills, not onto phrasings, and the unfamiliar question stops being a threat: you are being asked for evidence of adaptability, and you have that.

Second, the last row is different in kind. Every other question asks you to prove a skill by describing a situation. The weakness question asks you to prove self-awareness by describing yourself accurately — which is the one moment where an answer that flatters you scores lower than an answer that does not. If you want the vocabulary for naming your own strengths precisely, the full list of soft skills gives each one an observable indicator rather than an adjective. The same incidents, written as outcomes, are also what belongs on your CV — and an interviewer who has read the CV will usually open by asking you to expand one of them.

Four mistakes that cost the offer

1. The story with no difficulty in it. If everything went smoothly, nothing was demonstrated. Choose the incident where you were uncomfortable, not the one with the tidiest ending.
2. The weakness that is a hidden compliment. "I am a perfectionist" or "I care too much" is recorded as a lack of self-awareness — which is exactly what the question is testing. Name a real weakness, the moment it showed, and what you do about it now.
3. Blaming somebody else. Even when it is entirely true, the interviewer is not hearing a fact about your old manager. They are hearing how you will talk about them in your next interview.
4. The result that never lands. Every story needs an ending: what changed, and what you have done differently since. Without it, the interviewer has to score a fragment, and fragments score low.

Underneath all four sits the same failure of preparation. Candidates rehearse the stories they enjoy telling instead of the stories that prove the skills the role is short of. Read the job advert and underline every verb that describes behaviour — coordinates, presents, negotiates, prioritises. Those are the skills you will be asked to evidence, roughly in that order, and each one deserves an incident chosen in advance.

That only works if you know which of your skills are genuinely the strongest, and self-assessment on this is reliably generous in the wrong direction. A diagnostic gives you a score per skill, so you spend your best three stories on the three that will hold up under a follow-up question.