How is an interview answer scored?

What a simulated interview returns

The estimate comes from ONE simulated round and is not a prediction. It says how a panel would have read these answers — not how a particular employer will decide.

The three sub-scores: substance, structure, fit

Substance measures whether your answer contains something checkable: a concrete project, a real number, a situation you actually lived through. Swiss interviewers are unusually evidence-oriented — self-marketing theatre ("I am a huge team player") counts for less here than almost anywhere else, and one verifiable example of it counts for more. An answer without a provable core stays low on this sub-score no matter how fluent it sounds.

Structure measures whether the answer has a recognisable thread: the situation, your task, what you did, the result — for behavioural questions, the STAR logic, without you having to name it. Taking three run-ups, repeating yourself and trailing off without a conclusion loses points here even when the content is right.

Fit measures whether the answer suits the specific position, the company size and Swiss conventions. Part of that is knowing what is normal in a Swiss interview: your salary expectation (the Lohnvorstellung), your notice period and your employer references (Arbeitszeugnisse, the written references every Swiss employer issues) are all ordinary topics. Deflecting the salary question because it would be impolite at home costs fit points, not goodwill.

The judgement comes from the role that asked

You assemble the panel yourself when setting up the session, as free text with up to five roles — say "Head of HR", "Engineering team lead", "Brand designer". Every question belongs to one of those roles, and your answer is explicitly rated the way that panel member would rate it.

That is more than a detail. An HR person probes motivation, team behaviour and salary; the domain expert drills into depth with concrete scenarios taken from the advert; someone from a different occupation checks whether you stay understandable beyond your own field. Answering the expert’s question with a motivation speech costs points — the same speech would have been right for the HR question.

The questions themselves are generated before the session from your CV and the job advert you paste in — or from a typical advert for the occupation if you do not have one yet. So you are not scored against a generic question list, but against the requirements you will actually face.

Company size moves the bar

You pick the company size during setup, and it shifts the yardstick in both directions. At a micro company of a handful of people, breadth beats depth: hands-on mentality, informal first-name culture (in German-speaking Switzerland, the du register), and an honest willingness to do everything. A perfectly structured corporate competency answer reads as oddly distant there — and is judged accordingly.

At a small company of up to about fifty people, ownership and pragmatism rank above process; at a mid-size firm, cross-team collaboration joins them. At a corporation, the formal register applies (Sie in German), and structured competency or STAR answers are not a bonus but the expectation.

If you are new to Switzerland, this is worth knowing about the real market too: the register question is not small talk. Which form of address a company uses is a genuine cultural signal here, and the practice tool deliberately mirrors it instead of flattening every interview into one style.

What comes back with the score

Each score arrives with two to three sentences of reasoning in the language you ran the interview in, plus one or two sentences on what a strong answer would also have included. That second part is the more useful one: it does not just say what was missing, but what belongs in its place — the concrete example, the number, the reference to the advert.

In the report you then see every question with your transcribed answer, the mark shown out of 10, and the reasoning underneath. That makes it possible to re-practise one weak answer deliberately instead of repeating the whole conversation.

If the rating step fails technically, your answer is not lost: the transcript is always saved and simply appears in the report without a mark. A missing number means "not scored", never "failed".

From single scores to the overall verdict

When you finish the session, a second, independent evaluation runs: the whole panel goes back through every question, answer and score — the way a real hiring panel debriefs after the candidate has left the room — and writes up its conclusion.

The result is an honest pass probability from 0 to 100 per cent for exactly this round, a summary of four to six sentences, a list of strengths, a list of concerns — and a short note per panel member on how that role experienced you. The separation is deliberate: per-answer scores tell you which answer to fix; the verdict tells you whether the round as a whole would have been enough.

Together they close the loop that real life leaves open: after an actual interview you almost never learn which answer cost you the job. Here it is written next to each question.

What the score cannot see — and what it costs

What is scored is the transcript, meaning the content of your spoken answer. Tone of voice, pace, pauses and body language do not enter the mark — for those you still need a human or a video recording. And the pass probability is an estimate for the practised round, not a promise: no practice tool knows the actual criteria of a real employer.

Trying it needs no account: the free taste generates two questions for your advert, and one answer of them is scored. With a free account you get five scored answers on the day you sign up and five per month after that; a full practice session runs to eight questions. On the trial or Premium (CHF 15 per month or CHF 95 per year) it is forty scored answers per month, eighty on the yearly plan — live voice minutes are the expensive part of the tool, which is why even the paid tier is capped. All figures as of 6 September 2026.