What is a good score in a practice interview?

Updated: 6 September 2026

In short

A rule of thumb for the SwissJobs.app practice interview: a per-answer score of 7 out of 10 or higher marks a solid answer, and the report colours the session's pass probability green from 70 per cent, amber between 45 and 69, red below that. The more meaningful definition of good, though, is beating your own previous attempt at the same setup — because the score measures your answer against a rubric, not your odds against the real, unseen competition for the job.

Practise the interview out loud, and see how you scored

An AI panel asks the questions and you answer by voice, the way the real call goes. Afterwards you get a pass probability and the moments that cost you.

Start a live practice interview

The two numbers are produced separately. Each answer is scored 0 to 10 right after you give it, and when you finish the session, the simulated panel debriefs on everything you said and estimates how likely this specific round would have been to succeed, as a value from 0 to 100 per cent. The colour thresholds at 70 and 45 are the report's own reading aid, there so you can place your result at a glance — not a scientifically derived boundary, and certainly not a hiring promise.

So it pays to define good twice: once in absolute terms, as orientation on the scale, and once relatively, as progress against your last run. The second definition is the honest one, and the tool is built around it — the redo function carries the position, company size and panel unchanged into the next session precisely so that two numbers are comparable at all.

Two scales: each answer gets 0 to 10, and each session ends with a pass probability of 0 to 100 per cent.
The report reads the probability like this: 70 per cent and up is green, 45 to 69 amber, below that red — a reading aid, not a verdict on your real chances.
Answers are measured against a rubric (substance, structure, fit), not against the other candidates — no practice tool can see those.
Only attempts at the same setup are comparable; that is exactly why redo carries position, company size and panel over unchanged.
Good means better than your last attempt. Stable 7s and 8s with the reasoning repeating itself means: move to the next question.
The number does not score tone, pace or body language — a video recording or a human is the better mirror for those.

Two numbers, two yardsticks

After every answer the report shows a score from 0 to 10, built from three sub-scores for substance, structure and fit — how that judgement is produced has its own page. When you close the session, a second, separate verdict follows: the pass probability for exactly this practised round, from 0 to 100 per cent, with strengths, concerns and a short note from each panel member.

Only the second number carries colours in the report: green from 70 per cent, amber between 45 and 69, red below. That is the tool's own framing, and it is deliberately built as a reading aid rather than a forecast. Green means: by this panel's rubric, this round would very likely have been enough. It does not mean you get the job.

For the per-answer ten-point scale the tool deliberately publishes no official threshold. As orientation, derived from the rubric: below 4 the verifiable core is usually missing — nothing concrete, nothing checkable. At 5 to 6 the content is there but the structure or the link to the advert is not. From 7 both hold. If you grew up with percentage grades, resist mapping 7 out of 10 to a mediocre 70 per cent — on this rubric it is a good result.

A rubric, not a ranking

The fundamental difference between a practice score and a real selection process: the practice score is criterion-referenced — your answer is held against a fixed standard. A real hire is a ranking. You are not measured against a rubric but against the other people applying for the same role, and no practice tool can see them: not how many there are, not how strong they are, not whether an internal candidate is already lined up.

That is why a pass probability of 85 per cent is not the same as an 85 per cent chance of the job. The number says: by this panel's rubric, for this position and this company size, your answers in this round would very likely have convinced. About hiring freezes, team chemistry or the candidate who has already done the same job elsewhere, it says nothing — the score does not know the room.

The number is still useful, and precisely because the rubric matches what Swiss recruiters typically apply: sober, evidence-first, little self-marketing. This matters doubly if you are arriving from a market with a different register — a polished, confident elevator pitch that would carry a US interview can score lower on fit here than a plainer answer built around one checkable example. A stable good score on this yardstick means you have mastered the part of a Swiss interview that can be prepared at all.

The most honest benchmark: your last attempt

The redo function carries position, company, company size and panel unchanged into a new session. That is not a convenience detail; it is the point. Only when the setup stays constant are two numbers comparable — and the session list is designed around exactly that comparison. A good score is, first of all, one that is higher than last time.

The reverse also holds: numbers from different setups cannot be read against each other. The same answer is scored differently in a five-person company than in a corporation, and a question from an HR lead demands different things than one from an engineering manager — by design, because what you are rehearsing is the interview actually ahead of you. A 6 in a demanding corporate setup can therefore mean more progress than an 8 in a relaxed one.

And place the first attempt realistically: answering out loud, in real time, without notes is the untrained half of applying for most people. The first round therefore usually lands lower than your knowledge would suggest. That is not a warning sign — it is the baseline everything after it is measured against.

A worked example: from 4 to 8

Say the setup is a mid-size company and the team lead on the panel asks: 'The advert says we are digitising our warehouse processes — where would you start?' The weak answer: 'I am very digitally minded, I learn fast, and I would love to take that on.' It is friendly and fluent — and contains nothing checkable: no lived example, no result, no reference to the advert. On the rubric, an answer like that typically sits around 4.

Under the score come two or three sentences of reasoning — and the more useful part, one sentence on what a strong answer would have included. That sentence is the rebuild plan. Here: a concrete case from your own experience ('At my last employer I moved stocktaking from paper lists to scanners'), a tangible outcome ('month-end closing has run without recounts since'), and a link back to the advert's own wording.

Same question, rebuilt answer: substance rises, fit rises, and structure comes along almost for free, because a lived case naturally has a beginning, an action and an end. An answer like that typically lands at 7 to 8. The point of the example: the improvement did not come from prettier phrasing but from a verifiable core — and that is exactly what the rubric measures.

When a score is good enough

The signal to move on is the plateau: when your scores sit stably at 7 to 8 and the reasoning starts repeating the same praise, another run of the same question buys little. Then the next open question is due — or a different setup: a stricter panel, another company size, the salary question instead of the technical one.

Tens everywhere are the wrong goal. An answer learned word for word can do well on a transcript rubric and still sound recited in a real room — interviewers hear the difference between a scaffold that holds and a script being read out. The purpose of practising is the scaffold: example, result, relevance — freely worded, slightly different every time.

Language is part of the setup too, and for readers applying from abroad it is often the deciding part: a good score in your strongest language says little about an interview that will be held in German or French. The practice interview runs in German, French, Italian or English — the most honest test before the real appointment is one full round in exactly the language the interview will use.

What a good score does not guarantee

What gets scored is the transcript of your answer — what you said. Tone, pace, pauses and body language do not enter the number. If delivery is what you want to work on, you need a different mirror: a video recording, a friend who listens, or a coach for the final polish. A green number replaces none of them.

Add the limit from the ranking chapter: no practice tool knows the real employer's criteria. A high pass probability is a good dress rehearsal — not a pre-decision. If you keep collecting real rejections despite stable good practice scores, the cause is more likely in the dossier, the choice of roles or the market than in another practice round.

Trying it needs no account: the first scored answer works without signing up; after that the free account's monthly allowance applies, and full use is part of the CHF 15 per month subscription — the details are on the page about how answers are scored. All figures as of 6 September 2026.

Scales, the report's colour thresholds (70/45) and the redo behaviour verified directly in the SwissJobs.app practice interview on 06.09.2026.

Practise an interview — live, with an AI voice and scoring

What our job index says about the Swiss market

Computed live from our own index, not quoted from a study. Shares only, as of today.

Language the advert is written in

Deutsch: 60%
English: 23%
Français: 13%
Italiano: 3%

Of adverts that state a language requirement, the share asking for

Deutsch: 70%
English: 43%
Français: 21%
Italiano: 3%

19% posted in the last 7 days · Largest markets: Zürich 18% · Bern 10% · Genève 5% · Basel 5%