Why do I keep failing interviews?

Updated: 5 September 2026

In short

Because you rarely sit the same exam twice. 66.1 per cent of open Swiss vacancies carry a job title that appears exactly once in the whole index, and 92.5 per cent of all distinct titles belong to a single employer. On top of that, 72.2 per cent of adverts put no level word in the title — no senior, no junior, no lead. There is no national standard bar to practise against: each employer sets one in the room. A run of rejections is therefore usually not a run of the same mistake, but a run of different yardsticks.

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 second half of the answer is arithmetic nobody enjoys. One advert is normally one seat. Several people are invited and one is hired, so most people who are ever interviewed are rejected — by construction, not by verdict. The Swiss cantonal careers service congratulates you on the invitation itself, because it means "a first hurdle" is already behind you; SECO calls it making the shortlist. Three or four noes in a row is what this process looks like when it is working normally.

It becomes a pattern only when you stop counting rejections and start recording where you stop each time. That is also the official line after a rejection: think about what you can still improve "in your application file and in the job interview" — two different problems, and the round you reach tells you which one is yours. Everything below separates them, and then names the handful of things you carry unchanged into every room. A recurring mistake can only live there.

How often a Swiss job title states a level, by the language the advert is written in

Share of open adverts whose title contains a level word — senior, junior, lead, deputy, specialist or expert. Own count of postings open on 4 September 2026; apprenticeships and internships excluded. Italian-language adverts are a small base.

Two thirds of interviews are one-offs: 66.1 per cent of open vacancies carry a title found nowhere else in the index. In engineering it is 74.4 per cent and in commercial and administrative roles 71.4, but in sales only 48.5 — in sales you really are sitting a recurring exam, in engineering almost never.
The English-language segment is the outlier, and if you arrived through it your expectations are calibrated wrong. 48.6 per cent of English-language adverts state a level in the title, against 18.4 per cent of German-language ones. Switzerland does not usually name levels; the international corner of it does.
That corner also has almost no junior door. Among English-language adverts, 17.2 per cent carry "Senior" in the title and 1.0 per cent carry "Junior" — about seventeen senior-titled vacancies for every junior-titled one.
So "we went with someone more experienced" is rarely a polite fiction. It is usually a literal statement about a level that was never printed, decided against a comparison you never saw.
The questions are the constant, not the bar. SECO names four recurring blocks — your career to date including training, your skills and personal qualities, your motivation for the search and for this specific application, and your suitability and career goals — plus the salary negotiation. A repeated failure almost always sits in one of those five, because only those five occur in every room.
Count exit points, not rejections. Never getting past the first conversation means something different from losing three times in a final round: the first is a presentation problem, the second is usually level, salary or availability.
Fix your salary number before the appointment. SECO advises defining your expectation in advance and looking up the usual pay for the industry and position; improvised in the room, it is the single most repeatable mistake across a series of interviews.
Ask for the reasons. The cantonal careers service explicitly recommends it after a rejection. Nobody owes you an answer, and it is still the only feedback in this whole process that does not come from your own memory.

First the arithmetic: when is it actually a run?

A job advert normally describes one chair. A handful of people are invited, one is hired. That means the large majority of everyone who has ever sat in an interview was rejected afterwards — not because they were bad, but because that is how the procedure is built. Rewriting your career story after the third no corrects a number that is not yet saying anything.

The invitation is itself a selection. The Swiss cantonal careers service congratulates the invited candidate outright: with the invitation, "a first hurdle" has already been cleared. SECO puts it more plainly — anyone invited has made it into the shortlist of possible candidates. Both mean the same thing. If you are getting invitations regularly, your application file works, and your problem is somewhere else.

The count becomes useful only when you separate two things: how many interviews you have actually had, and how far you got in each. Five interviews in a year is a thin sample and completely normal in a narrow field. Five interviews that all ended after the first round is a pattern — and a different one from five interviews that reached a final round three times.

Every interview is a different exam

The reason interviews are so hard to "practise" is visible in the adverts themselves. 66.1 per cent of open vacancies carry a title that occurs exactly once in the whole index, and 92.5 per cent of all distinct titles belong to exactly one employer. The role you interviewed for yesterday mostly does not exist a second time in that shape. There is no canon, no syllabus and no model answer, because there is no second instance of the same job.

How strongly this bites depends on the field, and the spread is wide. In sales only 48.5 per cent of adverts carry a one-off title: the same functions are advertised continuously in the same words, so a run of interviews there really is a run of similar exams. In engineering it is 74.4 per cent, in commercial and administrative roles 71.4, in IT 68.9. There, each conversation is a bespoke assessment, and comparing two rejections tells you correspondingly little.

The practical consequence is uncomfortable but it saves time: you almost never derive a common mistake from the content of two failed interviews. The questions were different, the team was different, the bar was different. The only thing you can compare is what you carried into both rooms unchanged — and that is a short list.

If you arrived through the English-language market, your calibration is off

Most people reading this question in English are either applying from abroad or working in the international part of the Swiss economy, and that part behaves differently from the rest of the country. 48.6 per cent of English-language adverts state a level in the title. Among German-language adverts it is 18.4 per cent, in French 25.7, in Italian 13.0. If your reference point is a market that labels everything Senior, Lead or Principal, a Swiss advert with a bare functional title reads as "open to anyone" — and it is not.

The same segment has almost no visible entry level. 17.2 per cent of English-language adverts carry "Senior" in the title, 1.0 per cent carry "Junior". That ratio has a direct consequence for a run of interviews: if you are early in a career, or restarting one after moving country, the roles you can find in English are systematically titled above where your documented track sits, and every one of those rooms will settle the level question against you at the end.

This is also where the competition changed. Swiss recruiters have been describing redundancies across pharmaceuticals, technology and industry since the spring, and searches that now take months rather than weeks for people with strong files. The person sitting in the same shortlist as you is more likely than two years ago to hold the same qualifications plus local experience, a Swiss reference and a permit that raises no questions. That is not a reason to prepare harder. It is a reason to check whether you are applying half a level above your evidence.

What stays the same in every room — the diagnosis lives there

If the bar varies, the diagnosis has to attach to the constants, and constants do exist. SECO lists the typical employer questions in four blocks: your career to date including education and further training; your skills and personal qualities; your motivation for looking and for this particular application; your suitability for the role and your professional goals. The salary negotiation comes on top, and the same source calls it the demanding part.

Those five blocks appear in every room, however bespoke the role is. That is exactly why a series of failures is nearly always a series inside one of the five. Two are the usual suspects. The first is your reason for leaving: formulated once and then repeated identically in every interview, including the point where it sounds defensive. The second is the salary number, which gets improvised whenever it was not fixed beforehand.

For salary the official advice is concrete and takes half an hour: define your expectation in advance, look up what the industry and the position normally pay — professional associations and the national salary calculator are named as sources — and argue from your performance and qualifications rather than from your needs. A number decided before the appointment is an answer. A number invented in the room is a mistake you then repeat in every later interview.

Where you stop is the diagnosis

A Swiss process has several stations and each tests something different. If you never get past the first conversation, the gap is between your file and you in person: the expectation your CV created is not redeemed in the room, usually because the examples are missing or because answers stay general where they need to be specific.

If you regularly reach the second round and lose there, the problem is another one. The second round is normally run by the manager you would report to, often with people from the team. It no longer tests suitability but collaboration, and at the same time the numbers get pinned down — salary, workload, start date, notice period. A pattern at that station points to level, money or availability, not to your self-introduction.

And if you repeatedly lose in a final round, you probably have no interview problem at all. You have a targeting problem: you keep applying where you are narrowly the second choice. It is the only one of the three diagnoses that better preparation does not fix. It is fixed by changing which adverts you answer — half a level lower, one field across, or where your documented track is the most obvious in the room.

The three things a foreign candidate is quietly re-tested on

The first is availability, and it is three questions in one: what permit you hold or would need, when you could start, and what notice period you are under. In Switzerland these are ordinary interview questions, asked early and answered in one sentence by local candidates. An answer that hedges here is heard as a risk, and the risk is compared against a candidate who has none. Decide the three answers before the appointment and give them without apology.

The second is the language of the room. Your CV states a level; the room tests it, usually in the first five minutes and sometimes in a language different from the one the advert was written in. In German-speaking Switzerland a conversation can open in dialect even when everything written was in standard German. Nobody expects you to speak Swiss German. Being visibly thrown by it in the first minute is what costs, and asking beforehand which language the interview will be held in is a normal, unremarkable question.

The third is documentation. A Swiss file rests on Arbeitszeugnisse, the written employer references issued at the end of each job, and on referees. SECO expects the file to note that references are available on request and suggests handing over a separate sheet of referees at the interview itself. If your previous employers were abroad and issue nothing comparable, say so once, plainly, and bring named referees with their function and phone number instead. Unexplained absence of documents is read as a gap; an explained one is read as a country difference.

The three rejection sentences and what they actually mean

"We decided on someone with more experience." In most cases this is meant literally and refers to the level, not to your answers. Check whether your relevant years can be added up in your file when they are spread over several jobs, and whether you are systematically applying to roles cut one level above your track.

"We liked you a lot, the other candidate simply fitted the team better." This is read as a courtesy and is usually a statement about the comparison rather than about you: the second round tests collaboration, and in a small team the decision goes to whoever fills the concrete gap. It is also the sentence that most often hides an unstated condition — availability, workload, start date.

And the third case: no sentence at all. Nobody owes you a reason. The cantonal careers service nevertheless recommends asking for the reasons after a rejection — briefly, politely, with the explicit intention of doing better next time. You will not always get an answer. When you do, it is the only feedback in this entire process that does not come out of your own memory.

The log: five lines after every interview

Because the yardstick changes every time, a pattern can only come from records, not from recollection. The cantonal careers service names three review questions for immediately after the appointment: what went well, what did not go so well, and what could I improve next time. Write them the same day rather than at the weekend — after twenty-four hours the exact wording of the questions is gone.

Add two lines that carry the actual diagnosis. First: how far did I get — first conversation, technical round, final round? Second: which question did I answer worst, and which of the five blocks does it belong to? Note as well which salary number you gave and whether it was fixed before you walked in. Five lines per appointment is enough.

After five appointments the sheet reads itself. The same question answered badly twice means prepare it. Always finishing after round one means the examples are missing. Always losing in the final round means targeting or level. And if you are registered with an RAV, keep in mind that job-search efforts only count once they are reported to the RAV or recorded in Job-Room — no tool and no private list replaces that step.

Where this answer stops

We see adverts, not rooms. How many people were invited to a particular interview, who sat at the table and what decided it in the end is known only inside the company. Everything counted here comes from the open adverts themselves and describes the conditions interviews happen under — not their outcome in your case.

The job title is also an approximation. Two employers can name the same work differently, and it then counts here as two different roles; conversely the same title can mean two very different jobs. The share of one-off titles therefore measures how inconsistently this market names things, which is exactly the right quantity for the question of whether a standard exam exists — but it says nothing about the number of positions.

And the level count has a known gap: apprenticeships and internships are excluded, and a large part of Swiss entry-level sits precisely there. The 1.0 per cent junior share measures junior-titled regular vacancies, not entry opportunities as a whole. The direction holds: in Switzerland, a level is usually not found in the title but settled in the conversation.

What people report

Swiss recruitment professionals repeatedly describe the decision as two-layered: who gets hired depends on qualifications, but who gets through at all depends on unwritten rules and small signals, with familiarity often outweighing merit. For a run of rejections that means part of what you are measured against is nowhere in writing — and that part never appears in a rejection sentence either.
Aggregated from a collection of more than 400 Swiss LinkedIn posts on applications, selection and rejections, including a February 2025 post by an author working in Switzerland on the unwritten rules of Swiss hiring. Reproduced as a pattern, not as a quotation.
A second observation from the same collection is, for this question, oddly reassuring: practitioners have described redundancy waves across pharmaceuticals, technology and industry since the spring, and job searches now running for months rather than weeks — for people with strong files who would have had several offers a few years ago. Whoever sits in your shortlist today sits next to more locally experienced candidates than two years ago.
Two posts from the same collection, May 2026: a professional working in Switzerland on how long current searches run, and a Swiss recruitment consultant on headcount cuts in pharmaceuticals, technology and industry. Reproduced as a pattern, without names and without reusing third-party text.
A third points the same way the official sources do: many vacancies are filled through contacts and recommendations, and the useful version of network help is not "let me know if I can do anything" but a specific introduction to a specific person. If you keep losing narrowly in interviews, access to the next role often moves your position faster than more preparation for it does.
Post by a professional working in Switzerland, May 2026, from the same collection; consistent with the cantonal careers service recommending that jobseekers tell their personal and professional circle about the search.

The market figures are our own count of adverts open on SwissJobs.app on 4 September 2026; we publish shares, never index sizes. Apprenticeships and internships are excluded — a large part of Swiss entry-level work sits there, so the junior share measures junior-titled regular vacancies rather than entry opportunities overall. A "one-off title" is a job title that, after lower-casing and removing bracketed additions and workload percentages, occurs exactly once in the whole index; different spellings of the same work therefore count as different titles. Level words counted in a title are senior, junior, lead or head, deputy, and specialist or expert designations. Fields are built from title patterns and overlap at the edges; the Italian-language advert base is small and its share is correspondingly less stable. The advert language is the detected language of the advert text. The statements on the shortlist, the typical employer questions, references and salary expectations come from arbeit.swiss (SECO) and from berufsberatung.ch and orientation.ch (SDBB/CSFO, on behalf of the cantons), all read on 5 September 2026.

Prepare the questions from the actual advert

Sources

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%