How often should I update my CV?

'Every six months' is a number with no mechanism behind it

Advice on this question almost always names an interval and almost never names a cause. That is the whole problem: an interval claims that time itself changes something about the document. A CV is not perishable. It does not spoil while it sits in a folder.

Test any such rule with one counter-question: what exactly is supposed to have become different in the last six months? If you held the same job at the same Pensum, earned no qualification and received no reference in that time, the honest answer is nothing. Revising without an event is rephrasing, and rephrasing without new content does not make a document better, only different.

There are exactly two reasons that genuinely require a change. First, something happened in your working life that belongs on the record. Second, you are applying to a specific advert, and that advert uses words your document does not. The first is rare. The second is frequent. Neither has anything to do with the calendar.

So this page gives no 'every X months' recommendation. It gives two measurements instead: how often the first reason actually occurs in Switzerland, and how large a difference it leaves behind.

How often the event really happens — the official figure

The Federal Statistical Office measures occupational mobility through the Swiss Labour Force Survey. Its press release of 24 October 2024 is titled 'One person in seven changed job in 2023' and opens: 'In Switzerland 14.7% of people who pursued a professional activity changed their job between 2022 and 2023.'

The second figure matters more for this question, and in the German and French editions it carries its own subheading: job changes happen on average after 5.6 years. Employed people aged 25 to 64 who moved between 2022 and 2023 had spent an average of 5.6 years with the same employer before leaving. Women moved faster than men, 5.2 years against 5.9.

Set the usual advice against that. Five and a half years is eleven half-years. Anyone revising 'every six months' therefore sits down with an unchanged document about eleven times before anything actually changes. That is not diligence, it is an occupation.

The newer figure sits on the FSO's occupational mobility topic page and refers to 2024: 19.5 per cent of people who were in work during 2024 left their post within a year, 12.9 per cent by changing job and 6.6 per cent by losing work or leaving the labour market. The larger number measures more than a move: it counts every departure from a post.

One caution for readers used to comparing countries: the press release's methodological annex distinguishes a gross rotation rate — everyone who left a post within a year, against the total number of employed people at the start of the observation period — from a net rotation rate, which counts only those who changed and were in work both before and after. Since 2010 internal moves within the same company can be measured separately as well. The 14.7 and 5.8 per cent figures on this page are net rates; the 19.5 per cent is gross.

A footnote that tells you something about reading Swiss statistics in English

The English edition of that press release is worth opening for its own sake. It runs to one paragraph. After the headline and the 14.7 per cent lead it says, in full: 'You can find the complete text of the press release in German, French or Italian', followed by three links.

Everything else quoted on this page — the 5.6 years, the age breakdown, the 34.5 per cent who changed occupation, the definitions of the two rotation rates — exists only in the German, French and Italian editions. The Federal Statistical Office publishes in the three official languages; English gets the headline.

That is worth knowing beyond this one question. If you are working in Switzerland in English and forming a picture of the labour market from English-language sources, you are reading the summaries of a body of statistics whose substance is published elsewhere. It is a good reason to open the German or French edition of anything from a federal office, even with a translation tool, rather than to rely on what an English page carries.

The same holds for the application guidance quoted further down: arbeit.swiss does publish its job-search advice in English, and the wording there is complete — but the underlying analyses it draws on generally are not.

The right rhythm depends on age — by more than a factor of four

A single recommendation cannot be right, because the events are not evenly spread. The FSO breaks the net rotation rate down, and the spread is wide.

Nearly one employed person in four aged 15 to 24 — 23.8 per cent — changed job between 2022 and 2023. Among 25 to 39 year olds with no child in the household the rate is 26.1 per cent for women and 23.3 per cent for men. Women in the same age band with at least one child under eight in the household are at 14.2 per cent, men at 15.5 per cent. From there the rate declines steadily to 5.8 per cent among 55 to 64 year olds.

For the question of rhythm this means: someone aged 22 has something new to enter every few years. Someone aged 58 has nothing in four years out of five. Between those two people sits a factor of more than four — and both are handed the same recommendation online.

The practical substitution is simple. Instead of an interval, put a condition in your calendar. Not 'revise in March', but 'revise when one of the following has happened'.

When the event comes, it is not one line

The second half of the official figures is the surprising one. A job change in Switzerland is rarely a step along the same track.

Between 2022 and 2023, 34.5 per cent of those who changed jobs also moved into a different occupation, defined by the sub-major groups of the Swiss occupational classification CH-ISCO. For 15.1 per cent the work-time percentage shifted by at least ten points after the move, and 38.1 per cent of employees earned at least ten per cent more in the new role at the same Pensum.

A detail in the same release quietly disposes of a common assumption: remote work does not appear to be a reason for moving. The share of employed people doing at least occasional telework was 41.4 per cent in the previous post and 41.8 per cent in the new one — effectively unchanged.

For the CV this forces an uncomfortable adjustment. The rare event is also the big one: for roughly a third of people, the next entry is not another item in the same list but the start of a different list. Someone who planes the document down every six months has gained nothing for that case, because the work it demands is a different kind of work.

What a change of occupation actually changes — measured against the adverts

If a third of moves cross into another occupation, the next question is measurable: how much of one field's vocabulary carries over into another? We counted this across the English-language adverts published on SwissJobs.app, comparing only within a single advert language so that language itself cannot produce the gap. Apprenticeship, internship and trainee adverts are excluded by title, because such an advert names the qualification it leads to rather than one you already hold.

A control ran alongside, without which the number would be worthless: before comparing two different fields, we compare two random non-overlapping halves of the same field. Equal numbers of adverts on both sides, forty draws, averaged.

The control is the real result. Two halves of the same field share only 60.2 per cent of their vocabulary on average — 64.1 per cent in IT and engineering, 58.4 per cent in office, sales and customer service, 58 per cent in hospitality and logistics. Roughly forty per cent of the difference arises with nothing having happened at all: same occupation, same week, same language, just different adverts.

Between two different fields the figure averages 40.8 per cent. Office and sales against hospitality and logistics is the closest pair at 53.2 per cent; IT and engineering against either of the other two sits at 34.5 and 34.6 per cent. So a change of occupation does not delete the document, but it removes a substantial share of the match it previously had.

The more important conclusion is about the normal case, not the move. If two samples of the same field already differ by around forty per cent, there is no state in which a CV 'matches the market'. There is only the state in which it matches an advert. A calendar rhythm promises the first and can at best deliver the second, and then only by accident.

The English-language Swiss market is one occupational world, not five

There is a second reason the comparison above ran across three fields rather than five, and it is the single most useful fact on this page for a reader arriving from abroad.

English-language adverts are a large part of the Swiss market — close to thirty per cent of everything published on SwissJobs.app. But they are not spread across the economy. Within that English-language slice, IT and engineering accounts for 23.9 per cent of adverts, office, sales and customer service for 7.9 per cent, hospitality and logistics for 4.7 per cent. Health and care accounts for 0.5 per cent and construction, trades and technical work for 0.7 per cent.

Compare that with the German-language slice, where health and care is 12.7 per cent, office and sales 13.9, construction and trades 9.4, IT and engineering 8.6, hospitality and logistics 8.4 — five fields of roughly comparable size.

The practical consequence has nothing to do with frequency and everything to do with what you maintain. If your CV exists only in English, the part of the Swiss market that can read it is narrow and concentrated. Keeping a German or French version is not a translation exercise you do once and file away; in most occupational fields outside IT it is the version that gets read at all. That is a reason to revise, and it is a better one than the calendar.

It also changes what 'up to date' should mean for you. Two language versions age independently, because they are read by different employers in different fields. Revising the English one on a schedule while the German one still lists a job you left two years ago is the most common way this goes wrong.

The official guidance names no interval — it names a moment

It is worth checking what the Swiss authorities themselves write. The application tips on arbeit.swiss are the State Secretariat for Economic Affairs' official guidance for jobseekers, and they exist in English. They cover the dossier in detail — and give no frequency for the CV. No 'every six months', no 'once a year', nowhere.

The one real timing instruction concerns the attachments: 'Request these from your former employers in good time.' That is not a frequency, it is a moment — and it applies to collecting, not to writing.

Where the same guidance does use the phrase 'up to date', it applies it to you rather than to the paper: continuing education is described as an important way to stay up to date professionally and to keep or improve your chances on the labour market.

So the official Swiss position is close to the one on this page. What is kept current is first the collection of evidence and second yourself. The writing comes afterwards.

The event list, and what to do in between

Seven occasions genuinely call for a change. An Arbeitszeugnis or Zwischenzeugnis arrives — file it the day it arrives. A qualification or certificate is issued, dated by the issuing body rather than by you. Your Pensum changes. Your function changes, even without changing employer. Your permit line changes. You leave a job, so an open end date becomes a closed one. And you start applying — at which point the rhythm is per advert, not per month.

Between two events the average gap runs to years, and the task in that gap is not writing. It is collecting evidence while it is still reachable. After five years the manager who would have signed your interim reference may well have moved on herself — statistically that is likely, since the same rotation rate applies to her as to you. The confirmation of a course you took has lost its file. The exact month a job started is only approximately remembered.

So keep a folder rather than a date. Into it go every reference and interim reference, every diploma and certificate, exact start and end dates with the month, every change of Pensum with its date, and a short note on each larger project in the year it ran. That is the only work that genuinely becomes more expensive by being postponed.

The CV itself can wait. It is not the archive but the extract drawn from it for one particular advert — and an extract takes twenty minutes when the archive is sound, and is impossible when it is not.

One exception is fair. If you have not touched the document for years and no longer know what it says, a single pass is worth doing — not because six months have elapsed, but because you have lost track of the contents. After that, the event list takes over again.

What a check can establish, and what it cannot

A check can tell you whether your document fits a particular advert: whether the language requirement is covered, whether the Pensum lines up, whether the advert's own words appear in the document. That is a statement about a pair of texts, and it is verifiable.

What no check can tell you is whether a CV is 'up to date'. That property does not exist as a state, and the measurement above shows why: two samples of the same occupational field already differ by around forty per cent. A document cannot match a spread, only a point within it.

Anyone selling a frequency recommendation is therefore selling a certainty that does not exist. The honest rule is duller: collect continuously, revise on events, tailor per advert — and judge the result against the one advert in front of you rather than against the calendar.