Remote jobs in Switzerland: what the advert actually promises

• Field or customer-site work12
• Foreign geography conflict4
• Explicit fully remote wording2
• Location-specific remote wording unresolved1
• No arrangement statement detected4
• Quantified home-office option1
• Unquantified home-office option24
• Unrelated use of remote or hybrid2

A remote label leaves several different questions unanswered

Can you avoid a daily commute? Work from home two days a week? Live anywhere in Switzerland? Travel to customers instead of an employer’s office? Those are different arrangements, even when a search result calls each one remote.

We reviewed every one of the 50 remote-tagged records in Panta’s 4 November 2026 snapshot of fresh-backed records stored as physical Swiss workplaces. The audit covered 19 recorded employer groups. Twenty-four advertised an unquantified homeworking or hybrid option, one offered a quantified option, and two explicitly used fully-remote wording. Another 12 described field or customer-site duties. The remaining 11 require other explanations, including four with explicit contrary foreign-location evidence.

This is an audit of what a stored label establishes. It is not a finding that only two Swiss jobs permit fully remote work. Some unquantified choices may also allow it. Nor does it measure a national remote-work share, a return-to-office trend, actual schedules or permission to work abroad.

First, most records have no workplace field

The underlying cohort contains 14,536 fresh-backed canonical records across 710 employer groups, classified as physical Swiss workplaces in the frozen index. These are recorded classifications, rather than individually verified geographical conclusions.

Stored workplace field, ignoring capitalization	Records
Missing	13,565
Hybrid	525
Remote	50
Homeoffice possibility	23
Onsite / on-site	372
Jobsharing	1

Thus 93.3% have no stored workplace value. That does not mean those employers refuse homeworking, or even that their adverts omit it. It means this structured field cannot settle the question.

Our broad multilingual text screen retrieved 1,512 records with homeoffice, hybrid, remote or related phrases among the untagged group. These are candidates for inspection. The words can describe technical subjects or explicit refusals, so counting matches as remote jobs would manufacture evidence.

The 467 fresh multi-country records and 34 records classified as remote and Swiss-eligible remain separate geographical populations. We did not add them to the physical-workplace denominator or assume that a Swiss employer permits work from any country.

What the 50 remote-tagged records said

The complete metadata audit reconciles as follows:

Reviewed wording or issue	Records
Unquantified homeworking/hybrid option	24
Quantified homeworking option	1
Explicit fully-remote wording	2
Field or customer-site duties	12
Explicit foreign-location conflict	4
Location-specific remote ambiguity	1
No arrangement statement detected	4
Unrelated use of remote/hybrid	2

The two explicit fully-remote cases were Lightly project-based contractor opportunities. That contract context matters: a task-based arrangement is not automatically a salaried employee role. We did not infer worldwide legal eligibility from work-from-anywhere language.

Other adverts offered meaningful flexibility without a number of homeworking days. Ascento described a remote or Zurich-office choice. Veeva’s company Work Anywhere wording offered home or office choices, sometimes alongside travel obligations. These may permit no routine office attendance; our narrow explicit-wording category does not decide that for them.

One Kanton Luzern leadership advert offered up to three homeoffice days per week. That is a ceiling to discuss, rather than a promise that every week will contain three homeworking days. An 80–100% employment percentage elsewhere in the advert answers a different question.

Home-based field work is a real arrangement, but a different one

A Medtronic sales advert explicitly used the home office as an organizational base while saying most working time would be spent with customers. Another clinical-support role involved operating-room work. An Alcon advert combined a Swiss homeoffice base with substantial travel. Miele installation and service roles described physical customer work.

Those can be attractive arrangements for someone who wants autonomy or no daily employer-office commute. They do not establish that the work can be done from a laptop at home. We kept 12 records with explicit field/customer-site duties separate from the homeworking-option categories.

Two further matches demonstrated why context matters. EPFL’s remote sensing was a research subject. Adaptyv’s hybrid sales and customer-success role combined functions. Neither phrase independently established homeworking permission.

Geography needs its own check

Four remote-tagged records conflicted with their stored Swiss geography: two Alcon territory adverts concerned Korea, Unit8 explicitly required relocation to Abu Dhabi, and Wingtra restricted candidates to Eastern Europe, the Baltics or the Balkans.

We retained all four in the 50-record metadata audit so the index problem remains visible, and excluded them from Swiss arrangement interpretation. Removing them leaves 46 records across 17 employer groups. A separate Lightly advert listed remote options and several countries, but its body did not unambiguously establish a Swiss remote option; that remained unresolved.

These are specific observed discrepancies, not an estimate of all geographical errors. The practical lesson is to read the role’s actual location and residency conditions before treating a remote badge as eligibility.

What we found when the tag was missing

We also reviewed a deterministic sample of 40 untagged phrase-matching records from 40 employers, selecting one per employer to reduce repeated benefit templates. Seven stated a numeric homeworking option and 24 mentioned an unquantified option. Examples included a two-day telework arrangement, an up-to-40% possibility and 40% homeoffice described as guidance. One benefits bullet mentioned two days without specifying the time denominator; we retained that uncertainty.

Three cases described conditional company policies. Volg/LANDI made availability function-dependent. A Compass barista advert’s homeoffice benefit explicitly referred to service-centre employees. Conversely, a hospital cleaning advert listed a general homeoffice benefit without an explicit restriction; we recorded the wording and left actual applicability open, rather than deciding eligibility from the job title.

Two records explicitly said nonremote and four used the terms for unrelated subjects. A separate 20-employer untagged screen-negative sample contained 18 bodies with no detected arrangement statement and two unavailable descriptions. Flexible working arrangements alone were not treated as homeworking.

These samples diagnose how information differs between fields and prose. They do not estimate all employers’ practices or prove that screen-negative adverts exclude homeworking. Coding was AI-assisted, with independent agent challenges and iterative screening corrections, rather than human double coding or a held-out accuracy study.

Employer concentration matters here too

Kanton Luzern contributed 16 of the 50 remote-tagged records (32%): 15 unquantified homeoffice possibilities and the one quantified ceiling. Removing that group leaves 34 records across 18 employers, including nine unquantified options. The downloadable pack shows both distributions.

A concentrated source can shape an apparently precise tag statistic. Removing it exposes that dependence; it does not make the remaining index representative of Switzerland.

How this complements other research

Indeed/OECD research examines advertised remote and hybrid work using more than one billion postings across 20 countries and 55 occupations in 2019–2023. Its historical design answers a different question from this frozen clarity audit. The Indeed Remote Tracker methodology also illustrates that phrase-based remote/hybrid series are broader than verified fully remote opportunities; Switzerland is not in the publicly listed tracker countries.

JobCloud’s 2025 parent survey studies work/family experience among respondents. Its online sample included over 1,000 people, 93% resident in Switzerland and 80% women. It motivates a question about usable flexibility, but cannot be benchmarked against our advert counts. The FSO’s annual teleworking-at-home series through 2025 measures employed people’s reported work practices. It is neither an advert disclosure rate nor a fully remote vacancy count.

For an applicant, the useful follow-up is concrete: how many office or customer-site days does this role require, where must you live, and is homeworking a role-specific agreement or a conditional company benefit? The dated evidence pack and methodology show exactly which wording this report counted.