Gender Composition of Swiss Company Officers
Swiss commercial-register notices identify changes to a company's registered officers and signatories. A notice names the person and their function. It also records signing authority. It does not record gender.
Of the Swiss officer-change records that could be classified from a first name, 27.0% received a female label. The analysis covers 243,261 classified person-company-year records from notices published between 15 July 2025 and 27 July 2026.
Legal form produced the clearest difference. The female-labelled share was 37.9% for sole proprietorships. It was 27.9% for GmbHs and 23.9% for AGs. Canton differences were smaller and harder to interpret because language and publication format affect the measurement.
These figures describe people attached to published officer changes. Serving officers without a published change fall outside the coverage. Gender is inferred from first names. The resulting labels support aggregate estimates. Individual classification falls outside the method's valid use.
Where the differences are largest
By legal form
Legal form separates the groups more clearly than canton. Female-labelled records account for 37.9% of the sole-proprietorship group. For GmbHs, the share is 27.9%, compared with 23.9% for AGs. The confidence intervals for the three estimates do not overlap.
Company structure may help explain the gap. Sole-proprietorship records are closely tied to an individual owner, while AG records include board and management appointments. This dataset shows the difference. Its scope cannot identify which appointment practices produced it.
By type of change
Appointments have a 29.9% female-labelled share, compared with 25.5% for departures. The pattern is consistent with a rising female share among serving officers. This one-window flow cannot establish a change in the standing population.
Role changes have the lowest female-labelled share in the dataset, at 19.1%. These events cover people who remain registered while changing function, often around chair or delegate roles. The data does not show why this group differs from appointments.
By canton
Canton estimates range from about 23% to 32%. Because many confidence intervals overlap, small differences in rank are too uncertain to interpret.
Table view
| Canton | Units | Classified | Female | Share | 95% interval |
|---|---|---|---|---|---|
| UR | 853 | 832 | 264 | 31.7% | 28.7–35.0% |
| TI | 11'809 | 11'412 | 3'594 | 31.5% | 30.6–32.4% |
| BS | 11'186 | 10'580 | 3'237 | 30.6% | 29.7–31.5% |
| GL | 1'016 | 983 | 293 | 29.8% | 27.0–32.7% |
| GR | 5'743 | 5'504 | 1'611 | 29.3% | 28.1–30.5% |
| ZH | 57'535 | 54'481 | 15'823 | 29.0% | 28.7–29.4% |
| AG | 13'072 | 12'513 | 3'618 | 28.9% | 28.1–29.7% |
| BE | 23'947 | 23'032 | 6'556 | 28.5% | 27.9–29.1% |
| AR | 1'557 | 1'487 | 418 | 28.1% | 25.9–30.4% |
| SH | 2'072 | 2'004 | 548 | 27.3% | 25.4–29.3% |
| JU | 1'552 | 1'475 | 402 | 27.3% | 25.0–29.6% |
| SO | 5'306 | 5'106 | 1'374 | 26.9% | 25.7–28.1% |
| LU | 11'183 | 10'773 | 2'854 | 26.5% | 25.7–27.3% |
| SG | 12'881 | 12'417 | 3'270 | 26.3% | 25.6–27.1% |
| TG | 6'021 | 5'822 | 1'531 | 26.3% | 25.2–27.4% |
| BL | 6'294 | 6'028 | 1'549 | 25.7% | 24.6–26.8% |
| OW | 1'272 | 1'239 | 314 | 25.3% | 23.0–27.8% |
| GE | 19'034 | 16'482 | 4'165 | 25.3% | 24.6–25.9% |
| NW | 1'616 | 1'577 | 397 | 25.2% | 23.1–27.4% |
| VS | 9'157 | 8'795 | 2'182 | 24.8% | 23.9–25.7% |
| NE | 3'422 | 3'088 | 741 | 24.0% | 22.5–25.5% |
| VD | 16'481 | 14'739 | 3'488 | 23.7% | 23.0–24.4% |
| AI | 788 | 769 | 180 | 23.4% | 20.6–26.5% |
| FR | 6'126 | 5'529 | 1'295 | 23.4% | 22.3–24.6% |
| ZG | 18'456 | 17'355 | 3'958 | 22.8% | 22.2–23.4% |
| SZ | 6'785 | 6'492 | 1'477 | 22.8% | 21.7–23.8% |
| Legal form | Units | Classified | Female | Share | 95% interval |
|---|---|---|---|---|---|
| Sole proprietorship | 4'308 | 3'872 | 1'466 | 37.9% | 36.3–39.4% |
| GmbH | 59'921 | 54'966 | 15'340 | 27.9% | 27.5–28.3% |
| AG | 137'904 | 131'474 | 31'407 | 23.9% | 23.7–24.1% |
| Change | Units | Classified | Female | Share | 95% interval |
|---|---|---|---|---|---|
| Appointed | 93'519 | 87'399 | 26'116 | 29.9% | 29.6–30.2% |
| Restated | 54'709 | 51'773 | 14'251 | 27.5% | 27.1–27.9% |
| Departed | 91'613 | 86'661 | 22'073 | 25.5% | 25.2–25.8% |
| Role changed | 18'195 | 17'428 | 3'333 | 19.1% | 18.5–19.7% |
Ticino's position is largely a classification artefact. Several common Italian male names receive a female label when language context is ignored. Rerunning the lookup with the dictionary's Italian setting moves Ticino from 31.5% to 24.4%.
The figures for Geneva, Vaud, Neuchâtel, and Fribourg carry a different limitation. Their notice formats require another extraction path, which recovered officer changes from a smaller share of all publications. The raw counts are therefore less comparable with those from cantons that use a labelled officer block.
After allowing for those measurement issues, the canton spread is modest and many estimates overlap. The chart supports broad comparisons only. Rankings of neighbouring cantons remain too uncertain for precise interpretation.
How we calculated it
The source material is 126,409 officer-change notices published in the Swiss Official Gazette of Commerce between 15 July 2025 and 27 July 2026. We parsed the notices into events and matched them to Zefix company records.
Each event records the register change and details of the people involved, including their functions.
We counted one record for each person-company combination in each calendar year. Repeated notices about the same person at the same company within that year still count once. A person registered at four companies contributes four records. After applying that rule, the dataset contains 258,036 person-company-year records.
We assigned a dictionary label from the first given name using the open
gender-guesser library. The parser first separates surnames from
given names, then removes honorifics. The dictionary left 14,775 records
unclassified because the name was missing or ambiguous. That is 5.7% of the
dataset. All shares on this page use the remaining 243,261 classified records,
with a 95% confidence interval.
The dictionary gave 27.0% of classified person-company-year records a female label. The 95% confidence interval is 26.9% to 27.2%. In total, 65,773 records received a female label and 177,488 received a male label. In aggregate, that provides an estimate of the share associated with women among the officer changes covered here.
This window cannot establish a trend. The female-labelled share was 26.8% in the covered part of 2025 and 27.2% in the covered part of 2026. That difference could reflect seasonality or a change in the mix of published events. Several complete years would be needed to assess a direction.
How to read the estimate
This is a flow of published changes
Long-serving officers at companies with no change during the window are absent. A person can also contribute one record for each company they serve, so the result weights people with several mandates more heavily. Cross-company deduplication would risk merging different people who share a name. Any gender difference in the number of mandates held would affect the estimate.
The register captures who can legally bind the company, which overlaps heavily with leadership. A CFO with signature authority appears, while a head of product who lacks it falls outside the dataset. The population is people with registered signing authority attached to a published change. This legal group is narrower than Swiss business leadership as a whole.
First names are an imperfect proxy
The register does not report gender, so this analysis assigns a label from the first given name. That works reliably for many names and poorly for names that are uncommon or used differently across languages. The error rate cannot be measured from this dataset, and the aggregate estimate can shift if errors are concentrated in one direction.
Ticino illustrates the risk. Ignoring language context gives several common Italian male names a female label and shifts the canton's estimate by seven percentage points. The same issue can occur wherever Italian names are common. The available records do not provide enough language context to correct it consistently outside Ticino.
The dictionary could not classify 5.7% of records. Those records are excluded from every share on this page. If their gender composition differs from the classified records, the headline estimate will be biased. This dataset cannot establish the size or direction of that bias.
Canton comparisons need extra care
French-language notices and Fribourg's format require a separate extractor. It produced officer-change records from a smaller share of all notices than the extractor used elsewhere, though canton publishing practices account for some of that difference. Name splitting also left more records unclassified. Their canton shares come from a subset whose representativeness has not been established. Treat the Romandie counts as less complete and small differences between cantons as uncertain.
Using the second token as the given name raises the estimated female-labelled share from 24.3% to 25.9%. This sensitivity test shows how much the parsing choice can move the result.
Use the figures as a description only
The register can show differences by legal form or event type. Causal explanation falls outside its scope. For workforce and management statistics based on reported gender, the Federal Statistical Office publishes the relevant series. This register analysis answers a narrower question about people with registered signing authority who appeared in a published change.
Why this matters for B2B sales
Registered signatories matter when a sale involves ownership or legal approval. They can also play a role in procurement. They are a narrower group than the operational contacts who evaluate many products. For teams measuring coverage across markets, this analysis shows that list composition varies more by legal form than by canton in this sample. First-name inference supports aggregate analysis only. Classifying an individual prospect falls outside its valid use.
How this data appears in Prospex
Prospex company profiles show the officers listed in the commercial register, along with their functions. The profiles also show signing authority. We update those details as new register notices are published.
Prospex shows the register record. Gender remains unlabelled on individual officer profiles.
See an example market, then open a company profile to view its registered officers.