Salary Transparency Observatory

How many job ads say what they pay

Since 7 June 2026, job ads in Italy must state the starting pay or a pay range. This page measures how many actually do, and how much they offer.

Loading the data…

The data is aggregated and does not show single companies. Combinations with few ads or few companies are hidden, so some filter choices show less detail.

-

of job ads state the pay

Median salary
-
Job ads
-
Companies
-
Vague wording only
-

Who states the pay

Share of job ads that give a figure or a range.

By country

Italy has had the obligation since 7 June 2026. Germany, the Netherlands and France had not yet transposed the EU directive by that date, so they serve as the comparison.

View as table

By sector

The selected sector is highlighted.

View as table

By region

Where the job is located. Ads for remote work are counted apart.

View as table

By contract

Only ads that state the contract type.

View as table

By work arrangement

On site, hybrid or remote, as stated in the ad or implied by the role (stores, plants, clinics).

View as table

Before and after the law

By publication date, across all ads online. "Before the law" covers the 12 months before 7 June 2026. Older ads still online do not represent the market of that time (see the methodology).

View as table

How much they offer

Median of the bottom and of the top of the stated range, in gross euros per year. The four charts share the same scale.

  • Bottom of the range
  • Top of the range

By job function

View as table

By seniority

Only ads that state the level or the years of experience required.

View as table

By region

View as table

By work arrangement

View as table

Where the work happens

Share of job ads by work arrangement, in each sector. "Not stated": the ad does not say and the role does not imply it.

  • On site
  • Hybrid
  • Remote
  • Not stated
  • Hidden (too few ads)

Work arrangement by sector

View as table

Download the data

The same aggregate data as the charts, with counts, shares and medians: one row for each combination of filters (period, sector, job function, work arrangement) and dimension. "all" means no filter.

aggregates.csv

Data license: CC BY 4.0. You can reuse it with attribution.

Methodology and limits

Sources

The ads come from company career sites, read through the public APIs and feeds of their applicant tracking systems (Greenhouse, Lever, Ashby, Workday, Workable, Recruitee, Teamtailor, Oracle Recruiting, SuccessFactors). No logins and no third-party job boards are used. Locations are matched to official lists of municipalities: ISTAT for Italy, Eurostat (LAU 2023 and NUTS 2021) for the other countries.

Collection rules

The collection respects robots.txt and text and data mining opt-outs (art. 70-quater of Italian Law 633/1941). Requests are spaced out and stop when a service asks for a pause. The ad text is not republished, and recruiters' personal data is not stored.

Protecting the companies

Only aggregate data is published. A combination appears if it covers at least ads from at least different companies and no company has more than of the ads. When only one combination in a group is hidden, the smallest of the others is hidden too, so it cannot be worked out by subtraction. For the same reason, a filtered figure is hidden when removing the filter would leave a small group, such as the few ads of one company.

What is measured

An ad "states the pay" if it gives a figure or a range, in the text or in the dedicated field of the applicant tracking system. Monthly figures are converted to gross annual salary over 14 monthly payments in Italy, the Italian norm, and over 12 in Germany, France and the Netherlands (where the 8% holiday allowance, paid separately, is not added); internships always over 12. The main analysis covers ads published from 7 June 2026, the deadline for every EU country to transpose the pay transparency directive (2023/970).

Countries

Italy is the main subject: its obligation applies from 7 June 2026. Germany, the Netherlands and France are observed as a comparison, with the same method: according to the published legislative timetables they had not transposed the directive by that date. Ads are counted in the country where the job is located; the few ads for remote work anywhere in Europe, with no country, are left out.

Work arrangement

Taken from the applicant tracking system when it has a field for it, otherwise from the ad text. When neither says, roles that can only be done in person (store, kitchen, plant, network and clinic staff) are counted as on site.

Limits

  • Only ads online since are observed: older ads still online are not a sample of the market before the law.
  • The data covers of the companies analyzed. Companies whose systems have no public source are left out, so the results describe these companies, not the whole market.
  • Pay, seniority and work arrangement are extracted from the text with automatic rules, which are being checked by hand on samples of ads.
  • The publication date means slightly different things across applicant tracking systems.

Last data update: .