Pay gap reporting — a checklist before the June 2026 deadline

Article 9 of the EU Pay Transparency Directive defines nine metrics that companies with 100 or more employees must report to national authorities. For companies between 150 and 499 employees, the first reporting cycle opens in 2026. For companies between 100 and 149, it opens in 2027.

Nine metrics sounds manageable. It is not, for most SMEs, because six of the nine depend on infrastructure that most SMEs have never built — and the three that do not are not the ones regulators weight most heavily.

This article maps the nine metrics, what each requires, and the order in which the underlying work needs to happen. For a full overview of what the directive requires across all obligation areas, the SME guide covers the complete scope. This article focuses specifically on the reporting structure.

What are the nine pay gap metrics?

EU member states may add to this list during national transposition. They cannot remove from it.

01 Gender pay gap in ordinary base pay

The observed difference in average base pay between female and male employees, expressed as a percentage. The directive does not specify mean or median — most national transpositions will. Reporting both is the more defensible approach.

Data requiredbase salary for every employee, by sex.

02 Gender pay gap in complementary or variable components

The observed difference in average bonus, commission, or variable pay between women and men.

Data requiredall variable pay components separated from base. If your HRIS exports a combined total cash figure, this metric cannot be produced without reclassifying the underlying data first. This is the most common data problem in companies between 100 and 300 employees.

03 Median gender pay gap in ordinary base pay

The median of base pay by sex and the observed difference between the two medians. The median is more resistant to distortion by outliers than the mean. It is also, for that reason, harder to explain away.

Data requiredsame as metric 1, sorted by value to find the midpoint of each distribution.

04 Median gender pay gap in complementary or variable components

Data requiredsame as metric 2, sorted.

05 Proportion of female and male employees receiving variable pay

What percentage of women and what percentage of men received any variable pay component in the reporting period. This metric surfaces access differences that salary data alone does not show.

Data requireda flag per employee indicating whether variable pay was received, by sex.

06 Proportion of female and male employees in each pay quartile

The share of women and men in each of four equal salary bands: bottom 25%, second 25%, third 25%, top 25%. This is the metric that most directly shows whether observed differences between women and men are structural or concentrated in specific parts of the distribution.

Data requiredall employees ranked by total pay, divided into four equal groups, with sex breakdown per group.

07 Gender pay gap by category of worker

The observed pay difference broken down by groups of employees doing equal work or work of equal value — not by department, not by title, but by roles that are comparable when evaluated against skill, effort, responsibility, and working conditions.

Data requireda documented, gender-neutral job evaluation methodology that groups roles by comparable weight, applied consistently across all functions. This metric cannot be produced from an org chart. It is also the metric that triggers the 5% threshold under Article 10.

08 Proportion of female and male employees in the company

Simple headcount ratio between women and men. It contextualises every other metric.

Data requiredtotal employee count with sex breakdown.

09 Proportion in each management category

Female and male representation by management level — individual contributor, team lead, senior manager, executive.

Data requireda consistent level taxonomy applied across all employees, with sex breakdown per level. Inconsistently coded management levels in the HRIS are the most common obstacle here.

What are the four directive articles behind the numbers?

Article 4 establishes the principle of equal pay for equal work and work of equal value. It is the legal foundation that makes metric 7 meaningful. Without a documented, gender-neutral job evaluation framework, category-level reporting has no defensible basis.

Article 9 defines the reporting obligation — the nine metrics, the frequency, and the timelines. Reports must be submitted to a national monitoring body and made publicly available. Format requirements depend on each member state's transposition rules.

Article 10 defines the 5% threshold. If the pay gap in any comparable worker category exceeds 5% and cannot be explained by objective, gender-neutral criteria, a joint pay assessment is mandatory. Metric 7 is the direct input. It is the metric regulators examine first.

Article 7 covers employee information rights — the right to request average pay data for comparable roles, broken down by sex. Not part of the annual report, but the data infrastructure for Article 7 responses and Article 9 reporting is the same. Building it once serves both.

Worth noting

"Category of worker" in metric 7 is not your HRIS department structure. It reflects the directive's concept of comparable work — roles equivalent in skill, effort, responsibility, and working conditions. A category that contains only employees of one sex will attract regulatory attention regardless of how it was constructed. The methodology is what auditors examine, not the outcome alone.

What order should the work happen in?

The nine metrics are not independent. Some require infrastructure that others assume. Building in the wrong order wastes time.

Start with the data foundation. Every metric requires complete employee data with sex coded consistently, pay components separated, and management levels assigned uniformly. If your system exports combined total cash, metrics 2 and 4 require manual reclassification before anything else can proceed.

Then build the category framework. Metric 7 — and therefore the 5% threshold — depends entirely on how roles are grouped. This requires a documented, gender-neutral job evaluation methodology. It takes longer than any other single piece of preparation and cannot be done in the final weeks before a deadline.

Then calculate in dependency order. Metrics 1, 3, 5, 8, and 9 require only clean employee data. Metrics 2 and 4 require separated pay components. Metric 6 requires sorted total pay. Metric 7 requires the category framework.

Then document the methodology. The report is not just the numbers. It must include a description of how categories were constructed, how pay components were classified, and how metrics were computed. The documentation is what makes the numbers auditable. Numbers without methodology are not a report — they are data.

Then prepare public disclosure. Article 9 requires that reports be made publicly available, not just submitted to a monitoring body. Format depends on national transposition rules across EU member states.

Where do most SMEs discover the problem?

The most common obstacle is not data volume. It is data structure. A 200-person company can extract base salary by sex in minutes. That same company may have never separated bonus from base in any structured way, assigned management levels consistently across departments, or documented any basis for comparing roles across functions.

Metric 7 is where "we have the data" and "we can report it" stop meaning the same thing.

01 Data exists, structure does not

Salary data is in the HRIS. Pay components are not separated. Management levels are inconsistently coded. No category framework exists. Metrics 1, 3, and 8 can be produced. The others cannot. Worth considering: the missing metrics are the ones regulators weight most heavily. The data structuring work is the critical path, and it takes months.

02 Structure exists, documentation does not

Pay components are separated. Levels are consistent. A category framework exists — informally, in the HR team's collective understanding. But none of it is documented in a way an external party can review. Worth considering: an undocumented methodology is, in regulatory terms, equivalent to an absent one. This is the position where the most progress can be made in the least time.

03 Data, structure, and documentation all exist

All nine metrics can be produced from existing data. The methodology is documented and gender-neutral by design. Categories are defined, applied consistently, and auditable. Worth considering: the remaining work is validation, disclosure format, and submission logistics. This is a manageable position.

The difference between Position 1 and Position 3 is not company size. It is whether pay decisions were ever made against a documented structure — or only against individual negotiation and whatever the market seemed to suggest that quarter.

Where does pay diagnostic work actually start?

The smallest observable step is understanding which position your company occupies — not across all nine metrics at once, but across the five principal obligation axes of the directive.

ReadinessCheck™ takes about 20 minutes and requires no salary data. It produces a position view by axis — including data readiness, category framework, and reporting infrastructure — scored observationally, with the patterns hardest to close identified first.

It is not a legal opinion. It is not a compliance certification. It is a structured observation, useful in deciding where to direct the next month of preparation.