DIY Guide: Conducting an EU Pay Transparency Pay Gap Audit

SkillsTrust Team

• 4 minute read

If your company has a gender pay gap greater than 5% in any job category and you have 100 or more employees in EU countries, you could face legal risks under the EU’s Pay Transparency Directive, effective June 2026.

This article breaks down the key elements of the 5% pay gap threshold, how to calculate your gender pay gap, and the steps you can take to check how ready you are for compliance. Plus, learn how tools and processes can simplify the journey to compliance.


What Is the 5% Pay Gap Threshold?

The EU’s Pay Transparency Directive requires companies to take action when the gender pay gap exceeds 5% within any job category of workers performing equal or equivalent work.

If this threshold is exceeded, companies must:

  • Remedy the gap within six months, or

  • Conduct a joint pay assessment with employee representatives to address the issue.


How to Calculate Your Gender Pay Gap Under the Directive

To work out whether you have a gender pay gap of 5% or more in any job category, you need to:

  1. Create a Job Framework or Architecture 

Think of job architecture as the foundation of your compliance with the EU Pay Transparency Directive. It’s a data structure for organising jobs into comparable categories based on objective factors including:

  • Role responsibility level - include the nature and scope of both formal and informal duties. Tasks like managing people or information, handling equipment and overseeing financial resources. 

  • Skills needed for the role - this includes knowledge, abilities and attitudes.

  • Effort levels - consider the physical, emotional or mental energy needed to meet the workload and demands of the role.

  • Working conditions - assess the environment and other factors like working hours, physical workspace, processes and tools.

This list covers the elements needed to ensure gender-neutral role assessments. You can add additional elements but you must ensure they are gender neutral. It’s worth noting that the EU Pay Transparency Directive states: “In particular, relevant soft skills shall not be undervalued.” 

Typically, each job category is made up of a job family (grouping roles with similar skills and tasks) and a job level (grouping roles based on the required level of skill, responsibility, and effort, working conditions for the role). You will need to decide on the right number of families and job levels for your organisation. You can find out more in our article on building a compliant job levelling framework.

Once you’ve settled on your approach, the framework must be applied across the entire organisation to level each job. This work creates a matrix of roles, grouped by job family and level (or grade) so you can easily identify which roles do work of equal or equivalent value. 

  1. Collect detailed compensation information

Now it’s time to get all your pay gap dataset together. Article 9 of the EU Pay Transparency Directive details the reporting data inputs. These include: 

  • Base salary for all workers expressed as an FTE.

  • Details of any bonuses and incentives.

  • Overtime pay, housing and food allowances, payments in the case of dismissal, statutory sick pay and statutory required compensation.

  • The cash value of benefits like health insurance, pension contributions, and company car allowances.

  • Additional perks like stock options, profit-sharing plans and other non-monetary benefits.

The good news is, you may already have some or all of this data for any national gender pay gap reporting requirements your government may require. 

  1. Bring it all together

Once your data is collected, match the roles from your job architecture with the corresponding pay data to build a complete pay gap dataset. Use this data to identify whether there are any gaps greater than 5% in any job category. 

You’ll need to calculate and report:

  • The overall mean and median gender pay gap.

  • The mean and median gender pay gap for additional or variable pay components.

  • The proportion and percentage of female and male workers receiving additional or variable pay components.

  • The proportion and  percentage of female and male workers in each quartile pay band.

  • The mean gender pay gap between workers by categories of workers split out  to show basic pay or salary and additional or variable pay components.

To calculate your gender pay gap:

  • Compare the average gross hourly earnings of all male and female employees for base pay and additional or variable pay components.

  • Do this for the overall population and for the people in each level or grade of your job leveling framework (you’ll see this referred to as a ‘worker category’ in the legislation and a lot of reward commentary). 

“The information in this article is based on the broad data gathering and reporting requirements set out in the EU directive. However, Member States may create more detailed rules locally. So think of this information as a minimum requirement and look out for local legal updates to ensure compliance for your business.” Michelle Dervan, Founder, SkillsTrust


Ok, We’ve Got a 5% Pay Gap - Now What?

Should your analysis reveal a pay gap of more than 5% in any job category, you must identify whether you have any objectively justifiable reasons for the difference. 

  1. Identify what’s driving pay differentials of 5%+

The following metrics are very helpful in identifying the underlying cause of the pay gap:

  • Representation by men and women in the job category.

  • The compa-ratio for men and women in the job category.

  • Any individual outliers that are paid outside of the typical pay range and may be skewing the numbers. 

This can include objective factors related to your job leveling framework as well as elements that can impact pay levels. For example:

  • Performance differences (backed up by robust performance management processes and records)

  • Experience levels

  • Geographic location

From this analysis, you may have some people with a justifiable pay gap and others without. 

  1. Remedy unjustified gender pay gaps within six months

You have six months to remedy the difference in the average pay level to below 5% for people with unjustified pay gaps. This typically means adjusting compensation. But you could also review your reward policies to ensure they drive fair pay practices.

  1. Carry out a joint pay assessment if you can’t reduce your unjustified pay gaps

If you cannot reduce the pay gap to less than 5%, you must carry out a joint pay assessment with any workers’ representatives to identify underlying causes. 

Unjustified pay differences must be corrected within a reasonable period of time. Most Member States have yet to publish their legislation which will likely set timescales around this work. 

You must make the joint pay assessment available to workers and workers’ representatives and communicate it to your local monitoring body. 

  1. Establish and maintain ongoing monitoring and reporting processes

This isn’t a one-and-done task as a key part of the directive is to ensure your approach delivers the intended outcomes. You’ll need to establish a schedule and processes to ensure sustained compliance and monitor the effectiveness of corrective actions and policy change. 

Don’t forget to boost employee engagement by keeping management and employee representatives up to date with your progress.


How SkillsTrust Can Help

Whew! There’s a lot to consider there. Especially for hard-pressed HR teams. Which is where software developed specifically for these tasks comes in. Look for platforms that:

  1. Inventory, analyse and update your job documentation.

  2. Evaluate each role using objective, gender-neutral criteria that meets the EU Pay Transparency Directive.

  3. Create a transparency-ready job architecture with job families and levels.

  4. Consolidate the relevant pay data to conduct the exact pay gap analysis needed to maintain compliance plus additional calculations to help you identify pay gap causes. 

  5. Provide reporting which not only meets your pay gap reporting responsibilities but identifies gender data, outliers, compa-ratios by men and women plus any risks that need action.

Wondering where to get hold of software like this? Take a look at SkillsTrust. Designed specifically for SMEs and mid-market organisations, our platform saves you a significant amount of time and delivers a simple solution at a sensible cost. 

Find out more by reading our FAQs. Or book your free pay transparency check-up

Read more articles

Michelle Dervan

2 Nov 2026

All About Averages | Pay Analytics 101

The Importance of the Average When talking about pay data, the term average appears time and time again, but what does it actually mean? Put simply, an average salary gives us one figure to help describe pay across a group of employees. It takes all the salary values from that group and calculates a figure that would represent a ‘typical’ employee in that group. This is useful because it allows us to compare pay between different employee groups without having to look at every individual salary. The average can become an even more powerful metric when we begin to understand and analyse the different ways it’s calculated. It might seem counterintuitive that a figure like the average can be calculated in different ways and possibly provide different answers. After all, how can the same set of salaries produce different averages? The key is that each method looks at the data differently, helping us bring a different part of the picture into focus. There are two types of average that are particularly important when it comes to understanding pay: the mean and the median. So, how do they work, and what can each tell us about pay? Let's introduce them using a simple example company, PeopleTech Ltd. The Mean: What would everyone earn if pay were shared equally? The mean is what many people think of when they hear the word average. It shows us what each employee would earn if the company's total pay were shared equally among everyone. In everyday conversation, the terms mean and average are often used interchangeably. However, this isn't technically accurate, as the mean is just one of the ways to calculate an average, as we'll explore later on. To calculate the mean, we simply add together all the salaries and divide the total by the number of employees. Let's look at the example company, PeopleTech Ltd. Imagine the five employees earn the following annual salaries: €30,000 + €40,000 + €40,000 + €50,000 + €100,000 = €260,000 PeopleTech Ltd. has a total annual salary bill of €260,000 across its five employees. To calculate the mean, we divide this total by five: €260,000 ÷ 5 = €52,000 So, the mean salary at PeopleTech Ltd. is €52,000. In other words, if the company's total salary bill were shared equally between all five employees, each person would receive €52,000. The mean is useful because it gives us an overall picture of pay across the company. However, as we'll see, it doesn't always reflect what most employees actually earn, particularly when there are significant differences between the highest and lowest salaries. The Median: What is the salary in the middle? But what happens when one or two employees earn significantly more than everyone else? This is where the median can give us a different perspective. While the mean looks at the total of all salaries, the median looks directly at the middle of the data. Imagine arranging all our employees in a straight line, from the lowest-paid to the highest-paid. The salary of the person standing in the middle would represent our median salary. Let's look at PeopleTech Ltd. again. Using the same salaries as before, we arrange them from lowest to highest: €30,000 | €40,000 | €40,000 | €50,000 | €100,000 The middle salary is €40,000, making this the median salary for PeopleTech Ltd. What makes the median particularly useful is that it is not affected by unusually high or low salaries, often referred to as outliers. In our example, let's assume the €100,000 salary belongs to the CEO, which is double the next highest salary. This unusually high salary pulls the mean upwards, making the average salary appear higher than what most employees actually earn. The median, however, is not affected by this unusually high salary. At €40,000, it gives us a different perspective on what a typical employee earns in PeopleTech Ltd. This is particularly useful for HR professionals because it helps us understand what employees typically earn, without the results being skewed by a small number of particularly high or low salaries. Neither figure is wrong. They simply tell us different things about how pay is distributed across the company. Looking at both together gives us a more complete picture than relying on either one alone. One thing to note: when there is an even number of employees, there won't be one person sitting exactly in the middle. In this case, we take the two middle salaries and calculate their average to find the median. How does this all relate to the EU Pay Transparency Directive? The EU Pay Transparency Directive requires employers to report on both the mean and median gender pay gap. These gaps are calculated by finding the average pay for male employees and the average pay for female employees, and then measuring the difference between them. The difference is expressed as a percentage of men's average pay. For example, a gender pay gap of +10% means that women's average pay is 10% lower than men's average pay. But why does the Directive want to look at both figures? The size of a company's mean gender pay gap and median gender pay gap can tell us different things about how pay is distributed across the organisation. If a company's mean gender pay gap is larger than its median gender pay gap, this could suggest that higher-paid employees are having a greater influence on the overall gap. Think back to our PeopleTech Ltd. example, where the CEO's €100,000 salary pulled the mean upwards. A similar effect could occur in an organisation where the pay gap between most male and female employees is relatively small, but a highly paid position, such as the CEO role, is occupied by a man. Even though the difference in pay across the wider workforce may be relatively modest, this one high salary could pull the mean upwards and result in a much larger overall gender pay gap. On the other hand, if a company's median gender pay gap is larger than its mean gender pay gap, this could suggest that the difference in pay is more evident across the wider workforce, rather than being driven primarily by a small number of highly or lower paid positions. In other words, the gap may be more noticeable among employees around the middle of the pay distribution, rather than being concentrated at the top. This could highlight wider differences in how men and women are paid across the organisation, and may warrant further investigation. It is important to remember that neither the mean nor the median tells us exactly why a gender pay gap exists. They are indicators that help us identify patterns and differences in pay, rather than explanations in themselves. By looking at both measures together, HR professionals can begin to build a clearer picture of pay within their organisation, identify areas that may need further investigation and take a more informed approach to understanding and addressing any gaps.

Read article

Michelle Dervan

2 Nov 2026

All About Averages | Pay Analytics 101

The Importance of the Average When talking about pay data, the term average appears time and time again, but what does it actually mean? Put simply, an average salary gives us one figure to help describe pay across a group of employees. It takes all the salary values from that group and calculates a figure that would represent a ‘typical’ employee in that group. This is useful because it allows us to compare pay between different employee groups without having to look at every individual salary. The average can become an even more powerful metric when we begin to understand and analyse the different ways it’s calculated. It might seem counterintuitive that a figure like the average can be calculated in different ways and possibly provide different answers. After all, how can the same set of salaries produce different averages? The key is that each method looks at the data differently, helping us bring a different part of the picture into focus. There are two types of average that are particularly important when it comes to understanding pay: the mean and the median. So, how do they work, and what can each tell us about pay? Let's introduce them using a simple example company, PeopleTech Ltd. The Mean: What would everyone earn if pay were shared equally? The mean is what many people think of when they hear the word average. It shows us what each employee would earn if the company's total pay were shared equally among everyone. In everyday conversation, the terms mean and average are often used interchangeably. However, this isn't technically accurate, as the mean is just one of the ways to calculate an average, as we'll explore later on. To calculate the mean, we simply add together all the salaries and divide the total by the number of employees. Let's look at the example company, PeopleTech Ltd. Imagine the five employees earn the following annual salaries: €30,000 + €40,000 + €40,000 + €50,000 + €100,000 = €260,000 PeopleTech Ltd. has a total annual salary bill of €260,000 across its five employees. To calculate the mean, we divide this total by five: €260,000 ÷ 5 = €52,000 So, the mean salary at PeopleTech Ltd. is €52,000. In other words, if the company's total salary bill were shared equally between all five employees, each person would receive €52,000. The mean is useful because it gives us an overall picture of pay across the company. However, as we'll see, it doesn't always reflect what most employees actually earn, particularly when there are significant differences between the highest and lowest salaries. The Median: What is the salary in the middle? But what happens when one or two employees earn significantly more than everyone else? This is where the median can give us a different perspective. While the mean looks at the total of all salaries, the median looks directly at the middle of the data. Imagine arranging all our employees in a straight line, from the lowest-paid to the highest-paid. The salary of the person standing in the middle would represent our median salary. Let's look at PeopleTech Ltd. again. Using the same salaries as before, we arrange them from lowest to highest: €30,000 | €40,000 | €40,000 | €50,000 | €100,000 The middle salary is €40,000, making this the median salary for PeopleTech Ltd. What makes the median particularly useful is that it is not affected by unusually high or low salaries, often referred to as outliers. In our example, let's assume the €100,000 salary belongs to the CEO, which is double the next highest salary. This unusually high salary pulls the mean upwards, making the average salary appear higher than what most employees actually earn. The median, however, is not affected by this unusually high salary. At €40,000, it gives us a different perspective on what a typical employee earns in PeopleTech Ltd. This is particularly useful for HR professionals because it helps us understand what employees typically earn, without the results being skewed by a small number of particularly high or low salaries. Neither figure is wrong. They simply tell us different things about how pay is distributed across the company. Looking at both together gives us a more complete picture than relying on either one alone. One thing to note: when there is an even number of employees, there won't be one person sitting exactly in the middle. In this case, we take the two middle salaries and calculate their average to find the median. How does this all relate to the EU Pay Transparency Directive? The EU Pay Transparency Directive requires employers to report on both the mean and median gender pay gap. These gaps are calculated by finding the average pay for male employees and the average pay for female employees, and then measuring the difference between them. The difference is expressed as a percentage of men's average pay. For example, a gender pay gap of +10% means that women's average pay is 10% lower than men's average pay. But why does the Directive want to look at both figures? The size of a company's mean gender pay gap and median gender pay gap can tell us different things about how pay is distributed across the organisation. If a company's mean gender pay gap is larger than its median gender pay gap, this could suggest that higher-paid employees are having a greater influence on the overall gap. Think back to our PeopleTech Ltd. example, where the CEO's €100,000 salary pulled the mean upwards. A similar effect could occur in an organisation where the pay gap between most male and female employees is relatively small, but a highly paid position, such as the CEO role, is occupied by a man. Even though the difference in pay across the wider workforce may be relatively modest, this one high salary could pull the mean upwards and result in a much larger overall gender pay gap. On the other hand, if a company's median gender pay gap is larger than its mean gender pay gap, this could suggest that the difference in pay is more evident across the wider workforce, rather than being driven primarily by a small number of highly or lower paid positions. In other words, the gap may be more noticeable among employees around the middle of the pay distribution, rather than being concentrated at the top. This could highlight wider differences in how men and women are paid across the organisation, and may warrant further investigation. It is important to remember that neither the mean nor the median tells us exactly why a gender pay gap exists. They are indicators that help us identify patterns and differences in pay, rather than explanations in themselves. By looking at both measures together, HR professionals can begin to build a clearer picture of pay within their organisation, identify areas that may need further investigation and take a more informed approach to understanding and addressing any gaps.

Read article

The information on this page is not intended to serve and does not serve as legal advice. All of the content, information, and material on this website are only for general informational use.

Copyright © 2026 SkillsTrust. All Rights Reserved.

The information on this page is not intended to serve and does not serve as legal advice. All of the content, information, and material on this website are only for general informational use.

Copyright © 2026 SkillsTrust. All Rights Reserved.

The information on this page is not intended to serve and does not serve as legal advice. All of the content, information, and material on this website are only for general informational use.

Copyright © 2026 SkillsTrust. All Rights Reserved.

The information on this page is not intended to serve and does not serve as legal advice. All of the content, information, and material on this website are only for general informational use.

Copyright © 2026 SkillsTrust. All Rights Reserved.