Guide · Job evaluation

Will the EU make you evaluate jobs?

Not by name. But it asks for pay structures that can compare any two jobs on four criteria, which is what job evaluation does.

Short answer: Not in so many words. The EU pay transparency directive requires pay structures that let you judge whether jobs are of equal value on skills, effort, responsibility and working conditions. It names job evaluation only as the fix after a joint pay assessment. In practice, a company with more than a few dozen distinct jobs needs a documented, gender-neutral job evaluation to meet that test.

Levels that don't compare

Picture a 180-person software company headquartered in Austin, with 70 people in Dublin and Amsterdam. It has a tidy career framework. Engineering has levels E1 to E7. Customer success has C1 to C6. Sales has its own ladder. Each was built by the leader of that function, with care. Each makes sense on its own.

Then the board's audit committee asks a question after a briefing on the EU pay transparency directive: Do we now need job evaluation? The head of people starts to answer and stops. Nobody has ever compared a C4 customer success manager with an E4 engineer. The level names look parallel but mean different things in each family. And the E4 engineers earn about 30% more than the C4 managers, most of whom are women.

The company's lawyer in Dublin points out that the question will not come from the board for long. Under the directive, any worker can ask for the average pay of colleagues doing the same work or work of equal value [1]. The first time a C4 asks whether her work is of equal value to an E4's, the company needs a better answer than "they're in different families".

This company is a composite of situations common at this size. It is not a client and the figures are illustrative.

Why the answer isn't a simple yes or no

The directive never says "every employer must run a job evaluation". What it says is close to that and in some ways harder.

The structure has to make the comparison possible. Employers must have pay structures that ensure equal pay for equal work or work of equal value. Those structures must allow an assessment of whether workers are in a comparable situation, using objective, gender-neutral criteria that include skills, effort, responsibility and working conditions. Relevant soft skills must not be undervalued [1]. Job evaluation, in one sentence, is a systematic way of comparing jobs by what they involve rather than by who does them or what they pay. That is what the directive describes.

Job evaluation is named once, as a fix. Where a joint pay assessment finds unjustified gaps, the remedy must include an analysis of the employer's gender-neutral job evaluation and classification systems, or the setting up of such systems [1]. A company that reaches that point without one builds it under scrutiny.

The EU's own guidance is voluntary. In March 2026 the European Commission and the European Institute for Gender Equality published a step-by-step toolkit on gender-neutral job evaluation and classification [2]. It offers different methods for organizations of different sizes [2] [3]. It is not binding, it is not the only way to comply and law firm commentary is clear that following it does not guarantee compliance [3] [5].

So the honest answer is: Not by name, but in substance for most companies past a certain size. The tension sits between two good goals. The first is proportionality: A 180-person company should not run the process a 20,000-person one runs. The second is defensibility: Whatever the company does has to answer the equal value question for any pair of jobs a worker might raise.

Headcount is a poor guide to how much evaluation a company needs. The number of distinct jobs matters more. A 400-person warehouse business with 15 roles can compare them all in a structured conversation. A 120-person software company with 60 distinct roles across eight functions cannot. The second company is the one most likely to be asked a cross-family question, because its jobs are so different that nobody has ever lined them up.

Five ways companies get the question wrong

Each of these is a reasonable first reaction. Each leaves a company unable to answer the equal value question when it is asked.

"We have levels, so we're covered"

What it looks like: The company points to its career framework. Every role has a level, every level has a description, so the structure is assumed to be in place.

Why smart teams fall into it: A good leveling framework does a lot of the work. It describes scope, impact and expectations. It feels like it must cover comparison too.

What it costs: Most frameworks compare jobs inside a family, not across families. They answer "is this engineer an E4 or an E5?" but not "is an E4 engineer's work of equal value to a C4 customer success manager's?" The directive's question is the second one. Work of equal value usually means different jobs, often in different functions, which is exactly where family ladders are silent. Our guide on how many job levels you need looks at ladders themselves. The equal value question needs a common yardstick across them.

"The market sets the value of each job"

What it looks like: Different families are paid differently because their markets differ. Engineers cost more to hire. The company treats market pay as the measure of a job's value.

Why smart teams fall into it: Market pricing is how most growing companies set pay. It works well for hiring. It feels objective because the company did not choose the numbers.

What it costs: The directive's criteria describe the work, not the market for it [1]. Market pay may still be part of a reason for a pay difference. It does not answer whether two jobs are of equal value. Those are separate questions and the law asks both. A company that has only market data can say what each job costs to hire. It cannot say whether the jobs are equal in value. And where female-dominated roles have long been paid less in the market, using the market as the yardstick carries that history straight into the structure.

A more defensible pattern keeps the two apart. The evaluation places jobs by their content. Where the market for one job runs well above its evaluated place, the company pays a visible premium for that reason and reviews it. The premium then has to justify itself each year, instead of disappearing into the structure.

"We'll download the EU toolkit and score everything"

What it looks like: The company takes the official toolkit's spreadsheets and has HR score every role against them over a few weeks.

Why smart teams fall into it: It is free, official and comes with templates. If the EU wrote it, it must be the safe choice.

What it costs: The toolkit is guidance, not a safe harbor [3] [5]. Its methods need real preparation, including job profiles built from questionnaires and interviews. The larger method expects a substantial scoring exercise [3] [4]. Some national systems already prescribe their own approach. In Spain, for example, a point-factor method is mandatory for companies with 50 or more workers and the official national tool, not the EU toolkit, carries the presumption of compliance [5]. In Italy, collective agreements set the main framework for equal value [6]. And a method run mechanically can still carry bias. Where weights are chosen by the evaluators, commentators warn that factors such as physical strength can end up weighted more heavily without anyone intending it [4].

"Soft skills are too vague to score"

What it looks like: The evaluation scores technical knowledge, budget responsibility and decision rights in detail. Communication, emotional effort and dealing with distressed customers are folded into a single vague factor or left out.

Why smart teams fall into it: Technical skills are easier to describe and evidence. Scoring empathy feels subjective.

What it costs: The directive says relevant soft skills shall not be undervalued [1]. EU guidance warns that evaluators tend to overlook psychosocial and interpersonal demands, which are more common in female-dominated roles [4]. A customer success manager who handles escalations all day carries real emotional effort and responsibility for client relationships. An evaluation that cannot see that will rank her job below an engineer's for reasons the law does not accept. The cost shows up in exactly the comparison the C4 is likely to raise.

"We'll deal with it after the first report"

What it looks like: Job evaluation goes on the list for later, after reporting dates are clear and the national law is final.

Why smart teams fall into it: National laws are late and some details are uncertain. Waiting for clarity seems prudent.

What it costs: The evaluation is the foundation for almost everything else the directive asks. It decides which roles count as equal work for the information rights, which categories the pay gap report uses and how any gap is explained [1]. The right to ask about equal-value pay applies to employers of every size, without waiting for a report [1]. A company that leaves the evaluation until after its first report builds its categories without it. If a joint pay assessment follows, it then builds the evaluation anyway, with workers' representatives watching and the gaps already known. Our guide on what triggers a joint pay assessment covers that sequence.

What a proportionate, defensible approach looks like

There is no single required method. There are qualities any method should have. In our view a sound approach has these.

  • It compares across families, not only within them. Any two jobs in the company can be placed on the same yardstick. The career ladders hang from it rather than replacing it.
  • It uses the law's four criteria, at least. Skills, effort, responsibility and working conditions are all present, with soft skills described as carefully as technical ones. Job-specific factors can be added where they matter.
  • It scores the job, not the person. Evaluators work from the job's content, not from who holds it or what they earn. Current pay is kept out of the room until the scoring is done.
  • Its weights are explained and tested. If some factors count more than others, the reason is written down. The result is checked to see whether the weighting favors roles mostly held by one sex.
  • It fits the company's size. A company with 25 distinct jobs does not need the method built for 2,500. A simpler method that is documented and applied consistently is stronger than an elaborate one nobody maintains.
  • It involves the people affected. Workers' representatives, where they exist, see the criteria and the method before the results. Where there are none, employees can at least see how jobs are compared.
  • It connects to everything else. The same evaluation feeds the grades, the pay ranges, the categories in any pay gap report and the answers to individual pay requests.
  • Disagreements have a home. When a manager or employee disputes where a job sits, there is a named way to raise it and a small group that decides, using the same criteria. The decision and its reason are recorded.
  • It is kept current. New and changed roles are evaluated when they are created, not at the next big review.

None of this is a method to copy. Each quality is a test a reviewer can put to whatever approach a company already has.

The calls only you can make

Several of the choices behind job evaluation are about what the company believes, not what the law requires.

How much weight the market gets. Some companies let internal value set the structure and handle market premiums as visible, separate additions. Others start from the market and use evaluation as a check. Both can be defended. They lead to different pay structures and different conversations.

Whether every family shares one grade structure. One structure across the company makes equal value comparisons straightforward. Separate structures per family follow markets more closely and make cross-family questions harder to answer.

Which method to use. The EU toolkit, a national method where one exists, a commercial framework or one built for the company. The choice depends on country, size and how much the company wants to own its method.

Who does the scoring. An internal panel knows the jobs and owns the result. An outside evaluator brings consistency and distance from the people in the roles. Many companies combine the two: Outside scoring, then a calibration session with managers who know the work. Who sits on the panel shapes how much the result is trusted, especially by the people whose jobs it places lower than they expected.

How open to be. Some companies publish their evaluation criteria and how jobs map to grades. Others share only the result. More openness answers employee questions before they are asked. It also invites challenge.

What to do when evaluation and pay disagree. An evaluation will often place two jobs together that the market pays very differently. Closing that gap costs money. Explaining it requires a reason the company is willing to stand behind.

The law and the guidance

  • The directive. Pay structures must allow assessment of equal value on objective, gender-neutral criteria including skills, effort, responsibility and working conditions, without undervaluing soft skills [1]. Member states must provide analytical tools to support this [1]. Job evaluation is named directly as part of remedying gaps after a joint pay assessment [1].
  • The deadline. Member states had until 7 June 2026 to transpose the directive [1]. Our note on EU pay transparency, country by country tracks where each stands.
  • The EU toolkit. Published on 26 March 2026 by the European Institute for Gender Equality with the European Commission, it sets out methods for micro, small and medium and larger organizations [2]. Sources describe the size bands for each method differently, so check the toolkit itself [3] [4]. It is voluntary [3].
  • National methods. Spain already requires a point-factor method for employers with 50 or more workers. Germany's existing tools are expected to be adapted rather than replaced. France's existing reporting system is unlikely to change much [5]. Italy relies on collective agreements as the main framework [6].

Our view

We think the directive does not need to say "job evaluation" for the answer to be yes, in practice, for most companies with more than a few dozen distinct jobs. It asks a question only job evaluation answers well: Is this work of equal value to that work?

We would start with what the company has. Good leveling frameworks are most of the way there. What they usually lack is a common yardstick across families and careful treatment of soft skills. Adding those, in a method sized to the company and documented well enough to explain, is a few weeks of work. Building it later, inside a joint pay assessment, is a different kind of project.

We would argue hardest against treating market pay as the measure of value. It is the most common answer companies give and the one the directive was written to move past.

Sources

  1. Directive (EU) 2023/970 on pay transparency, Articles 3, 4, 7, 9, 10 and 34 (EUR-Lex), checked 2026-10-25
  2. European Institute for Gender Equality, EU-wide guidelines on gender-neutral job evaluation and classification: Step-by-step toolkit (26 March 2026), checked 2026-10-25
  3. Baker McKenzie, European Union: Toolkit launched to support pay transparency compliance (7 April 2026), checked 2026-10-25
  4. Lewis Silkin, EU guidelines for gender-neutral job evaluation: Compliance with the Pay Transparency Directive (27 March 2026), checked 2026-10-25
  5. Littler, Gender-neutral job evaluation in the EU: Assessment of the utility of the new EU toolkit (2026), checked 2026-10-25
  6. Littler, Italy implements the EU Pay Transparency Directive: A guide to the final decree (4 June 2026), checked 2026-10-25

What triggers a joint pay assessment?

A 5% gap in any category of workers can trigger an EU joint pay assessment. Five ways mid-size employers get caught and what a sound position looks like.

Which job evaluation method to use?

Pick the lightest method that still defends equal value. Ranking, paired comparison or points at about 200 roles: Five traps and what a sound choice looks like.

Pay decisions the EU wants written down

EU staff must be able to see the criteria behind their pay and progression. Five traps in writing them down and what sound, checkable criteria look like.

See Resources for every guide and the self-check.