Salary survey – how to do it, step by step
Share
By Mallbutiken · Fact-checked September 30, 2026 · Approximately 7 minutes reading time
All employers must conduct a pay survey every year. The purpose is to detect, remedy, and prevent pay differences between women and men that are related to gender. The work encompasses both how pay is set and differences between equal or equivalent work. For employers with ten or more employees at the start of the calendar year, there is also a requirement for written documentation.
Which employers must conduct a pay survey?
The obligation under the Discrimination Act applies even to small employers. Documentation requirements, however, vary based on the number of employees at the start of the year. The requirement for an annual survey is therefore not limited to workplaces with at least ten employees.
| Number of employees | What is required? |
|---|---|
| 1–9 | Annual survey and analysis. No specific written documentation requirement for the pay survey according to Chapter 3, Sections 13–14. |
| 10–24 | Annual survey and written documentation of the pay survey work. | 25 or more | Written documentation of the entire process of active measures, which includes the pay survey. |
Even where written documentation is not mandatory, notes are useful. They make it possible to follow up on the previous year's assessments and understand why a difference was considered objective.
How to conduct a pay survey in six steps
1. Determine responsibility, data, and cooperation
Appoint the person in charge and schedule time with those who have knowledge of tasks, salaries, and salary setting. Employers and employees must cooperate. Plan the cooperation from the start, not just when the report is finished.
Decide on a common measurement date and which types of compensation are to be examined. Document how part-time salaries and variable compensation are made comparable. Limit access to individual data to those who need the information for their work.
2. Review salary provisions and actual practice
Gather salary criteria, collective agreements where applicable, principles for supplements, benefits, and bonuses, as well as routines for recruitment and salary revision. Examine how the rules are actually applied. A neutrally formulated criterion can have a different effect when implemented.
For example, ask: is responsibility assessed in the same way in administrative and technical roles? If flexibility is rewarded, what does that mean concretely and how can the criterion be assessed without unjustified differences?
3. Group equal work
Start from the actual content of the tasks, not just the title or department. Two people with different titles can perform equal work, while the same title can hide different responsibilities. Write a brief justification for the grouping so it can be reviewed.
4. Assess which jobs are equivalent
Equivalent work does not need to involve similar tasks. Assess the job's requirements and nature as a whole: knowledge and skills, responsibility, effort, and working conditions. The job evaluation refers to the work, not how valued a certain person is.
A point-based model can help you work consistently, but the law does not require a specific point scale. Describe what each level means and check the result qualitatively. A decimal in a spreadsheet is not in itself a legal determination.
5. Analyze the differences
Examine pay differences between women and men in equal work. Also perform the comparisons required by law between female-dominated jobs and equivalent jobs that are not female-dominated. An additional comparison concerns non-female-dominated jobs with higher pay despite lower requirements.
Test every explanation against the data. "Experience" needs to be specified: what experience, why is it relevant, and is the criterion applied consistently? "The market" also needs to be made concrete. Note what has been clarified and what remains to be investigated.
6. Decide on measures and follow up
Identify salary adjustments and other necessary measures. Set the responsible person, cost, timing, and follow-up. The documentation's schedule should assume that necessary salary adjustments are carried out as soon as possible and within three years at the latest. It is not a valid reason to routinely push all measures three years into the future.
Example: a pay difference is the start of the analysis
Fictional example: In a group performing equal work, the women's comparable average salary is 32,000 SEK and the men's is 34,000 SEK. The difference is 2,000 SEK. Using the men's salary as a comparison base, the difference is approximately 5.9 percent: 2,000 / 34,000 × 100.
This alone does not show whether discrimination is occurring. The next step is to examine the individuals behind the average, the content of the work, and the factors that actually influenced the salary. In a small group, a single salary can significantly impact the average.
A useful analysis note can follow this structure: observation → possible explanation → verified data → assessment → measure or further investigation. Do not write "no discrimination" simply because the difference is small. The annual survey does not have a general safe harbor for differences under five percent.
What should the pay survey document?
- The results of the survey and analysis.
- Which adjustments and other measures are needed to address gender-related pay differences.
- Cost calculation and schedule for necessary salary adjustments.
- How last year's planned measures were implemented and evaluated.
- How the employer and employees have cooperated.
To make the conclusions understandable, the working material should also explain the selection process, measurement date, grouping, job evaluation, and calculation methods. Save the methodology description with the report so that next year's work can build on the same foundation.
