Run a pay equity analysis

What this achieves

A pay equity analysis compares average pay between groups of employees and reports where the difference between them is large. It produces cohort comparisons and a list of individuals whose pay sits away from their grade midpoint.

Required role: Administrator.

Before you start

Grades need to be assigned, because individual outliers are measured against the midpoint of the employee’s grade. Employees without a grade cannot be placed against one.

Read the section on small groups below before you choose your dimensions. It is easier to decide what you will look at than to unsee a result.

Steps

Go to Compensation → Pay Equity.
Select Run New Analysis.
Enter a Name.

Example: Q2 2026 Pay Equity Review.
Choose the Scope, and the Department if you scoped it to one.
Choose the Dimensions to compare on.
Select Run Analysis.
Open the analysis under Past Analyses to read it.

Options on this screen

Option	Description
Name	What the analysis is called. It identifies the run in the list afterwards.
Scope	Company compares across the whole organisation; Department limits the analysis to one.
Dimensions	Gender, Department, and Grade are the comparisons available.

What the cohorts table shows

Each row compares two groups on one dimension.

Option	Description
Dimension	Which comparison the row is making.
Group A and Group B	The two groups being compared, each with the number of people in it.
Avg A and Avg B	The average salary of each group.
Gap %	The difference between the two averages, as a percentage.
Severity	How large the gap is, from None through Low, Medium, and High to Critical.

An overall equity score summarises the run. The screen states how to read it: higher is better, and 100 means no significant cohort gaps were detected.

A gap is a difference between two averages. It is not, by itself, an explanation — two groups can differ in average pay because of grade mix, tenure, location, or role, and the analysis compares what it is told to compare rather than adjusting for those things.

Flagged employees

Alongside the cohorts, the analysis lists individuals whose salary sits away from what their grade midpoint would suggest, showing their Current salary, an Expected figure, and the gap between them.

The Recommendation column is the product’s suggestion for a person to consider. It is input to a decision, not a decision — an increase happens only if someone proposes it through a compensation cycle or records a pay change, and both of those are separate deliberate actions.

Small groups identify people

Every cohort shows how many people are in it, and the smaller that number is, the more the average tells you about individuals rather than about a group.

This compounds when dimensions are combined with a narrow scope. A department is already a small group at most companies; splitting a small department by a second dimension can leave a cohort of two or three people whose pay can be worked out by anyone who knows who they are.

Example: at HC Corp UK Ltd, Engineering has 34 people and Finance has 9. A gender comparison inside Finance is a comparison between two handfuls of named individuals, and the People department at six is smaller again. The same comparison across the whole company is a different kind of result.

Decide before you run what group sizes you are willing to look at and who will see the output, and apply that rule to the results you like as well as the ones you do not. The product reports the counts; the judgement about what is safe to circulate is yours.

Where this stops

This analysis describes your own pay data. It does not tell you whether a gap is lawful, what you are obliged to report, or what you must do about anything it shows.

Those are questions for your own legal and reward advisers, who can weigh the result against the obligations that apply to you in the places you employ people. Take the figures to them rather than acting on a severity label.

What happens next

The analysis is kept under Past Analyses, so a later run can be compared against it. Any pay change arising from what you find is made through a compensation cycle or as a pay change on the employee record — the analysis itself changes nobody’s salary.