How to Conduct a Pay Equity Audit: A Step-by-Step Guide for Employers

## The short answer

A pay equity audit is a structured analysis that compares what you pay employees doing similar or equivalent work, then tests whether any pay differences are explained by legitimate factors (experience, performance, location) rather than protected characteristics such as sex, race or age. To run one: define comparable role groups, gather clean pay and demographic data, run a like-for-like comparison, isolate unexplained gaps, document the reasons, and remediate where you cannot justify the difference. Done well, it is a recurring health check rather than a one-off project.

## Step 1: Decide the scope and the question

Before touching data, agree what you are testing. Most audits ask one of two questions: *Are we paying men and women fairly for equal work?* (an equal pay analysis) or *Do pay outcomes differ across groups regardless of role?* (a pay gap analysis). These are different. Equal pay compares like-for-like roles; a pay gap looks at the whole workforce average and is heavily driven by who sits in which roles.

Decide which protected characteristics you can analyse based on the demographic data you hold lawfully, and whether the audit covers base pay only or total reward (bonus, commission, allowances, equity).

## Step 2: Build comparable groups

You cannot compare a software engineer to a warehouse operative. Group employees into roles of *equal value* — work that is similar in skill, effort, responsibility and conditions. Useful grouping methods include:

- **Job architecture or grading levels**, if you have them

- **Job families** (e.g. engineering, finance, customer support) split by level

- **A job evaluation scheme** that scores roles on consistent factors

The quality of your groups determines the quality of your audit. Groups that are too broad hide real differences; groups that are too narrow leave too few people to compare.

## Step 3: Get the data clean

Pull base pay, variable pay, hours, start dates, job level, location and the demographic fields you are analysing. Common data problems to fix first:

- Pro-rate part-time salaries to a full-time equivalent before comparing

- Standardise pay to a single currency and frequency (e.g. annualised)

- Remove or flag leavers, new joiners mid-period and people on long-term absence

- Reconcile job titles that mean different things in different teams

## Step 4: Compare and isolate unexplained gaps

Within each comparable group, compare median and mean pay across the groups you are testing. Where you see a difference, the central task is to ask: *can a legitimate, consistently applied factor explain it?* Acceptable explanatory factors typically include experience, tenure, performance rating, qualifications required for the role and geographic pay zones — provided they are applied even-handedly.

For larger populations, a regression analysis lets you control for several factors at once and reveals the gap that remains *after* accounting for them. That residual, unexplained gap is what matters.

## Step 5: Document the reasons

For every gap you accept as justified, write down why, with evidence. "He negotiated harder" is not a defensible reason and often encodes bias. "She is one grade junior and joined nine months ago" is. This documentation is your defence if a claim arises and your roadmap for the next audit.

## Step 6: Remediate

Where a gap cannot be justified, plan corrections. Options include:

- Targeted pay adjustments for affected individuals

- Bringing outliers into a defined pay range over an agreed period

- Fixing the upstream process (starting-salary rules, promotion criteria) that created the gap

Never fix an unequal-pay problem by cutting the higher earner's pay — reduce the gap by raising the underpaid party. Set a budget and prioritise the clearest, largest unexplained gaps first.

## Step 7: Make it recurring

A single audit is a snapshot. Pay drifts every time you hire, promote or run a pay review. Build the audit into your annual reward cycle so you catch new gaps before they compound. Many organisations now run continuous monitoring so a flag is raised the moment a new hire's offer would create an unexplained gap.

## Common pitfalls

- **Comparing job titles instead of job content** — titles are inconsistent

- **Ignoring variable pay** — bonus and commission often hide the largest gaps

- **Auditing once and stopping** — gaps reopen quickly

- **Skipping documentation** — undermines both defence and follow-up

This is the kind of structured, data-heavy, repeatable analysis that benefits from purpose-built tooling; neart.ai builds enterprise-grade products in the HR and payroll space designed for exactly this sort of recurring fairness analysis.

## Takeaway

Run your audit in clear steps: scope it, group comparable roles, clean the data, isolate the unexplained gap, document justified differences, remediate the rest, and repeat it on a schedule. The unexplained residual — not the headline average — is the number that tells you whether your pay is fair.