What is the difference between a pay equity audit and a pay equity analysis?

In practice the terms are used interchangeably. When distinguished, the analysis is the statistical work of comparing pay within comparable groups, and the audit is the full exercise around it: scoping, data collection, the analysis itself, remediation and documentation.

Pay equity audit: how to conduct a pay equity analysis, step by step

Q: What is a pay equity audit?

A pay equity audit is a structured review of compensation data to find pay differences between employees doing comparable work that cannot be explained by legitimate factors such as experience, performance, location or tenure. It compares pay within groups of comparable jobs, tests whether gaps track protected characteristics like sex or race, and produces a remediation plan for the differences that have no defensible explanation.

Q: How do you conduct a pay equity audit?

Six steps: define scope and involve counsel to protect the work, pull compensation and demographic data from payroll and HRIS, group comparable jobs, analyze pay within each group using cohort comparison or regression depending on company size, remediate unexplained gaps with a funded plan, and document the methodology so next year starts from a baseline.

Q: How often should you do a pay equity audit?

Annually is the working standard, timed a quarter before the merit cycle so findings can be fixed inside it rather than as awkward off-cycle adjustments. Companies growing fast or entering pay transparency states often add a lighter mid-year check on new hires, since new offers are where drift usually re-enters.

Q: What data do you need for a pay equity audit?

Base salary, bonus and equity for every employee, plus the legitimate factors that explain pay: job title and level, department, location, hire date, time in role, performance ratings and relevant prior experience, along with self-reported demographic data such as sex and race or ethnicity. Most of it lives in payroll and the HRIS; the usual gap is a clean job architecture that says which roles are comparable.

Q: Are pay equity audits legally required?

No US federal law requires a pay equity audit. The federal Equal Pay Act and Title VII prohibit discriminatory pay, and several states sweeten voluntary audits: the Massachusetts Equal Pay Act gives employers who complete a good-faith self-evaluation and make progress on fixing gaps an affirmative defense to state pay discrimination claims. Separately, California requires large employers to file annual pay data reports, which is a disclosure duty rather than an audit.

Jul 20, 2026 11 min read By the Wagelist team

A pay equity audit is a structured review of your compensation data to find pay differences between people doing comparable work that you cannot explain with legitimate factors such as experience, performance, location or tenure. The output is three lists: gaps that are explained, gaps that need fixing, and the fixes with a price tag. For a company under 200 people the whole exercise typically takes a few weeks, and most of that is data cleanup, not statistics.

This guide walks through the six steps, sized for small and mid-size companies rather than the Fortune 500 version with a consulting team attached. It also covers the legal privilege question you should settle before pulling a single row of data, and the audit's quiet dependency: you cannot test pay against a standard if you have no standard, which is why salary bands come first.

Why audits stopped being optional

No federal law forces you to run one. The federal Equal Pay Act of 1963 and Title VII prohibit discriminatory pay, and state equal pay acts in California, New York, Colorado, Massachusetts and elsewhere apply broader comparable-work standards. What changed is visibility. In the states covered by pay transparency laws, your ranges are now public in every posting, and your own employees read them. A gap you have not found is a gap someone else will, either an employee comparing a posting to their paycheck or a plaintiff's lawyer doing the same at scale.

There is also a carrot. Under the Massachusetts Equal Pay Act, an employer that completes a good-faith self-evaluation of its pay practices and makes reasonable progress on fixing what it finds gets an affirmative defense to state pay discrimination claims. Other states have looked at similar safe harbors. And California's separate pay data reporting regime means the state already receives large employers' pay distributions by demographic group every year, whether or not the employer has looked at the numbers itself. Auditing first means you see your problems before a regulator does.

Step 1: Define scope, and involve counsel before you start

Decide what you are testing (base salary only, or base plus bonus and equity), which populations are in scope, and which protected characteristics you will test against. Sex and race or ethnicity are the standard pair, because they map to the data most companies already collect for EEO purposes.

Then the uncomfortable part: talk to employment counsel before running the numbers. An audit you run casually produces a discoverable spreadsheet that says, in your own words, that you found unexplained gaps and by how much. An audit directed by counsel for the purpose of legal advice can often be protected by privilege, which changes what happens to the findings if you are ever sued. This is not a reason to skip the audit; unremediated gaps are the real exposure. It is a reason to set the structure up properly, decide in advance that you will fund fixes for whatever you find, and keep the working files tight.

Step 2: Pull the data

The dataset is one row per employee, and every column is either pay or a legitimate reason pay differs:

Column group	Fields	Usual source
Compensation	Base salary, target bonus, most recent bonus paid, equity grant value, FTE percentage	Payroll, cap table
Job architecture	Title, level, job family, department, manager, location or pay zone	HRIS
Legitimate factors	Hire date, time in role, last two performance ratings, relevant prior experience	HRIS, performance tool
Demographics	Sex, race/ethnicity (self-reported, EEO categories)	HRIS / EEO records

Expect the cleanup to take longer than the analysis. Titles will not match levels, two departments will use the same title for different jobs, and someone's "experience" will turn out to be a guess typed in three years ago. Fix the rows, not the findings.

Step 3: Group comparable jobs

The audit's central judgment call is which jobs are comparable. State standards phrase it as equal work, substantially similar work, or comparable work, but the operational version is the same: group roles by skill, effort, responsibility and working conditions, not by title. A "customer success manager" who renews enterprise contracts and one who answers support tickets are different jobs wearing the same title.

If you maintain salary bands, the grouping is nearly done: a band is already a claim that the roles inside it are comparable, anchored to a level and a market rate. That is the practical reason to build bands before auditing. Testing pay inside band-and-level groups also gives every finding an immediate reference point: compa ratio against the band midpoint tells you who inside the group is paid where, and why a gap exists in dollars rather than abstractions.

Step 4: Analyze, sized to your headcount

The method depends on how many people sit in each comparable group, not on ambition:

Company size	Method	What it looks like
Under ~50	Side-by-side cohort review	List everyone in each group with pay, tenure, performance and compa ratio, and explain every gap over a threshold you set in advance (5 percent is common). Statistics add nothing at this size; judgment is the tool.
~50 to 250	Cohort review plus group medians	Same review, plus median pay and median compa ratio by sex and by race within each group. Medians beat averages here: one founder-era salary can drag an average anywhere.
250+	Multiple regression	Model pay as a function of the legitimate factors, then test whether sex or race still predicts pay after controlling for them. This is where a statistician or a specialized platform earns its fee.

Whichever tier you are in, run the queries you will actually reread: median pay by group and sex, compa ratio distribution by group and race, new-hire offers of the last 12 months versus the incumbents they sat next to. If SQL is the bottleneck between you and your own HRIS export, an AI data analyst that takes plain-English questions gets you the group-by tables without waiting on an engineer.

One trap at every size: do not average away your findings. A healthy company-wide number can hide a serious gap in one department, which is exactly where a claim would be filed from. Run the analysis at the comparable-group level and only then roll it up.

Step 5: Remediate, with a budget and a sequence

Findings sort into three piles. Gaps explained by legitimate factors get documented and closed. Gaps that are really banding errors (someone hired under the range minimum, a band nobody refreshed) get fixed through the band. Unexplained gaps that track a protected characteristic get raises, and the raises go to the underpaid person; lowering anyone's pay to close a gap is illegal under the federal Equal Pay Act and its state counterparts.

Two practical rules. First, fund remediation before the audit starts, because finding a gap and sitting on it is worse than not having looked; a typical outcome for a first audit lands somewhere under 1 percent of payroll, but your number is your number. Second, fix the process that created the gaps, or next year's audit rediscovers them: anchor offers to the band, not to negotiation stamina, and check geographic pay differentials that quietly encode a demographic pattern.

Step 6: Document, and put it on the calendar

Write down the methodology: which groups, which factors, which thresholds, what you found in aggregate, what you fixed and when. That document is the difference between a one-off cleanup and a defensible practice, and in Massachusetts it is literally the substance of the affirmative defense. Then schedule the next one. Annual, one quarter before the merit cycle, is the cadence that lets findings become merit adjustments instead of awkward off-cycle raises. Fast-growing teams add a mid-year check on new offers, because offers are where drift re-enters.

Frequently asked questions

What is a pay equity audit?

A structured review of compensation data to find pay differences between employees doing comparable work that legitimate factors such as experience, performance, location and tenure cannot explain. It ends with a remediation plan for whatever cannot be defended.

How do you conduct a pay equity audit?

Define scope and involve counsel, pull pay and demographic data from payroll and the HRIS, group comparable jobs (bands make this nearly automatic), analyze within groups with cohort review or regression depending on size, fund and apply fixes, then document the method and repeat annually.

Are pay equity audits legally required?

Not by US federal law. Equal pay statutes prohibit discriminatory pay but do not mandate audits. Massachusetts rewards a good-faith self-evaluation with an affirmative defense to state claims, and California separately requires 100+ employee companies to file annual pay data reports with the state.

How often should you do a pay equity audit?

Annually, timed so the findings can be paid for inside the regular merit cycle. Add a lighter mid-year review of new-hire offers if you are hiring fast, since new offers are the main way gaps reappear between audits.

What data do you need for a pay equity audit?

Pay (base, bonus, equity), the legitimate factors (level, department, location, tenure, performance, relevant experience) and self-reported demographics. The hard part is usually not the data but the job architecture that says which roles are comparable.

Can you lower salaries to fix pay equity gaps?

No. The federal Equal Pay Act and state equivalents explicitly prohibit reducing any employee's pay to equalize a difference. Remediation means raising the underpaid side, which is why the budget conversation belongs at the start of the audit, not the end.

The audit is the checkup; bands are the diet

A pay equity audit finds the damage after the fact. Salary bands prevent most of it from happening: when every offer, raise and promotion is anchored to a documented band, the unexplained gaps mostly stop appearing, and the annual audit turns into a fast confirmation instead of a cleanup. If you are starting from zero, build the bands first with the pay band calculator, then run your first audit against them; the posted ranges your state now requires will come out of the same, already-audited numbers.