What you'll learn
- What a Pay Equity Audit Actually Is
- Privilege Structure: Why How You Run the Audit Matters as Much as the Findings
- Building the Comparable Groups: Where Most Audits Go Wrong
- What Counts as a Legitimate Explanatory Factor — and What Doesn't
- From Finding to Remediation: Getting the Fix Funded
A pay equity audit sounds like a straightforward statistics exercise, and the analysis itself often is. What trips companies up is everything around the analysis: whether it's structured to protect the findings under attorney-client privilege, whether the comparison groups are built correctly, whether the control variables are genuinely legitimate or quietly baking in the same bias the audit is supposed to catch, and — most consequentially — whether a confirmed gap actually gets funded and fixed instead of sitting in a slide deck for a year. This guide covers the methodology, the privilege structure that determines whether your own audit can be used against you, the factors that hold up as legitimate controls and the ones that don't, and how to turn a finding into a funded remediation before it becomes a bigger liability than the gap itself.
What a Pay Equity Audit Actually Is
Quick answer
A pay equity audit is a statistical analysis of whether employees performing substantially similar work are paid comparably, after accounting for legitimate, job-related factors like experience, tenure, performance, and location. It is not a simple average comparison — 'women in engineering earn 8% less than men in engineering on average' can be entirely explained by a legitimate factor like tenure distribution, or it can reflect a real, unexplained gap. The whole point of the audit is separating the two, and that separation requires a regression model, not a spreadsheet average.
The standard methodology groups employees into comparable roles — sometimes called similarly situated employee groups — based on job function, level, and location, then runs a multivariate regression that controls for legitimate compensation drivers and isolates the residual difference associated with gender, race, age, or other protected characteristics. A statistically significant residual gap after controlling for legitimate factors is what regulators and plaintiffs' attorneys look for, and it's the number that should drive your remediation budget, not the raw unadjusted average that tends to generate headlines but doesn't tell you where the actual problem sits.
Audits typically run annually or before a major compensation event — an IPO, an acquisition, a new pay transparency law taking effect in a state you operate in. The frequency should track your risk exposure: a company with high turnover, frequent off-cycle hiring, and multiple acquisitions integrating different pay structures needs to audit more often than a stable company with a mature, well-documented leveling framework, because those factors are exactly what introduces new pay gaps between audit cycles.
Privilege Structure: Why How You Run the Audit Matters as Much as the Findings
Quick answer
Run the audit under attorney-client privilege, engaging outside employment counsel to direct the analysis, not HR or People Analytics running it independently and looping in legal afterward. The distinction matters because an audit run outside privilege is discoverable in litigation — if a pay discrimination claim surfaces later, opposing counsel can request the audit, and a document showing the company identified a gap and remediated it slowly, partially, or not at all becomes powerful evidence of knowing disparate treatment. An audit run under privilege, by contrast, is generally protected from discovery, which gives the company room to investigate honestly without every finding becoming a future liability.
The practical mechanics: outside counsel retains a statistician or economist as their expert, directs the scope and methodology, and the resulting analysis and any draft findings are communicated through counsel rather than circulated broadly in email or shared drives. This isn't about hiding wrongdoing — it's about being able to run a genuinely thorough audit, including exploratory analysis that might show a larger gap than expected, without that exploratory work itself becoming a trial exhibit. Companies that skip this step because it feels like unnecessary overhead often regret it the first time a pay discrimination claim actually surfaces.
Decide before the audit starts what will and won't be shared externally or with employees, and get alignment from legal, HR, and executive leadership on that plan. Some companies publish a summary finding — 'we conducted a pay equity analysis and made adjustments totaling $X to close identified gaps' — without publishing the underlying methodology or individual-level data. Others say nothing publicly and address findings purely through the compensation review process. Either can be defensible; deciding reactively, after the audit is already complete and someone's asking what it found, is not.
A pay equity audit run without attorney-client privilege protection creates a discoverable document that shows the company knew about a gap and either fixed it slowly or didn't fix it at all — the privilege structure isn't legal theater, it's what determines whether your own audit becomes evidence against you.
Building the Comparable Groups: Where Most Audits Go Wrong
Quick answer
The single most common methodology failure is grouping employees too broadly or too narrowly. Group too broadly — comparing everyone with the title 'Manager' across the whole company regardless of function — and you dilute a real gap in one department with no gap in another, producing a misleadingly clean top-line number. Group too narrowly — treating every unique combination of title, department, and years of experience as its own comparison group — and you end up with groups of one or two people, which makes statistical significance meaningless and lets real gaps hide inside noise.
The right approach uses your job architecture, not job titles, as the grouping unit. If you have a leveling framework — IC3, IC4, M2, M3, or similar — group by level and function, since that reflects actual comparable work better than title, which varies wildly in how consistently it's applied across a company that's grown through multiple hiring managers and possibly acquisitions. Companies without a mature leveling framework often discover, in the process of preparing for a pay equity audit, that building one is a prerequisite — you can't meaningfully compare pay for 'similarly situated' employees if you don't have a consistent definition of what similarly situated means.
Location adds another layer of complexity that shouldn't be flattened away. A national pay equity analysis that ignores cost-of-labor differences between a role based in San Francisco and the same role based in a lower cost-of-labor market will show gaps that are fully explained by location-based pay bands, not discrimination. Include location tier as a legitimate control variable, but be precise about it — location adjustments should reflect your actual documented pay philosophy, not be backed into as a way to explain away a gap that a regression is showing you.
What Counts as a Legitimate Explanatory Factor — and What Doesn't
Quick answer
Tenure, level, function, location, and documented performance ratings are standard, defensible control variables in a pay equity regression. Prior salary is where audits get legally risky: several states and cities have banned the use of prior salary history in setting new hire pay specifically because it perpetuates historical discrimination, and including prior salary as a control variable in your equity analysis can end up validating the very bias the analysis is supposed to detect. If your hiring practice already excludes salary history (as required in many jurisdictions), don't reintroduce it through the back door of your audit methodology.
Performance ratings deserve scrutiny before you treat them as a clean control variable. If performance ratings themselves show a pattern correlated with gender or race — a common finding in organizations that haven't audited their performance calibration process — then using performance rating to explain a pay gap is circular: you're using a potentially biased input to explain away a potentially related outcome. Before relying heavily on performance rating as a control, it's worth running a separate check on whether ratings themselves show demographic patterns that need their own investigation.
'Negotiation' is not a legitimate control variable, even though it's an intuitively appealing explanation for pay differences — 'she just didn't negotiate as hard.' Research consistently shows negotiation outcomes themselves are affected by gender and other demographic factors, including how assertively negotiation is received differently depending on who's doing it. A pay equity analysis that treats negotiation history as an acceptable reason for a gap is building bias into the audit's own logic rather than identifying it.
Related reading
From Finding to Remediation: Getting the Fix Funded
Quick answer
An audit that identifies a gap and doesn't get funded remediation is worse, from a legal exposure standpoint, than never having audited — it creates a documented record that the company knew and didn't act with reasonable speed. Build the remediation budget request before the audit completes, not after, so finance and executive leadership are not blindsided by a number that needs to move fast. A rough estimate based on a preliminary read of the data, communicated early, gets budget conversations started while there's still time to act within the same compensation cycle.
Prioritize remediation by statistical significance and gap size, not by which department complains loudest or which fix is administratively easiest. The employees with the largest, most statistically robust unexplained gaps should be adjusted first, ideally within the same compensation cycle the audit completes in — not deferred to 'next year's planning' for a finding that's already been identified as a legal exposure. Track remediation completion the same way you'd track any other compliance deadline, with an owner and a date, not an open-ended commitment.
Communicate remediation adjustments carefully. Employees receiving an equity adjustment generally should not be told it's specifically because of a pay equity finding tied to their gender or race — this creates its own set of complications, both in how the adjustment is perceived and in what it implies about historical treatment. Standard practice is to communicate adjustments as part of a broader compensation review or market alignment process, while documenting internally, through counsel, exactly which adjustments were equity-driven and why. Get legal input on the specific language used before any adjustment conversations happen.
The hardest part of a pay equity audit isn't running the regression — it's getting finance to fund the remediation before the next salary planning cycle. Audits that surface a gap and then sit unfunded for a year are worse than not auditing at all, because now there's a paper trail showing the company knew.
Frequently asked questions
Common questions about hr strategy and how InCruiter helps teams solve them.
InCruiter Editorial Team
AI Hiring Research · Interview Intelligence · Enterprise Talent Strategy
The InCruiter editorial team covers AI-driven hiring, interview intelligence, and modern talent acquisition strategy. Our guides draw on platform data from 2,000+ hiring teams, conversations with talent leaders, and published research in industrial-organizational psychology.
