How to Run a Pay Equity Audit Before Your Ranges Go Public

Why the Day You Publish Ranges Is the Worst Time to Discover a Pay Gap

Picture this: your 80-person professional-services firm is three weeks from its California job-posting deadline. Your HR team — you, plus a part-time coordinator — has just finished building salary ranges for the first time. The bands look clean on a spreadsheet. Then a senior account manager walks into your office and asks, calmly, why her job posting lists a range that starts $12,000 below what the company is already paying a male peer in the same role.

You don't have a good answer, because you didn't look.

This is the single most common pay-transparency miscalculation small and mid-size employers make: treating range publication as the finish line when it is, in fact, a starting gun. Once ranges are visible — in postings, to current employees, or to regulators — any gap between what you're paying people and what the range implies becomes a documented, reviewable, potentially litigable fact.

Under California SB 642, effective January 1, 2026, the statute of limitations for willful pay-scale civil actions runs six years. Each non-compliant posting is its own violation event. That is not a compliance environment that forgives an audit you skipped.

The good news: a structured pay equity audit before you publish is entirely achievable for an SMB HR team without a dedicated compensation analyst. This article walks you through the five core steps, what to do when you find a gap, and how to document everything so the audit itself becomes a defensible record.

Step 1 — Pull a Complete, Matched Pay Dataset

A pay equity audit only works if the data underneath it is clean and complete. Before you open a pivot table, build a single flat file that includes, for every active employee:

• Employee ID (not name — use IDs for the working file to reduce anchoring bias during analysis)
• Job title and job band/level as defined in your leveling framework
• SOC code (US) or NOC code (Canada) if you've mapped roles to a benchmark occupation
• Current base salary or hourly rate
• Hire date and date of last pay change
• Time in current role/level (different from tenure — someone can have five years at the company but two years at this level)
• Location (state/province — relevant for multi-jurisdiction employers)
• Performance rating from the most recent cycle, if ratings exist and are applied consistently
• Any equity adjustments or off-cycle increases in the prior 24 months

The protected-class fields — gender, race/ethnicity — belong in a separate file keyed to employee ID, joined only for the regression or group analysis step. Keeping them separate while you clean the pay data reduces the risk of unconscious anchoring.

If your data lives in a spreadsheet today, take one careful pass before analysis: research consistently finds that the overwhelming majority of business spreadsheets used in real decisions contain errors. A single formula error in a salary column corrupts every ratio you calculate downstream. If you're unsure, have a second person spot-check a sample of rows against the source system of record before you proceed.

Step 2 — Define Comparable Groups (This Is Where Most Audits Go Wrong)

The legal and analytical standard for pay equity is not "people with the same title." It is people doing substantially similar work — comparable in skill, effort, responsibility, and working conditions. If your organization hasn't mapped roles to a formal job band structure, now is the time to do that groundwork, because the validity of every finding downstream depends on whether your groupings are defensible.

For each comparable group you define, document:

1. The job band or level (e.g., Individual Contributor — Level 2)
2. The functional family (e.g., Operations, Finance, Customer Success)
3. The compensable factors that distinguish this group from adjacent levels — scope, decision authority, required credentials, customer impact

If you need a starting framework, see our complete guide to job band structure for how to build job levels that hold up under scrutiny.

One practical judgment call at this stage: how finely do you slice? A group of two people is not analytically meaningful, but a group of 40 people across three genuine sub-functions may obscure real inequities within those sub-functions. A reasonable rule of thumb: groups of five or more are the minimum for meaningful statistical analysis; if a group is smaller, flag it for qualitative review rather than excluding it.

Step 3 — Calculate Compa-Ratios and Flag Outliers

Once your groups are defined, the primary analytical tool is the compa-ratio — the ratio of an employee's current pay to the midpoint of their assigned salary band. A compa-ratio of 1.00 means the employee is paid exactly at midpoint; 0.85 means they're paid at 85% of midpoint; 1.15 means 15% above.

A compa-ratio tells you where an employee sits relative to the market anchor. A pay equity audit tells you whether the pattern of compa-ratios across demographic groups has a defensible explanation.

For a primer on calculating and interpreting compa-ratios, see Compa-Ratio Explained.

With compa-ratios calculated for every employee in a comparable group, you're looking for two types of signals:

Distribution spread within a group. If the compa-ratio range in a 10-person group runs from 0.80 to 1.25, that spread represents a 45-point difference in pay positioning. That is not automatically a problem — but it requires explanation. What compensable factors account for it?

Demographic clustering. Join your protected-class file and ask: do employees of a particular gender, race, or ethnicity cluster systematically at the low end of the compa-ratio distribution within a group, even controlling for tenure in role and performance rating?

Worked example (illustrative — not drawn from actual employee data):

Suppose you have a comparable group of eight customer success managers at Band IC-2, with a band midpoint of $72,000. After calculating compa-ratios, you find:

Employee	Gender	Yrs in Role	Salary	Compa-Ratio
A	F	3.5	$61,200	0.85
B	F	2.0	$63,000	0.875
C	F	4.0	$64,800	0.90
D	M	2.5	$72,000	1.00
E	M	1.5	$70,560	0.98
F	M	3.0	$75,600	1.05
G	F	5.0	$67,320	0.935
H	M	4.5	$79,200	1.10

The female employees average a compa-ratio of approximately 0.89; the male employees average approximately 1.03. The gap is not explained by time in role — Employee A has 3.5 years in role and a lower compa-ratio than Employee E with 1.5 years. This group warrants a root-cause review.

The US Bureau of Labor Statistics has documented a persistent national gender pay gap: in 2023, women working full-time had median weekly earnings of $1,005 — 83.6% of the $1,202 median for men. An audit that surfaces a similar pattern within your organization is not a fluke; it is a concrete version of a structural dynamic that has been measured at national scale. The question your audit must answer is whether any of the gap has a legitimate, documented business explanation — and whether the rest gets remediated.

Step 4 — Distinguish Explained from Unexplained Gaps

Not every gap is inequity. Compensation differences within a group can reflect:

• Legitimate compensable factors: a specialized certification, geographic differential pay, a retention agreement tied to a documented flight risk, a premium paid at hire during a tight labor market for a scarce skill
• Administrative drift: no explanation, no documentation — just different managers making uncoordinated decisions over time
• Systemic inequity: a pattern where a protected class consistently bears the administrative drift while another consistently receives the benefit of doubt

Your job at this step is to sort each identified gap into one of these three buckets. For every employee whose compa-ratio sits more than a defined threshold below group average (a common starting point is 5 percentage points below the group median), document the review:

1. What compensable factors, if any, are on file that explain this position? Documented performance history, hire notes, offer letters, retention agreements.
2. Are those factors consistently applied across the group? If one employee's below-midpoint position is explained by "hired during a slower market," is that explanation also applied to similarly-situated employees who were hired in the same window?
3. Is there a pattern by protected class within this explanation category?

If a gap is unexplained — no documented factor, no consistent application — it belongs in the remediation queue. If a gap is partially explained, it may warrant partial remediation and fuller documentation going forward.

This analysis is the core of what makes a pay equity audit defensible. The goal is not to produce a clean-looking spreadsheet; it is to produce a documented record showing that the organization reviewed pay, found explanations where they exist, and acted on gaps where they don't. See our guidance on building an audit trail for compensation changes for how to structure the documentation layer.

Step 5 — Build a Remediation Plan Before You Publish

Finding a gap and publishing ranges anyway — without a remediation plan — is worse than not auditing. It creates a documented record of a known problem left unaddressed.

A remediation plan for SMB HR teams should be structured and realistic:

Prioritize by magnitude and legal exposure. Employees in jurisdictions with active pay transparency enforcement (California, Colorado, and now Ontario and British Columbia) where you are about to publish ranges should be addressed first. An unexplained gap that becomes visible through a posted range in a state with per-violation penalties is a compounding risk.

Model the cost before the conversation. Calculate the total annualized cost of bringing below-range employees to a defensible position — either to the band minimum or to a percentile that reflects their compensable factors. For most SMB organizations, the number is smaller than expected, and knowing it in advance prevents sticker shock from derailing the process.

Set a timeline by cohort, not all at once. If budget constraints prevent full remediation in a single cycle, document a phased plan with committed dates. A phased plan with documented intent is a stronger legal position than no plan.

Update the ranges and the data together. Once remediation is processed, update your compensation system of record so compa-ratios reflect the new positions. This is the moment to establish a cadence — most SMB organizations run a pay equity review annually, aligned with the merit cycle.

One downstream risk that often goes unmanaged: employees who learn their pay is below the published range — whether from a new posting or from a direct conversation — may reconsider their tenure. The connection between pay positioning and attrition risk is real; see how pay transparency affects attrition risk for a fuller treatment.

For the documentation and compliance coverage of the remediation itself, the Pay Transparency Compliance Hub has jurisdiction-specific checklists for California, Colorado, Ontario, and British Columbia.

What a Completed Pay Equity Audit Looks Like

When you've run all five steps, you should have:

• A matched, cleaned pay dataset with compa-ratios for every active employee by comparable group
• A documented classification of every identified gap: explained (with the factor on file), partially explained, or unexplained
• A remediation queue with cost model and timeline for unexplained gaps
• A sign-off record: who reviewed, what was found, what action was taken, and when

That documentation package is what gives a pay equity audit its legal and organizational value. It doesn't need to be elaborate — a structured spreadsheet with a consistent decision log is sufficient for most SMB organizations. What matters is that the review was real, the findings were recorded, and the remediation was acted on.

One more thing worth stating plainly: this audit is not a one-time project. Publishing ranges for the first time is the catalyst; the audit itself should become an annual process, run before each merit cycle close and before any new job-posting push in a regulated jurisdiction.

---

If you'd rather start with a pre-built structure than build the analysis from scratch, the Pay Equity Audit Workbook gives you a ready-to-use template: the comparable-group definition sheet, the compa-ratio calculator, the gap classification framework, and the remediation cost model — all in one structured file designed for HR teams without a dedicated compensation analyst.

If you're at the stage of building the underlying bands before you can run the audit, Job Band Builder's pricing plans start with the tools you need to define levels, set min/mid/max ranges against BLS OEWS benchmarks, and generate the structured data the audit requires. ```