Skip to content
CompensationMay 11, 2026 11 min read

Pay equity in 30 days: the operator's playbook nobody publishes.

Most pay-equity guides are written by consultants who want a 6-month engagement. Here's how to run a defensible audit in 30 days with a spreadsheet, a statistician's hour, and the right framing…

PJ
Pawan Joshi
Global HR & Operations
Share

If you Google 'pay equity audit,' you'll find a hundred pages telling you to hire a consultant. That's because the people writing the pages are the consultants. The truth is uglier and more useful: a competent People + Finance team can run a defensible first-pass audit in 30 days, find the 80% of gaps that matter, and queue up remediation before the next compensation cycle.

Here's the playbook. It's not the gold-standard regression-on-everything version — that's a 90-day project with legal privilege. This is the version that catches the indefensible gaps before they become a lawsuit, a Glassdoor post, or a regulator's letter.

Why 30 days, not 6 months
63%
of identified pay gaps are explainable by 4 variables: role, level, location, tenure
Payscale 2024 study
$3M+
median cost of a Title VII pay discrimination settlement (US)
EEOC, 2024
11 states
in the US now require proactive pay-equity reporting
As of Jan 2026
30 days
is enough to catch the gaps that explain 80% of legal and brand risk
Author benchmark
6 sections · tap to expand

Week 1 — Get the data right

  • Pull comp, role, level, location, tenure, gender, race, manager from HRIS into one flat file.
  • Engage legal (in writing) under privilege before any analysis. This single email protects the work product.
  • Define the 'similarly situated' bucket: same job family + same level + same country. Smaller buckets are noise.
  • Decide your pay measure: base only, base + bonus, or total cash. Be consistent.

Week 2 — Run the unadjusted gap

  • For each bucket, compute mean and median pay by gender and (where you have it) race.
  • Flag any bucket where the gap exceeds 5% OR the bucket is too small to be statistically meaningful (n<10).
  • Resist the urge to explain gaps away yet. Just list them.

Week 3 — Adjust for legitimate factors

  • Add controls one at a time: level, tenure, performance rating, location.
  • After each control, recompute. The 'adjusted gap' is what remains after legitimate explanations.
  • Anything above ~2% adjusted gap is a real issue. Anything above 5% is urgent.

Week 4 — Remediate and document

  • For each unjustified gap, propose a remediation amount that brings the person to the bucket's adjusted-fair midpoint.
  • Bundle remediations into the next comp cycle when possible (cleaner narrative, less individual exposure).
  • Document method, controls, decisions, and edge cases. This is what makes the audit defensible.
  • Schedule the next audit for 6 months out. Pay equity is a maintenance discipline, not a project.
Anti-patterns that turn an audit into a lawsuit
Don't
  • Run the audit without legal privilege
  • Share raw results in a wide Slack channel
  • Remediate downward (cut anyone's pay)
  • Promise 'no gap' publicly before remediation ships
  • Wait for the perfect dataset before starting
Do
  • Engage legal in writing on day 1
  • Limit access to need-to-know on a single drive
  • Remediate up; freeze overpaid if needed
  • Publish methodology and direction of travel
  • Start with the data you have, iterate

Most pay gaps don't come from a sexist manager in a corner office. They come from a well-documented cognitive bias: anchoring. Kahneman's research shows that the first number mentioned in a salary negotiation sets the reference point for everything that follows — and decades of audit-study research (Goldin & Rouse, Bertrand & Mullainathan) show that women and minority candidates systematically state lower anchors, accept the first offer more often, and are coached less aggressively by recruiters. Multiply that across thousands of hiring decisions and you get a 7-12% structural gap with no individual villain.

The implication for your audit: looking for the bad actor misses the mechanism. Look for the patterns — which job families, which managers, which hiring quarters — where anchoring effects compounded. That's where remediation has the most leverage.

What changed in 2025-2026
16 US states
now require salary range posting on job ads as of Jan 2026
NCSL state tracker, 2026
EU Directive
Pay Transparency Directive enforced June 2026 — requires gap reporting for orgs >150 employees
EU Council, 2023/970
$28M
largest US pay equity settlement of 2025 (single employer)
EEOC enforcement summary, 2025
3 years
median lookback period regulators use when calculating back-pay liability
EEOC guidance, 2024

A 400-person healthtech ran their first audit in 2024. The unadjusted gap was 11.4%. After controlling for level, tenure, and location, the adjusted gap was 2.8% — meaningful but not catastrophic. The interesting finding was that 78% of the residual gap came from one job family (sales) and one practice (negotiation-based offers). They didn't restructure comp company-wide. They moved sales offers to a fixed grid based on level + experience and re-audited 6 months later. Adjusted gap dropped to 0.6%. The total remediation cost was $340K — less than one Title VII filing.

  • Engage legal under privilege in writing on day 1, not day 10.
  • Use 4 controls only for the first pass: role, level, location, tenure. Resist 'one more variable.'
  • Set a hard threshold: anything >2% adjusted gap in a bucket of n≥10 gets remediated.
  • Never remediate downward. Freeze and let market catch up if you must.
  • Bundle remediation into the next comp cycle for narrative cleanliness.
  • Schedule the next audit for 180 days from today. Put it on the calendar before this one ships.
  • Publish methodology and direction-of-travel externally — not the numbers, but the discipline.
Found this useful? Share it.