Compensation Equity Audit

Ask where pay gaps are hiding and get a clear, adjusted breakdown by gender and level in seconds, so you walk into comp planning knowing exactly which roles need a second look.

Für Compensation Teams HR Business Partners People Analytics Legal & Compliance
Beispiel-Prompt für den Joy-Assistenten

Analyze our compensation data and show pay gaps by [gender] within [job family], flagging any adjusted difference wider than [5%] so I know where to look first.

So kann es aussehen

Ein anschauliches Beispiel. Ihr Ergebnis hängt von Ihren Quellen, Daten und Anfragen ab.

Analyze our compensation data and show pay gaps by gender within software engineering, flagging any adjusted difference wider than 5% so I know where to look first.

I compared base pay for the 214 people in software engineering, controlling for level, location, and tenure, using the compensation data connected to your workspace. Most levels come in within about a point, but two cross your 5% line.

Across the four engineering levels, the adjusted gap is small at the bottom and widens sharply at the top. Only one level clears your 5% threshold outright.

Adjusted pay gap, women vs. men, by level: software engineering (negative means women paid less)

L3 Engineer-0.8%
L4 Senior-1.2%
L5 Staff-6.4%
L6 Principal-4.1%

The L5 Staff level is the one to look at first: women are paid 6.4% less than men after adjusting for tenure and location, and that band holds 38 people, so it's not a one-person artifact. L6 Principal is next at 4.1%, under your threshold, but close enough to watch given it covers only nine people, where a single outlier moves the number.

The lower levels are clean: L3 sits at -0.8% and L4 at -1.2%, both inside normal noise. If you close the L5 Staff gap in this cycle, the department-wide adjusted gap drops from 3.1% to roughly 1.4%.

Want me to break the L5 Staff gap down by manager so you can see whether it traces to specific teams or hiring cohorts?

From question to flagged gaps in seconds

Compensation Equity Audit reads your connected compensation data at the moment you ask and returns an adjusted comparison (controlling for level, tenure, and location) so you're looking at real gaps instead of a raw average skewed by seniority mix.

  1. Connect your compensation data

    Point Joy at the compensation dataset you already maintain. It reads the fields it needs (base pay, level, tenure, location, gender) at the moment you ask.

  2. Describe the cut you need

    Tell Joy the dimension and the population: pay gaps by gender within software engineering, flagging anything past 5%. Set the threshold that matters to you.

  3. Review the chart and the read

    Joy returns an adjusted breakdown by level with a written takeaway, calling out which levels cross your threshold and by how much.

  4. Use it where you work

    Copy the figures and the takeaway straight into your comp-planning deck or your note to legal. Joy drafts the analysis; you decide what to do with it.

  5. Make it one click for your team

    Save this ask as a custom command on the assistant your team already uses, so anyone can run it in one step.

Machen Sie es sich zu eigen

Adjusted Comparisons

Controls for level, tenure, and location so you see real gaps, not artifacts of who's senior.

Threshold Flags

You set the line (3%, 5%, whatever your policy is) and Joy flags every level that crosses it.

Chart In The Answer

The gap by level lands as a chart right in the chat, with the numbers repeated in the written read.

Drill By Segment

Follow up to break a flagged gap down by manager, team, or hire cohort to find the source.

Multi-Dimension Review

Look at gaps by race/ethnicity or by full-time versus part-time, not just gender.

By Business Unit

Compare adjusted gaps across departments to see where the exposure concentrates.

Pre-Cycle Check

Run it before merit planning so managers get budget guidance where gaps exist.

Post-Adjustment Recheck

After raises land in your data, ask again to confirm the flagged gaps actually closed.

Frequently Asked Questions

How does an AI pay equity audit control for legitimate differences?

Joy adjusts for level, tenure, and location before it compares, so a gap driven by seniority mix doesn't masquerade as a pay-equity problem. What's left after those controls is the adjusted gap you actually need to explain.

Can I set my own threshold for what gets flagged?

Yes. Tell Joy the line that matters for your policy (3%, 5%, or whatever you use) and it flags every level that crosses it and shows how far each one is over.

Does this write pay changes back into our system?

No. It's a read-only analysis. Joy reads your compensation data at the moment you ask and returns a chart and written read. Any raise or adjustment happens in your own comp process; you copy the findings into your planning.

Is the compensation data exposed to other employees?

Access follows the permissions you already set, so only people cleared for compensation data can run the audit. The analysis is a point-in-time answer in your chat, not a shared standing dashboard.

Can I look at more than gender?

Yes. You can ask for gaps by race/ethnicity, by employment type, or by other fields you track, and cut them within a job family or across business units to see where exposure concentrates.

Verwandte Vorlagen

Personalwesen

Headcount Dashboard

Ask for headcount by department and see it charted on demand.
Personalwesen

Turnover Heat Map

See where turnover is highest by team, tenure, or manager.
Personalwesen

Team Sentiment Analysis

Spot burnout trends by reading the questions teams ask most.

Ready to find your pay gaps in an afternoon?

Melden Sie sich für die Warteliste an und probieren Sie diesen Workflow bei der Einführung von JoySuite als Erste aus.

Zur Warteliste Alle Vorlagen durchsuchen