The Old Model Does Not Meet the New Standard
The traditional equity model in most large enterprises has three moving parts. Once a year, HR runs a pay equity study. Once a year, DEI publishes a representation report. Once a year, the CHRO presents a slide to the board. The problem is that hiring decisions are made every day, promotion decisions are made every quarter, and pay decisions are made every cycle. A once-a-year read on equity cannot influence a decision that has already been made a hundred times before the next study runs.
The regulatory regime is now catching up with that reality. The EU Pay Transparency Directive, adopted by the EU Commission in 2023 as Directive 2023/970, sets a June 2026 transposition deadline and requires employers to report gender pay gaps, act on unexplained gaps above the 5% threshold through a joint pay assessment, and give candidates access to pay information before interview. In the United States, the EEOC continues to expand its pay data collection program, and the OFCCP has raised the compliance bar for federal contractors. In the United Kingdom, the ONS gender pay gap reporting rules are being extended in practice, with growing pressure from the EHRC and from investors for ethnicity pay gap disclosure alongside gender.
None of these regulations is asking for an annual slide. They are asking for a data system that produces defensible answers on demand. That is what people analytics has to become.
Equity analytics is moving out of the annual report and into the operating rhythm of the HR function, sitting alongside the hiring, pay, and promotion decisions in real time rather than reviewing them after the fact. The organisations that will meet the new standard are the ones treating equity as a measurable operating discipline, not a compliance event.
The Four Layers Equity Analytics Actually Has to Cover
Talenbrium's people analytics framework organises equity measurement into four connected layers. Each layer answers a different question, each is built on different data, and each fails in a different way if it is measured alone. Reading the four layers together is what turns a set of dashboards into an equity discipline.
Layer 1. Pay Equity
The most measurable and the most regulated of the four. Pay equity analytics asks whether two employees doing work of equal value are paid equally, controlling for the factors that legitimately explain pay difference such as location, experience, and performance. The EU Pay Transparency Directive names this test directly. Eurostat's structure of earnings survey is the reference point for the EU-level pay gap benchmark, and BLS median weekly earnings data serves the same role for the US market.
Where most organisations get this wrong is in the internal-only view. A pay equity analysis that reads only against the internal pay structure will show equity even when the entire structure is set below external market for a particular demographic cohort. Talenbrium's external compensation model closes that gap by placing every internal pay band against the external market read for the same role, geography, and seniority, so pay equity can be tested both within the organisation and against the market the organisation hires from.
Layer 2. Promotion and Progression Velocity
Pay equity is the outcome. Promotion equity is where the outcome is produced. Talenbrium's analysis of promotion data across tracked employers shows a consistent pattern where the pay gap at the exit of a career stage is smaller than the promotion velocity gap that produced it. Two employees paid equally in the same role can be on very different trajectories to the next role, and the current pay equity view will not see it.
Promotion velocity analytics measures how long it takes an employee in a given demographic to move from one seniority band to the next, controlling for role, performance, and tenure. The gaps that appear in this data are often much larger than the gaps that appear in the pay equity view, because they compound across the career rather than showing up in a single pay cycle.
Layer 3. Hiring Funnel Diagnostics
The equity problem often starts before the offer letter. Hiring funnel analytics asks where the funnel narrows unevenly by demographic, from sourced candidate to applied, to phone screen, to onsite, to offer, to accept. A funnel that looks broadly representative at the top and homogeneous at the bottom is a funnel that is failing an equity test at a specific stage, and the specific stage is knowable from the data.
Talenbrium's hiring funnel analytics reads this stage-by-stage, benchmarked against the external labor market composition for the role. If the applied pool is representative of the labor market but the offer pool is not, the equity failure is between application and offer, not upstream in sourcing. That distinction changes the intervention.
Layer 4. Development Access and Stretch Assignments
The most invisible of the four layers, and the one most predictive of long-term equity outcomes. Development access analytics asks who gets the stretch assignments, the visible projects, the cross-functional exposure, and the sponsorship that produces the next promotion. This data usually does not sit in a system of record. It sits in project rosters, committee membership, and manager-level decisions that are not typically audited for equity.
Talenbrium's employer database and workforce pulse data help construct a proxy view of this layer, because access to development activity is a leading indicator of promotion velocity three to five years out. Reading Layer 4 today gives the CHRO a view of what Layer 2 is going to look like in the medium term.
The Four-Layer Analytics Model, Visualised
"A pay equity study run once a year cannot change a hiring decision made every day. Equity analytics that is not built into the operating rhythm of the function is measurement without discipline, and measurement without discipline does not close a gap."
Talenbrium Workforce Intelligence, Q2 2026Where Analytics Reinforces Bias Instead of Removing It
The counterintuitive finding from Talenbrium's people analytics work is that badly designed analytics can make equity outcomes worse, not better. A model trained on historical hiring, pay, or promotion data is a model that has learned the historical bias in that data. Deploying such a model into the hiring or pay decision loop does not remove bias, it operationalises and accelerates it. Reading which analytics practices build equity, and which ones entrench the gap they were designed to close, is where the practical work of the CHRO sits.
Analytics Practices That Build EquityAdopt
- External-benchmarked pay equityReads internal pay structure against the external labor market for the same role, geography, and seniority, so internal equity is tested against the market and not just against itself.
- Stage-by-stage funnel decompositionIsolates the specific hiring stage at which representation drops, so the intervention lands where the gap is created rather than being applied uniformly across the funnel.
- Promotion velocity by cohortMeasures time-to-next-band by demographic, role, and manager, exposing the promotion gaps that a pay gap view cannot see because they are still building.
- Manager-level equity signalsReads promotion, pay, and attrition patterns by hiring manager, identifying where inequitable decisions concentrate rather than treating the whole organisation as one average.
- Access to stretch assignmentsTracks who is on the visible projects and in the leadership development pool as a leading indicator of the next promotion cycle's outcome.
Analytics Practices That Entrench BiasRetire
- Predictive hiring models trained on historical hiresThe historical hire pool is the outcome of past bias, so the model learns to replicate that bias at scale. Removes friction from an unfair process rather than fixing it.
- Attrition risk models used for retention prioritisationRetention effort concentrated on high-flight-risk employees systematically underinvests in the cohorts less likely to leave, deepening the equity gap the model was supposed to help close.
- Sentiment analytics without action loopsMeasures employee experience but never feeds the finding back into a decision, producing an ever-larger dataset of unresolved equity signals.
- Once-a-year pay equity audits with no external benchmarkPasses an internal test and fails the external market test. Does not survive the joint pay assessment standard the EU Pay Transparency Directive is introducing.
Where the Regulatory Clock Is Loudest by Region
The equity conversation is regulated differently in different markets, and the analytics build has to respond to the regulator that will read the output. Talenbrium's regional view highlights where the pressure is arriving first, and what the analytics function is being asked to produce.
| Region | Primary Regulatory Driver | Analytics Requirement Rising Fastest | Severity |
|---|---|---|---|
| European Union | Pay Transparency Directive 2023/970, transposition June 2026 | Joint pay assessment for gaps above 5%, external market benchmarking | Critical |
| United Kingdom | ONS gender pay gap reporting, EHRC oversight, ethnicity reporting pressure | Ethnicity pay gap analytics alongside gender, quartile representation | High |
| United States | EEOC pay data collection, OFCCP federal contractor oversight, state-level pay transparency laws | State-by-state pay range disclosure compliance, EEO-1 component readiness | High |
| Ireland | Gender Pay Gap Information Act reporting extended to 150+ employees | Ready-to-publish quartile analysis, bonus and benefit-in-kind gap | High |
| Germany | Entgelttransparenzgesetz, Directive 2023/970 alignment via national transposition | Works-council-facing pay structure documentation, joint pay assessment readiness | Moderate |
| Spain and Netherlands | National equal pay registers, EU Directive alignment | Standardised pay category documentation, external comparability | Moderate |
What Good Looks Like: Five Operating Principles
Talenbrium's people analytics work with HR functions across sectors and regions produces a consistent set of principles for organisations that are turning people analytics into an equity discipline rather than a reporting exercise.
Under the EU Pay Transparency Directive, an unexplained gender pay gap above the 5% threshold triggers a joint pay assessment obligation. Talenbrium's analysis suggests that a material share of large employers will cross the threshold on their first published report simply because their internal analytics have never been tested against an external comparability standard. The workforce planning window to build the analytics capability to meet that standard is now, not after the first report goes out.
The Talenbrium View
Equity analytics is not a technology project. It is not a dashboard project. It is a decision system project, and the goal is that every hiring, pay, promotion, and development decision is made with equity data visible at the point of decision, benchmarked against an external market read, and captured for audit. The organisations that reach that standard will not need a separate equity audit, because the audit will be a summary of the decision system rather than a review of it after the fact.
People analytics does not build an equitable workplace by producing a better annual report. It builds one by changing what data is in front of the decision-maker at the moment the decision is being made. That is the operating change the new regulation is asking for, and it is the change the best HR functions are already building.
Want to test your people analytics against the new standard?
Talenbrium's people analytics team helps HR functions read their pay, promotion, hiring, and development data against external market benchmarks, and stress-test their equity analytics against the EU Pay Transparency Directive, EEOC oversight, and ONS reporting standard. A short conversation is enough to see whether a deeper review is worthwhile.
