Beyond Diversity Metrics, A 6-Step Racial Equity Audit Framework

A Practical Framework for What Your Data Is Not Showing You

Most organisations believe they do not have enough data to understand what is happening to their Black employees. Most organisations are wrong. The data exists. It has not been read in the right terms. This article provides a practical framework for the racial equity audit that most HR functions have never run.

In our study of 1,039 Black professionals, we found that aggregate data acts as a veil. As a researcher, I’ve seen that ‘90% employee satisfaction’ often hides a 0% promotion rate for Black talent. This audit is the ‘structural survey’ your DEI strategy is missing.

The Data You Already Have

A racial equity audit does not require building new data infrastructure from scratch. It requires reading existing data through a lens that most organisations have never applied.

Every organisation running annual performance reviews holds years of rating data. Every organisation with a promotion process holds records of who was put forward, who was successful, and at what level. Every organisation with a pay structure holds salary data that can be disaggregated. Every organisation that conducts exit interviews holds qualitative data about why people leave.

What most organisations do not do is analyse this data by race with sufficient granularity to see the structural patterns beneath the aggregate. Aggregate diversity data obscures the specific, cumulative, and compounding inequities identified by the Excellence Tax research. The audit begins by disaggregating what you already hold.

Step 1: Disaggregate Your Performance Review Data

Performance review data is the starting point for a racial equity audit because it sits at the intersection of every other mechanism the Excellence Tax operates through. It captures the outcome of tone policing, the result of sponsorship gaps, the consequence of the double performance standard, and the language through which racialised judgements become official organisational record.

Pull the last three years of performance review ratings. Disaggregate by race. Cross-reference with the seniority level of the reviewer and the reviewed. Calculate the average rating differential between Black employees and their non-Black peers at equivalent grades, tenure, and function.

In the majority of organisations, this analysis will produce a gap. Black employees will show lower average ratings than non-Black employees at equivalent levels, despite evidence that their objective output is comparable. If you do not find this gap, your data collection may not be capturing race accurately enough to produce a meaningful analysis.

Next, analyse the ratings against objective output metrics. Project delivery records, sales figures, client satisfaction scores, or whatever measurable performance data your organisation holds. A rating gap that is not explained by objective output differential is evidence of a process problem.

Finally, analyse the correlation between performance ratings and subsequent promotion decisions. If lower ratings for Black employees are translating into lower promotion rates, and those lower ratings are not explained by objective output, you have documented the mechanism through which performance review bias produces advancement inequity.

The question to bring to your leadership team: if the ratings are producing racialised outcomes that are not explained by performance, what is the ratings process actually measuring?

Step 2: Audit Your Performance Review Language

Rating scores tell part of the story. The language used in written reviews tells a different and equally important part.

Pull a representative sample of written performance reviews from the last two to three years. Remove identifying information. Analyse the language used to describe performance, development needs, and leadership potential across racial groups.

Look specifically for the following patterns:

Descriptors applied to Black employees that describe personality or style rather than performance: “abrasive”, “difficult”, “intense”, “emotional”, “hard to manage”, “not quite the right fit”, “needs to work on presence”, “can come across as quite direct”. Cross-reference whether the same descriptors appear in reviews of non-Black employees describing equivalent behaviours.

Vague developmental feedback that cannot be acted on: “would benefit from reflecting on how they are perceived”, “needs to build stronger relationships with stakeholders”, “should work on executive presence”. Assess whether this type of feedback appears more frequently in reviews of Black employees than non-Black employees, and whether it is accompanied by the specific examples that would make it actionable.

Conditional language that undermines positive assessments: “performs well when given clear direction”, “shows potential but needs to develop confidence”, “good work this year, though there is room to grow into the role more fully”. Analyse whether this qualifying pattern appears more frequently in positive reviews of Black employees than in equivalent reviews of non-Black employees.

The language audit is qualitative work. It requires time and a degree of analytical judgment. It also produces some of the most compelling evidence available for the leadership conversation about structural change, because it makes the mechanism visible in your organisation’s own documents.

Step 3: Map Your Promotion Patterns

Promotion data sits at the outcome end of the Excellence Tax mechanisms. By the time a Black professional fails to receive a promotion they deserved, multiple structural mechanisms have already operated: tone policing has shaped how they are perceived, performance review bias has shaped their record, and the sponsorship gap has shaped whether anyone advocated for them in the calibration meeting.

Map your promotion data over the last five years. For each promotion decision, record the grade of the person promoted, their tenure at that grade, their most recent performance rating, their function, their manager, and their race.

Calculate the promotion rate for Black employees versus non-Black employees at equivalent tenure and performance ratings. A gap at this level, after controlling for performance rating, is strong evidence of a structural barrier operating between performance and advancement.

Then examine the calibration meeting process. Who presents candidates for promotion. Whether there is a standard for the evidence required to support a promotion recommendation. Whether the same quality of evidence is required for all candidates regardless of who is advocating for them. Whether the informal advocacy of senior sponsors, the act of going into a room and putting someone’s name forward, is systematically less available to Black professionals than to their non-Black peers.

The sponsorship gap is one of the most consequential structural mechanisms the Excellence Tax research identifies, and one of the least formally addressed. Most organisations have no process for ensuring Black professionals have access to senior advocates. They have mentoring programmes. Mentoring and sponsorship are not the same thing. A mentor offers guidance. A sponsor goes into rooms and advocates. The gap between the two is where careers stall.

Step 4: Analyse Your Attrition Data

Exit interview data is rarely analysed with sufficient racial granularity to reveal what it actually contains. Many organisations do not ask about race in exit interviews. Many that do collect the data do not analyse it by race. And many that analyse it by race do not cross-reference it with the tenure, seniority, and performance profile of the people leaving.

Run the following analysis. Of the Black employees who left your organisation in the last three years, what was their average tenure, their most recent performance rating, and their seniority level at exit. Compare with equivalent figures for non-Black employees who left in the same period.

If Black employees are leaving at earlier tenure, at higher performance ratings, and at lower seniority levels than their non-Black peers, your organisation is losing its highest-performing Black talent before it reaches the levels where it would produce the most value. This pattern is what the Excellence Tax produces. It is visible in attrition data when you look for it.

Supplement the quantitative analysis with qualitative data. Where exit interview responses are available, analyse the language Black employees use to describe their reasons for leaving. Look for the themes the Excellence Tax research identifies: feeling unseen, unrecognised, held to a different standard, unable to progress despite strong performance. These themes, appearing consistently across exit interviews over multiple years, constitute organisational evidence of the mechanism.

Step 5: Run a Pay Gap Analysis by Race

Gender pay gap reporting is mandatory for UK organisations above a certain size. Race pay gap reporting is not currently mandatory, though the direction of policy travel suggests it will become so. Forward-thinking organisations run this analysis now.

Disaggregate your pay data by race, controlling for grade, function, and tenure. A pay gap that is not explained by these variables is evidence of a structural inequity in how Black professionals are being compensated relative to their non-Black peers.

Cross-reference the pay analysis with the promotion and performance review analysis. Pay gaps that are partly explained by lower promotion rates for Black professionals at equivalent performance levels are not straightforwardly pay discrimination. They are evidence of a structural mechanism in the promotion process that produces a pay outcome as a downstream effect. Addressing the pay gap requires addressing the promotion mechanism, not just the pay structure.

Step 5: Collect Qualitative Data Directly

Quantitative data tells you what is happening. Qualitative data tells you how it is experienced and what is producing it. A complete racial equity audit requires both.

Create the conditions for Black employees to share their experiences accurately and without fear of professional consequence. This requires anonymised data collection, ideally through an independent third party with the expertise to ask the right questions and the credibility to handle sensitive material safely. Internal surveys administered by HR functions rarely produce the candour that the qualitative dimension of a racial equity audit requires.

The questions that produce the most useful qualitative data ask about specific experiences rather than general satisfaction. Not “do you feel included?” but “have you ever received feedback focused on how you communicated something rather than on what you communicated?” Not “do you feel valued?” but “have you ever felt that a promotion decision did not reflect your performance level?” The specific questions produce specific answers that map onto the mechanisms the audit is trying to identify.

What to Do With What You Find

The racial equity audit produces data. The data requires a response. Two principles guide what comes next.

Publish the findings internally. Transparency about what the audit revealed, presented to the whole organisation with leadership accountability attached, signals a seriousness of intent that private acknowledgement does not. It also creates a baseline against which future progress can be measured.

Connect findings to specific structural changes with specific owners and timelines. An audit finding that performance review language is racialised requires a specific intervention: a language audit protocol, a training framework for reviewers, a structured second review process for developmental feedback. An audit finding that the promotion rate gap is not explained by performance requires a specific intervention: a structured sponsorship programme, a revised calibration process, defined evidence standards for promotion recommendations. The finding and the intervention need to be explicitly connected.

The Cost of Black Excellence Research Institute works with organisations to design and conduct racial equity audits, interpret the findings through the Excellence Tax framework, and build the structural change programme that addresses what the data reveals.