Diversity recruitment: measuring what actually matters
Most diversity recruitment metrics measure activity, not progress. Counting how many diverse candidates applied, or how many events you ran, tells you whether you were busy. It does not tell you whether your hiring changed. The teams that actually move the dial measure a small number of things well, at the points in the process where bias does its work, and they resist the temptation to celebrate the numbers that are easiest to produce.
This is a guide to measuring diversity recruitment in a way that means something, for anyone who hires in the UK and wants evidence rather than a dashboard that looks good in a board pack. It pairs with our broader piece on building balanced teams that last.
Measure the funnel, not just the top and bottom
The single most useful thing you can do is track representation at each stage of hiring, not only applications and hires. Applied, screened, interviewed, offered, accepted. When you see the proportion of women, or any under-represented group, at every stage, you find out exactly where they drop out.
This matters because the fix is different at each stage. If women apply in healthy numbers but fall away at CV screening, the problem is how CVs are read, and our piece on reducing bias in your hiring process addresses that directly. If they reach interview and then do not get offers, the problem is the interview itself or the panel. If they get offers and decline, the problem is your proposition or your reputation. A single top-line diversity number hides all of this. Stage-by-stage data points you at the actual leak.
The metrics that actually reveal something
A few measures carry real signal. Prioritise these over vanity numbers.
Pass-through rates by stage and group. The percentage of each group that moves from one stage to the next. A sharp drop for one group at one stage is the clearest evidence of where to intervene.
Offer-acceptance rates by group. If under-represented candidates accept offers at a lower rate, something in the experience, the pay, or the perceived culture is putting them off after they have met you.
Quality-of-hire and retention by group, over time. The honest long-run test. Diverse hiring that is followed by diverse people leaving within a year is not working, and only time-based data shows it. This is where many programmes quietly fail.
Source effectiveness by group. Which channels actually bring you candidates who progress and get hired, rather than channels that simply generate applications. This stops you pouring budget into activity that produces volume and no outcomes.
The metrics that mislead
Some popular numbers do more harm than good, because they reward the wrong behaviour.
Raw application counts are the worst offender. A flood of applications from under-represented groups looks like success and means nothing if none of them get hired. It can even mask a process that filters those candidates out later. Counting activity, the number of job fairs attended or posts published, measures effort, not result, and effort is not the point.
Single aggregate diversity percentages, reported without the funnel behind them, let you miss the stage where the problem lives. And any metric that can be hit by hiring one visible senior person, while the pipeline beneath stays unchanged, invites exactly that kind of gaming. Measure the system, not the symbol.
Tie it to the pay gap
Diversity recruitment and the gender pay gap are the same problem viewed from two ends. Hiring and promoting women into higher-paid and senior roles is what closes a pay gap over time, so your recruitment metrics and your pay gap figures should tell a consistent story. If your hiring data shows progress but your published gap stays flat, you are probably recruiting women into junior roles and losing them before they rise. Our complete UK guide to gender pay gap reporting sets out how the published figures work and why representation across the pay quartiles is what moves them.
The Office for National Statistics anchor is worth keeping in view: a median gender pay gap of 6.9% for full-time employees and 12.8% across all employees in April 2025. That gap is driven mostly by who holds senior and higher-paid roles, which is precisely what good diversity recruitment, measured properly, is meant to change.
Set up measurement you can trust
A few principles keep the data honest. Collect diversity information fairly and lawfully, which in the UK means handling it in line with equality and data protection law, keeping it voluntary and confidential, and being clear about why you are collecting it. Baseline before you start, so you can show change rather than guess at it. Review the funnel regularly rather than once a year, because a quarterly rhythm lets you act while a hire is still live. And report the uncomfortable numbers alongside the good ones, because a measurement culture that only surfaces wins is not measuring anything useful.
Above all, treat the metrics as a diagnostic, not a scoreboard. The point is not to produce a number that satisfies a board. It is to find the stage where good candidates are being lost, and fix it. Our practical guide to inclusive recruitment covers the process changes that the data will point you towards.
Frequently asked questions
What are the best metrics for diversity recruitment?
Stage-by-stage pass-through rates by group, offer-acceptance rates by group, and retention and quality-of-hire over time. These show where candidates are lost and whether diverse hiring actually sticks, which raw application counts and activity measures cannot.
Why are application numbers a poor measure of diversity?
Because they measure interest, not outcome. A large, diverse applicant pool means nothing if those candidates are filtered out at screening or interview, and a high application count can even disguise a process that rejects them later. Track what happens after they apply.
How does diversity recruitment relate to the gender pay gap?
They are two views of the same issue. A pay gap closes when women move into higher-paid and senior roles, so recruitment data and pay gap figures should align. If hiring looks diverse but the published gap stays flat, you are likely recruiting women into junior roles and losing them before they progress.
How often should we review diversity recruitment data?
Quarterly is a sensible rhythm for the funnel, so you can act while roles are still open, with a fuller annual review against your baseline and your published pay gap figures. Reviewing only once a year means problems run for months before anyone sees them.
Hiring women? Get RecruitHer early access for stronger signals and less inbox noise.
This is educational information, not legal advice. Collecting and using diversity data carries legal obligations under UK equality and data protection law. For impartial guidance contact ACAS or consult gov.uk.
Last reviewed: November 2026