Shortlists that never changed shape, and the four requirements that explained why.
A Midlands engineering group could not shift the composition of its senior shortlists. Full-market research and a specification review identified the artificial barriers doing the excluding.
Sector: Engineering. Market: West Midlands. Services used: Diversity hiring, Talent mapping.
A Midlands engineering group could not shift the composition of its senior shortlists. Full-market research and a specification review identified the artificial barriers doing the excluding.
The situation
A West Midlands engineering group had held a representation target for three years without measurable movement in its senior technical shortlists. Internal consensus had settled on the explanation that the available engineering population simply did not support the target. Nobody had tested that claim.
What we did
Two parallel workstreams. First, full-market identification across the West Midlands and the wider engineering belt, deliberately reaching beyond the client's usual competitor set into aerospace, energy, rail and industrial automation. Second, a written review of the role specification, testing every stated requirement against whether it was genuinely necessary. We also benchmarked what the available market actually looked like, so the target could be calibrated against reality rather than against a national statistic.
What the research found
Four findings that changed the decision.
Four requirements were narrowing reach without justification
A continuous-service requirement, a named-employer training-scheme preference, an inflated years-of-experience threshold and a degree requirement that predated the current role all filtered on career shape rather than on capability.
The referral channel was reproducing the existing team
A large share of prior hires had come through internal referral, which is efficient and, left unbalanced, structurally self-replicating.
The available market differed from the national picture
Representation in the specific engineering specialisms concerned was materially different from national workforce statistics — in some sub-functions better, in others worse. The target had been set against the wrong reference point entirely.
The funnel leaked at a specific, measurable stage
Candidates were being lost at a consistent point in the process, which pointed to an assessment issue rather than a reach issue.
The outcome
The four artificial requirements were removed, which expanded the qualified population substantially for all candidates. Referral weighting was rebalanced against directly researched candidates. The representation target was recalibrated against the actual available market. Shortlist composition changed over the following two hiring cycles, and the benchmarking evidence was subsequently cited in the client's board and ESG reporting.
The point of it
The work here was lawful positive action — widening the pool and removing artificial barriers. At no point were candidates screened on protected characteristics, and no quota was operated. The requirements that were removed had been excluding capable candidates of every description.
A note on confidentiality. Clients are never named without written consent, and we do not publish project values, candidate records or commercially sensitive figures. These accounts describe real engagement patterns with identifying detail removed. Where a specific number is not stated here, it is not a number we have published — please do not infer one.
How this engagement was structured
The commercial and methodological shape of the project, for readers assessing whether a similar approach would suit their own situation.
| Service | Diversity hiring research |
|---|---|
| Scope | Full-market research and representation benchmarking for a senior appointment where previous shortlists had lacked diversity. |
| Commercial model | Fixed project fee in pounds, agreed against a written scope before work began. No commission and no percentage of salary. |
| Data ownership | Every record produced transferred to the client permanently at the close of the project. |
| Lawful basis | Legitimate interest under UK GDPR and the Data Protection Act 2018, documented for the project with an agreed retention period. |
Method
What we actually did
The decisions that determined whether this project produced something usable.
Comparator population defined first
The realistic market for the role, level, sector and region — not the national workforce. The wrong denominator makes every subsequent figure meaningless.
Whole market researched, not the applying fraction
Because a shortlist drawn from applicants and referrals reproduces the demographics of those channels, which is what those channels do.
Specification reviewed for unnecessary barriers
Requirements that filtered disproportionately without predicting performance were identified and challenged — frequently the highest-impact part of the work.
Representation reported at every stage
Mapped population, longlist, shortlist and appointment, so the stage at which representation dropped was visible rather than assumed.
What this engagement demonstrates
The constraint was the specification rather than the market. Several requirements were removing a large share of the available diverse population without predicting job performance at all.
A note on how we present client work
Client identities and commercially sensitive details are withheld unless we hold explicit written permission to name them. Where figures are quoted they come from the engagement itself. Where an example is constructed to illustrate a method rather than describe a specific client, we label it as illustrative rather than presenting it as a case study. We would rather publish a less impressive page than an unverifiable one.
Questions we are asked about this work
Can we speak to a reference client?
In many cases yes, subject to the client’s agreement. Some engagements are confidential and will stay that way — which is itself a reasonable indication of how we would treat yours.
Would this approach work for our situation?
That is exactly what a scoping call establishes. Thirty minutes is usually enough to tell you whether the method transfers to your market and what the right scope and fee would be — including telling you if it does not transfer.
How much would a comparable project cost?
It is quoted as a fixed fee against a written scope, driven by market breadth, seniority and the depth of validation required. Because scope drives the fee, a meaningful figure needs a short conversation about what you actually need.
Recognise the problem?
Most of these engagements began with a client describing a symptom. Thirty minutes on a call is usually enough to work out what is actually causing it.