Mercer diagnoses.PayGapCheck proves it.

That sentence is the position of this product. It is short on purpose; the full version takes longer to write than the diagnostic itself takes to deliver.

Mercer charges €50k a year for the kind of data the diagnosis is built on. The diagnosis tells you what the market looks like. It does not tell you what your company looks like. Those are different statements.

For the 200-person HR Director with a June 2026 deadline, the difference is everything.

What Mercer actually is

Mercer's compensation surveys are the largest of their kind. They aggregate declarations from thousands of companies' HR teams, normalise the inputs into role-and-level distributions, and publish annually with a 6-month lag from data collection.

This is a useful product for the right buyer. The 1,000-person multinational with a dedicated compensation team uses the survey the way the methodology supports — sector benchmarking, annual band calibration, board reporting on relative positioning. The €50k is a rounding line on a multi-million-euro HR stack.

What Mercer does is diagnose. The diagnosis is precise on its own terms: "here is where your role bands sit relative to the market, on the published basis, at the moment the survey closed."

It is also six months old, sector-aggregated, and exactly the same data delivered to every other subscribing company. Three properties that matter less for the multinational and more for the 200-person SME.

What proof actually requires

The directive does not ask what the sector looks like. It asks what the employer's company looks like — on its own data, by category of equal work or work of equal value, with documented methodology.

That is a different artefact.

The sector distribution is the reference; the company's data is the subject. The HR Director presenting the report cannot point to Mercer and say "here is what we look like." The report shows what this company looks like. The sector survey is, at most, a footnote.

Proof has four properties Mercer cannot deliver:

The first two are properties of any clean payroll export. The second two are properties of the methodology built around it. Together they are what the report needs.

The next hire

Here is where the sentence gets sharp.

Every offer extended creates new patterns in the data. The candidate hired today at €68k for a Senior role, when the existing team's Seniors are clustered at €58k–€64k, just rewrote three things:

If the offer was constructed from a Mercer median, the offer is calibrated to the sector. It is not calibrated to this company's current state. The diagnosis told you what the market looks like; it did not tell you what your category looks like; you extended an offer based on the diagnosis; the offer just reshaped what your category looks like.

The next pay gap report includes that data. The methodology behind it does not.

This is the loop. Diagnosis informs the offer; the offer reshapes the company; the company reshapes the report. The diagnosis never sees what it set in motion.

When diagnosis is enough

This article is not arguing Mercer is bad. The 1,000-person multinational has a compensation team that translates the survey into operational decisions, a budget that absorbs €50k as a rounding line, and a board that recognises the brand reference as sufficient. For that buyer, Mercer is exactly the right product. Diagnosis is what they need. They can produce their own proof from internal systems.

The argument is about who Mercer is not the right product for. The 200-person SME with one HR generalist, no dedicated compensation team, and a June 2026 deadline cannot translate diagnosis into proof on its own. The €50k pays for the input; the output costs another €50k of HR generalist time, or another €30k of consultancy, or — most often — gets translated badly and produces a report that does not survive the joint pay assessment.

There is a segment-shaped hole between "diagnosis" and "proof" that has been there for a decade. The directive made it acute. The hole is what PayGapCheck is built for.

Stop diagnosing. Start proving.

PayGapCheck operates on the company's own data. Salary file in. The report — overall and per-category gaps, quartile distributions, every Article 9 metric, with the methodology documented per calculation — out. The reference period is the actual reference year. The categories are the company's actual categories. The output is what the company actually has to publish.

It does not replace the diagnostic. The sector reference is still useful as context — "we are at the 60th percentile" is a meaningful statement for the board conversation. But the report is built on internal data, not external aggregation.

For the next offer, HireGapCheck™ shows where the proposed offer sits against the existing cohort, with the documented rationale categories the directive's framework respects. Not "what does the market pay" but what does our category currently look like, and what does this offer do to it.

For the position view before either, ReadinessCheck™ surfaces where the company sits on the directive's principal obligations — without requiring any salary data. The diagnostic step that costs €50k everywhere else costs about 20 minutes here.

The shift from diagnosis to proof is the shift from "what does the market look like" to "what does our company look like." From "what is the average bonus gap" to "what is our bonus gap, per category, this year." From "we benchmark against Mercer" to "here is the calculation, the methodology, and the documentation."

The first question has been answered for decades. The second is what the directive is actually asking.