Corvus Retail Group: One Global Grading Structure Versus Four Labour Markets

Global (do this): job families, level definitions, the number of grades, and a single job-evaluation methodology applied consistently. The test is portability of meaning: a grade should describe comparable scope and accountability in every country, so that a 'Grade 5 engineer' can be discussed in a talent review across borders.

Local (must stay local): pay ranges themselves, built from local market data. Absolute pay cannot be equalized across the UK, Netherlands, Mexico and India - market rates, statutory layers and living costs differ, and equalizing would make Corvus uncompetitive in India and overpay in Mexico.

Where the global answer must stop - explicitly:

• The statutory and collective layers sit outside the base range. Mexico's profit sharing and thirteenth-month payment, the Dutch 8% holiday allowance and CAO terms are not 'range positioning' - modelling them inside the range produces a false comparator.
• Cross-country comparison must use compa-ratio and range penetration, plus total-reward percentile positioning, never currency amounts.
• The CAO covering 60% of Dutch staff limits unilateral range design; consultation is a constraint, not a detail.

On 12 months: the framework (families, levels, methodology, grade count) is achievable in 12 months. Rebuilding local ranges in four markets with one analyst per country and no global lead is not. Commit to the framework in 12 months and local range implementation over 24, and say so to the board - an unmet 12-month promise on pay data is far more damaging than a phased plan.

Recognize that the existing reports do not answer the question. UK gender pay gap reporting is a disclosure obligation (a set of prescribed calculations on a snapshot date), not a diagnosis - it tells you the gap, not why. The investor is asking about equity, which is a different and broader question.

Run four analyses, in this order:

1. Equal pay for equal work: like-for-like comparison within grade and job family, country by country. This is the legal exposure question and must be answered first.
2. Representation by grade: the reported gap is usually driven mainly by segregation - few women in the top band - not by unequal pay for the same job. If that is the case here, a pay-line fix will not move the number, and the board needs to hear it.
3. Mechanics audit: starting-salary rules, offer-negotiation practice, whether salary history is requested, and how people are placed within ranges. These are where small gaps are manufactured and where inexpensive fixes exist.
4. Transitions: promotion and attrition rates by gender at each grade boundary - a leak at the manager-to-senior-manager step compounds into a large top-band gap over years.

What not to promise: do not commit to 'closing the gap' by a date. Commit to a defined metric, a baseline, a remediation budget and a review cadence. Also flag that the legal frame differs by country - UK equal pay is a statutory claim route; EU pay-transparency rules will push toward disclosed ranges and gap reporting; India and Mexico rest on equal-remuneration provisions - so the same analysis can carry different consequences in each market.

Sequencing note: fix the mechanics (offer guidelines, range-placement rules, promotion calibration) before spending on remediation, or the gap reopens.

The trap first: matching counteroffers systematically resets your internal equity baseline. Each accepted counteroffer becomes the reference point for the next resignation, so the price of retention ratchets upward and you end up rewarding external offers rather than performance. A 30–40% uplift also tells you your ranges are stale - people are leaving for numbers your structure does not contain. That is a data point about the ranges, not just about the individuals.

Do this instead:

1. Segment the attrition. 24% overall may be concentrated in the first two years of tenure. If it is early-tenure, the driver is hiring profile and onboarding, and pay is not the lever - no amount of range adjustment fixes a bad hire profile.
2. Separate regrettable from non-regrettable. Reward spend should follow the roles that are genuinely scarce.
3. Re-benchmark and take a deliberate position on critical roles - for example median to 75th percentile for a defined set of scarce technical roles, funded by declining to chase every counteroffer.
4. Set a counteroffer policy: only for critical or scarce skills, capped, approved centrally, with a repayment clause if the employee leaves within 12 months.
5. Attack the non-monetary drivers. Top-band attrition at this level is frequently about career path and scope, not cash. A visible next-role path is cheaper than a 35% uplift and does not reset anyone else's expectations.

Link to the grading project: the new architecture is the opportunity to fix this structurally - if the framework is delivered without re-benchmarking India, you will have built a clean structure on top of stale data.