Pay bands across regions: explaining the same role at different rates

Two analysts in the same role were paid thirty per cent apart, and a shared document made the gap visible in one afternoon. The gap was defensible and the secrecy was not.

The document that showed the gap

A hiring plan circulated with salary ranges attached, and two analysts noticed their own roles in it. One was in Manchester, the other in Lisbon, same title and same band label.

The numbers differed by thirty per cent. Neither number was a mistake, and neither person had known the other’s until that Tuesday.

The Manchester analyst asked one question in the thread: is my work worth more, or is my postcode? Nobody answered for two days.

Location pricing is normal

Paying by market is what every company with offices has always done, and remote work did not invent it. It became visible, which is a different problem.

The logic is real, and it is not a cover story. The same salary buys different amounts of life in different places, and one global number breaks in both directions.

The failure was never the policy itself. It was running a policy that shapes people’s lives without ever stating, in writing, how it works.

Secrecy is the part that costs

A policy nobody can inspect gets filled in by imagination, and imagination is usually unkind. People assume the difference tracks favouritism rather than a table.

In our case the assumption was wrong and the damage was already done. Two people spent a weekend deciding whether they had been undervalued for years.

Publishing the method costs one difficult conversation and a slightly awkward document. Not publishing it costs a version of that conversation in every one-to-one, indefinitely.

Publish the method, not the salaries

We now publish how a band is set: a reference location, a multiplier per region, and the review date. Individual salaries stay private.

The method answers the question people actually ask, which is whether their number was arrived at fairly. The number alone never answers that.

It also constrains us. Once the method is written, an exception has to be argued against a document rather than decided in a corridor.

Anchor bands to a reference city

Every band starts from one named location and a multiplier table. Ours starts from the capital and treats every other region as a percentage of it.

The anchor is arbitrary and we say so. What matters is that it is stable, written down, and applied to every role in the same way.

Stability is what makes the system usable in a negotiation. A band that moves with the market every quarter cannot be quoted to a candidate.

The cost of living trap

Cost of living indices look precise and are mostly noise at the scale of a salary. Rent differs enormously inside one city, let alone one country.

We use broad regional bands instead of index maths. A three-tier table is cruder, easier to explain, and far harder to game.

The trap is pretending to precision. A candidate who finds a cheaper flat does not expect a pay cut, and the system should never imply one.

Explain what the band buys

A band is not a statement about the value of a person. It is a statement about what the company will pay for a role in a market.

That distinction sounds like corporate language until you use it in a real conversation. It separates the job from the person doing it.

It also gives the manager something honest to say when the answer is no. The role is priced in this market; your performance is discussed on its own terms.

The person above the band

Long-serving people drift above the band as the market moves. We froze the band during a flat year and three salaries quietly left it behind.

We now treat anyone above the band as a red-marked case with a written plan. Freezing their pay without a plan is how loyalty becomes resentment.

The plan is usually a one-off adjustment or a role change, decided once a year. What it cannot be is silence, which is what we chose first.

The person below it

A new hire below the band’s midpoint is normal. A person below the band’s minimum after two years is a decision somebody has avoided making.

We review those cases every February, before the salary round rather than after it. The order matters, because the round is where excuses get made.

Most were explained by a starting salary set in a hurry. Correcting them cost less than the two resignations that followed the previous review.

When someone moves country

A move to a cheaper region triggers a band change, and we now say so in the contract rather than at the airport. Nobody has been surprised since.

We phase the adjustment over two cycles instead of applying it at once. A sudden cut after a move reads as a penalty for a personal decision.

The reverse move is applied immediately, which is unfair in the short term and obviously correct in the long one. We say that out loud too.

What to say when you cannot match

Sometimes a candidate has an offer from a company with a single global band. We do not match it, and we say why in one paragraph.

The paragraph names the band, the region, and the review date. It does not claim the other offer is wrong, which candidates can always detect.

About one in five accepts anyway. The ones who do tend to stay, because they chose the role with the number in front of them.

The annual review of the bands

Once a year we check the multiplier table against what we actually paid to hire. Two regions were badly out of date and had been for eighteen months.

The correction cost money and bought back some credibility. The people in those two regions had been quietly assuming the table was fiction anyway.

We now publish the review date alongside the method. A rule with a known revision date is easier to trust than a rule with none.

Most pay policies are defensible in the abstract and awkward in a specific conversation. If your team read the method tomorrow, what would you change before that felt safe?