In retail, head office cost is not a head office problem

Support functions get cut one at a time, against a benchmark percentage of sales. The cost they actually create sits in the store manager’s week, and nobody owns that.

The benchmark measures the wrong thing

Head office cost is governed as a percentage of sales, compared against a peer set. It is an easy number to steer and it says nothing about which work should exist. A function can hit its target by removing the people who did the work while keeping the work itself, which simply moves it to the stores.

Every function is cheap on its own

Each support function sends work to the shop floor. A stock count, a compliance check, a campaign set-up, a new field in the rota tool, a survey, a photo of the display. Inside the function that request costs nothing, because it is one email. In aggregate it is the reason a store manager spends a third of the week on head office work and the rest dealing with what goes wrong.

No single function is accountable for how much of a store manager’s week head office uses, so each of them adds to it.

That makes it a decision rights problem before it is a cost problem. There is no single owner of what may be asked of a store, no budget for store time, and no forum where two functions have to trade against each other. Every function uses store time, but nobody hands it out or keeps track of it.

Size the functions against the demand they create

Give store time an owner and a ceiling. Make the functions bid against that ceiling the way they bid for budget, and publish what each one is currently taking. Then you can see which support work to keep: the requests that are worth the store hours they take.

The changes to the structure are usually smaller than people expect: Fewer coordination roles sitting between the functions and the estate, one clear owner for store-facing demand, and a decision right that did not exist before.

What this is not

Shared services and new systems change where the work is done, but not how much head office may ask of a store. A ceiling on store time keeps working when stores open or close, which in retail happens every year.

Sizing a support function is an OD Blueprint. If a plan for it is already on your desk, an OD Verdict is the faster way to find out whether it holds.