Why costs rise even when every function meets its targets

The review looks healthy. Buying has held margin. Logistics has brought down cost per case. Stores have stayed inside the labour plan. Planning has a forecast accuracy it can defend. Then finance asks why the total cost of serving the customer went up. The answer is usually sitting between the targets, not inside them.

A buyer protects margin by taking a larger pack or a longer lead time. The warehouse absorbs the pack because its target is throughput, not whether the store can sell the pack before it expires. The store protects labour by cutting a replenishment task. Planning protects forecast accuracy by smoothing a promotion the commercial team has already promised. Each choice is rational against its own scorecard. Together they create expedites, markdowns, transfer costs and stock that never meets a customer.

This is why a chain can be full of high-performing teams and still feel expensive to run. The measures were designed when each function could be improved on its own. Retail no longer works that way. Availability, cash and cost are properties of the whole path from supplier to shelf. A target that cannot see the next function will keep winning locally and losing collectively.

The repair is not a thicker dashboard of the same KPIs. It is a small set of shared outcomes that no single function can hit alone: on-shelf availability, inventory that turns, and the real cost of a fulfilled order, including the expedite and the markdown. Functions still need their own measures. Those measures have to be checked against the shared ones, or the monthly review will keep congratulating a system that is getting dearer.

Leaders usually discover this in the gap between the pack they bought and the shelf the customer sees. If every team met its target and the shelf is still wrong, the targets are describing effort. They are not describing the ecosystem.