From a food-waste pledge to a reliable operating routine

In food operations, loss arises at a small number of recurring points, and each has a different remedy.

Production planning sets how much is made. Where planning runs ahead of demand — building stock against an uncertain forecast — some of that stock will not sell within its life. The remedy is forecasting and planning discipline rather than any downstream intervention.

Purchasing sets how much raw material is bought. Order quantities driven by supplier discount tiers, minimum order quantities or delivery schedules rather than by production need create surplus that then has to be used or lost.

Storage determines how much survives. Temperature control, rotation and stock visibility determine whether material is used while it is still usable. Losses here are frequently the result of material being found late rather than of it deteriorating quickly.

Service forecasting determines how much prepared food is produced against uncertain demand — particularly relevant in food service, where production is often made ahead of service. Where demand is uncertain, the choice of how much to prepare in advance is a direct decision about expected surplus.

Surplus routing determines what happens to what is left. Whether surplus can be redistributed, sold through an alternative channel or used as animal feed or recovery determines whether it is a loss or a recovery. The safety and regulatory conditions applying to redistribution are matters for qualified advice, and they differ by jurisdiction.

The weekly cycle fails in a predictable way: it works while someone is watching and decays when attention moves elsewhere. The counter is to make each step someone's named responsibility, with a figure they are accountable for.

Ownership has to be real rather than nominal. A person who is accountable for loss at the storage stage needs the authority to change rotation practice, and needs their performance to be assessed on the outcome. Where accountability exists without authority, the owner can report the problem but not fix it.

The review step is the one that most often lacks an owner, and it is the step that produces learning. A review that compares loss by stage and reason week to week identifies drift early and confirms whether an intervention worked. Without it, the cycle records data without generating decisions.

Measurement in food operations competes with operational work, and a measurement burden that is too heavy is abandoned. The design principle is proportionate accuracy: measure precisely what you will act on, and estimate the rest with the method recorded.

For most businesses, that means precise measurement of the larger streams — the categories where volume or cost concentrates — and a lighter treatment of categories that are small or stable. The estimation log is what keeps this honest, distinguishing measured from estimated values so that a figure is not treated as more precise than it is.

It also keeps the cycle sustainable. A routine that takes twenty minutes a week will persist; one that requires an hour will not survive a busy period.

A useful test when designing the routine is to ask what decision each figure will inform. A number that no one would act on differently, whichever way it moved, is a number that need not be collected weekly. Pruning those out is what keeps the remaining measurements credible and maintained, because attention is finite and spreading it thinly across everything is how a measurement routine fails.