Why the monthly count fails in a hypermarket — and how to fix it
A monthly count does not fail because the rep is careless. It fails for four operational reasons that repeat in every large store — and none of them are fixed by adding people.
If your monthly count returns a different number every time, the problem is usually not the person counting. In large stores, four operational causes corrupt the number before it ever reaches your spreadsheet — and none are solved by adding staff or tightening instructions.
1. One item, four places
In a hypermarket the same product sits on the main shelf, in a box at the checkout, on an off-shelf display in the aisle, and in a carton in the back. Count only the shelf and the number is short. Count shelf and warehouse onto one line and you can never split it apart again.
The difference between two counts then stops meaning sales movement and starts meaning that the first rep counted the back room and the second forgot it. If you do not know where the stock was, you cannot know whether it was missing.
2. The count lands at the worst possible time
Month end is also target close, peak footfall, and fifteen minutes before the store shuts. The rep is standing at the shelf with a phone in one hand and a carton in the other, and is expected to record accurately.
Any process that needs five steps to log one item will be shortened under those conditions. Nobody will tell you it was shortened. The numbers will.
- Measure the logging step in seconds, not in features: how long does one item take?
- Every extra field must drive a decision you actually make. If it does not, remove it.
- Barcode scanning is not a luxury — it is the difference between three seconds and fifteen.
3. The warehouse has no signal
This is the most commonly missed point when buying a counting system. The stock room is at the back or below ground, and coverage there is weak or absent. Any tool that needs a live connection will stop at the most important part of the visit.
What happens in practice is that the rep writes on paper and keys it in later from somewhere with signal — restoring the manual error you bought the system to remove.
4. Expiry recorded later is expiry never recorded
An expiry date is only legible with the item in hand. If it is not captured at that moment, it will never be captured accurately. By the time the expiry list arrives a week later, the window for return or rotation has narrowed or closed.
Waste here is not an accident. It is the predictable result of delaying information that was free at the moment of counting.
What to ask for as a manager
- Counts separated by location inside the store, merged afterwards rather than before.
- Expiry captured with the item as it is counted, not in a later report.
- Full offline operation, demonstrated in a warehouse rather than promised in a deck.
- A report closed in store before the rep leaves, so it is never edited from memory.
- A variance log per store and per visit, so you can see where the problem repeats instead of generalising.
None of these are technical requirements. They are operational conditions. Any competent technical team can meet them once it understands the aisle — and the usual problem is that it has never stood in one.