Shrinkage is measured per aisle by counting, and ours runs at two tenths of a percent everywhere except aisle 9, which runs at eight tenths. The cameras see nobody, the gates never sound, and the product going is cat food.
I am a stock controller at a supermarket. Nine years, and the counting is my job.
How shrinkage is actually measured, because it is more exact than people think.
Every line has a perpetual inventory figure: what the system believes is on the shelf, from deliveries in, sales out and wastage recorded. Then you count. A counter walks the aisle with a handheld and counts the facings, and the difference between the count and the system is the variance.
Done weekly on high risk lines and monthly on everything else, and the variance rolls up per aisle and per category. Our store runs at about two tenths of a percent of sales, which is good and which I am quietly proud of.
Aisle 9 runs at eight tenths.
Aisle 9 is pet food. Four bays of it, dry and wet, and the loss is all in wet cat food in single pouches and small tins.
About forty pounds a week at retail. Two thousand a year. In a store that turns over a great deal more than that, which is why I have never been able to get anybody excited about it.
What is ruled out, and the loss prevention team have done most of it because this is exactly their job.
Delivery error. Every delivery is checked in against a note and the discrepancies are raised with the depot within twenty four hours. Aisle 9's deliveries reconcile.
Wastage not recorded. Out of date stock gets scanned into a wastage terminal, which is auditable, and the bins are checked.
The tills. Mis-scanning would show as a sales pattern and not a count variance, and the self checkout exception data for pet food is unremarkable.
The cameras. There is a dome at each end of aisle 9 and loss prevention have reviewed them in eleven separate blocks, some of them entire weekends at speed.
They have never seen anybody take more than they paid for out of that aisle.
The gates, which is the part I cannot get round.
Every pouch and tin in the high loss lines in aisle 9 has been source tagged since 2021. That was my idea and it cost real money: a security tag inside the outer film on every multipack and a hard tag strategy on the singles.
A tagged item going through the gates at the front sets them off. That is not a subtle system, it is a radio aerial in a door frame, and it is tested every morning with a test tag by the duty manager and the test is signed.
The gates work. They go off four or five times a day, mostly on somebody's old jacket with a tag left in it from another shop, and twice a week on something real.
They have never gone off on cat food.
Since source tagging the loss in aisle 9 has not changed by a single percentage point.
So it is leaving the store without passing through the gates, or it is not leaving the store.
I spent 2022 on the second one, because it is the cheaper answer and because stock that has not left a building is somewhere in the building.
Every shelf, every void above the shelves, behind the bays, in the warehouse, the compactor area, the staff room, the lockers with a union rep present, the yard, and the skip. A full store count over two nights with eleven people.
We found forty one pounds of stock in the wrong places, which is normal, and none of it was cat food.
What I checked in 2023, which is the thing I would most like somebody to explain.
The counts.
A variance is a difference between a count and a system figure, so it can be a counting error as easily as a loss. So I stopped trusting the handhelds and did aisle 9 myself, with a colleague, independently, on the same evening, twice a week for six weeks.
Our two counts agreed with each other every time, to the unit.
And the loss is still there, which means it is real stock that was on the shelf at nine in the evening and is not on the shelf at nine in the morning, in a shut shop, with no door open between midnight and six except the goods in door which is on camera and manned.
What happens about it.
It is in the weekly report as a known high shrink aisle, and the regional stock controller has looked at my figures twice and told me both times that my methodology is sound and that some aisles are like that.
Some aisles are like that. It is the most irritating true sentence in retail.
What I do.
Aisle 9 is counted weekly, source tagged, faced up last thing, and the figures go in the report.
The only thing I have changed is that I do not count it at night on my own.
It used to be my quiet job. The store is shut, the aisle is yours, and counting four bays of pet food takes about twenty minutes with nobody asking you where the bleach is.
In September 2023 I was halfway down it at about one in the morning with the handheld, working the bottom shelf of the wet food, which means you are crouched with your back to the aisle.
The noise a pouch of cat food makes when it is moved is a specific noise and I hear it three hundred times a week.
I heard it behind me, four or five times, at about the height of the second shelf, and when I stood up and turned round the facings on the second shelf were straight, because I had faced them up myself an hour earlier, and in the middle of the run there was a gap exactly one pouch wide.