Stock Take for a Store Chain: How Often to Count, How to Reconcile and What Shrinkage Is

A stock take (stocktaking, an inventory count) is the physical count of the goods in a store or warehouse, compared line by line with the stock the system says should be there. The system figure is the book stock; the counted figure is the physical stock. The difference between the two is the variance, and the part of it nobody can explain is shrinkage.

For a chain, the count does two jobs. It corrects the records that every replenishment order is calculated from, and it gives the accounts a closing stock figure and a loss for the period. The two jobs pull towards different methods, which is why most chains need more than one kind of count. This guide covers how often to count and what, how to reconcile book and physical stock, and how shrinkage is defined and measured. It sits under our wider guide to stock control for retail chains.

1. Why the record drifts from the shelf

Every sale, delivery, transfer and write-off moves the book stock by a number someone or something recorded. Each wrong scan, unbooked delivery or unrecorded breakage leaves the record a little further from the shelf, and nothing in daily trading pulls it back.

The scale is larger than most managers expect. A 2026 study of a leading UK grocer compared system and counted stock across 139,828 product and store records in 11 stores: 35.3% were accurate, 39.4% showed more stock in the system than on the shelf and 25.3% showed less¹. The retailer itself called these counts stocktakes.

The direction matters more than the count. When the system shows more than the shelf, replenishment waits for stock that is not there and the shelf stays empty: the out-of-stock guide shows how to spot those lines from sales data. The same study ran a store-wide count in one store and used a matched store as a control. Sales in the counted store rose by about 11% over the following two months, and the lift came from lines where the system had shown more than the shelf; lines with accurate or understated records showed no discernible change¹. A count, in other words, is not only an accounting chore. It puts products back on the shelf.

2. Full counts and cycle counts

There are two ways to organise counting, and they answer different questions.

A full count is the one auditors and finance teams know, and the only one that measures the loss for the whole store at a single date. Its weakness is the gap between counts: a record that went wrong in February drives orders until the next count in January. Cycle counting closes that gap by spreading the work, so that each line is counted on a schedule that fits its risk rather than the calendar.

3. Which products to count more often

Not every line deserves the same attention. In the UK grocery study, records were more often wrong for lines with higher average stock, more frequent deliveries and a short shelf life, and less often wrong for lines with more days on promotion, which the authors link to the extra checks a promotion brings¹. Perishable lines also gained more from being counted than the rest.

• Lines with large stock and frequent deliveries: more handling, more chances to book the wrong quantity.
• Fresh and short-life lines, where waste is often thrown away without being written off.
• Lines thieves target. The British Retail Consortium names alcohol, cosmetics, meat, baby formula and medicines as the goods organised offenders focus on².
• High-value lines, where one unit of error is worth a whole shelf of cheap ones.
• Lines whose record says stock but whose sales have stopped. These are the ones the count pays for fastest, and they sit close to dead stock, which is why the two need to be told apart before anything is marked down.

4. Running the count

• Set a cut-off: post every delivery, transfer and return before the count, and keep anything that arrives during it apart.
• Count blind. Counters should not see the book figure, or they will find it.
• Count in pairs or have a second person check a sample, and recount any line whose variance is large in value.
• Count the stockroom, the shop floor, displays and returns cages separately, then add them up.
• Whether the counters are your own staff or an outside stock taker, agree in advance how units, cases and weighed goods are recorded.

5. Reconciling book and physical stock

The count produces a variance for every line. The mistake is to post it straight to the accounts as loss. A large part of most variances is the record being wrong for reasons that have nothing to do with theft, and those reasons repeat if nobody finds them. Work through the causes in order, cheapest first.

Transfers deserve particular care in a chain: a transfer booked out of one store and never booked into the other shows up as a loss in the first and a gain in the second, and the two cancel only if someone looks at both. Positive variances, where the shelf holds more than the record, are a clue in their own right: they usually mean the stock was sold or received under another code. Our guide to investigating an inventory shortage shows the same narrowing applied to one large shortage across seventeen sites.

6. What shrinkage is and how it is measured

Shrinkage (shrink) is stock a retailer paid for and could not sell: stolen by customers or staff, damaged, spoiled, or lost to errors in receiving, pricing and recording. Part of it is known when it happens, such as a breakage written off or a theft caught on camera. The rest, unknown loss, appears only when a stock take finds less than the book says.

Shrinkage is usually expressed as a percentage of sales: the loss for the period between two counts divided by net sales over the same period. A store whose book stock at cost was £412,000 and whose count found £405,800 has lost £6,200; with £520,000 of sales in the period, that is a shrinkage of 1.2%. State the basis. A loss valued at selling price looks larger than the same loss at cost, so compare stores, periods and benchmarks only on the same one.

Published benchmarks are thin and mostly old. The US National Retail Security Survey of 2023 put the average shrink rate at 1.6% of sales for financial year 2022, with internal and external theft accounting for 65% of it³. In the UK, the British Retail Consortium's report for the year to August 2025 counts 5,451,000 incidents of customer theft costing £408m, but that edition switched to counting only verified incidents, dropping unknown loss, and states that the true cost is likely to be much higher². Police in England and Wales recorded 519,516 shoplifting offences in the year to June 2026, and the ONS notes that reporting rates for shoplifting are very low⁴.

That is why your own counts are the better benchmark: the same store against its own history, and each store against similar stores in the chain. Read the shrinkage figure alongside stock turnover, since stock that sits longer has more time to go missing, and feed corrected records into the safety stock you hold, so the buffer is not sized to cover errors the count could have fixed.