Inventory shrinkage is the loss of stock between purchase and sale, caused by theft, damage, miscounting, or administrative error.
Inventory shrinkage is the loss of stock between when it is bought and when it is sold or used - the gap between what your records say you hold and what is actually on the shelf. The defining feature of shrinkage is that it is discovered, not observed: nobody logs it as it happens, it simply surfaces when a cycle count or year-end stocktake finds less than the books promised.
What causes shrinkage
Shrinkage is a bucket for several different problems, and they need different fixes:
- Theft - external (shoplifting, van break-ins) and internal (stock walking out with staff). Small, valuable, resellable items are the classic targets.
- Damage and spoilage - dropped, crushed, expired, or weather-ruined stock that was thrown away without being written off, so the records still show it.
- Administrative error - miskeyed quantities, a delivery booked in twice, a unit-of-measure mix-up (a box of 12 received as 12 boxes), transfers between locations recorded on one side only.
- Supplier short-shipments - the invoice says 100, the pallet held 96, and nobody counted at the door. You paid for stock that never arrived.
- Counting error - some “shrinkage” is phantom: stock that exists but sits in the wrong bin or was counted wrongly. It distorts the numbers in both directions.
How to calculate the shrinkage rate
Shrinkage rate = (recorded inventory − physical inventory) ÷ recorded inventory × 100
A worked example: the system says a storeroom holds €50,000 of stock at cost. A full count values what is actually there at €48,500. Shrinkage is €1,500, and the rate is 1,500 ÷ 50,000 × 100 = 3%. Tracked per period and per location, the rate tells you whether things are improving and where the losses concentrate - one site running well above its siblings is a lead, not a coincidence.
Why shrinkage hurts twice
The first cost is the stock itself - bought, paid for, gone. The second is quieter: until the count, your records overstate what you hold. Purchasing trusts those records, so replenishment is skipped for stock that does not exist, and the result is a surprise stockout and customers pushed onto backorder for an item the system swore was available. Shrinkage is not just a write-off at year end; it is bad data feeding every decision in between.
How to reduce shrinkage
The fixes map to the causes:
- Count little and often. Rolling cycle counts catch discrepancies while they are small and recent enough to investigate, instead of letting a year of losses arrive as one unexplainable number.
- Check goods in at the door. Receiving against the purchase order, not the invoice, kills short-shipment losses.
- Record waste as waste. Make writing off a damaged item easier than binning it silently.
- Give stock a named owner. Losses shrink when a specific person answers for a specific storeroom, van, or count.
- Tighten who can edit records, and keep a trail of changes - many “losses” turn out to be edits. For teams that track equipment and consumables in one register, AMPthilly logs every checkout, return, transfer, and field edit in a per-item audit history, which makes tracing where a discrepancy crept in a lookup rather than an interrogation.
Reusable carriers deserve a special mention: pallets and totes leak constantly precisely because everyone treats them as nobody’s stock.
Related terms
- Cycle Count - rolling partial counts, the main tool for catching shrinkage early
- Stockout - the downstream failure when shrinkage leaves records overstated
- Backorder - accepted orders waiting on stock the records wrongly promised
- Inventory Turnover - how fast stock moves; slow-moving stock hides shrinkage longest
- BOM (Bill of Materials) - the component list that makes usage predictable enough to spot leaks