Inventory Shrinkage
Inventory shrinkage is the gap between what a warehouse's records say it holds and what a physical count actually finds — and the discipline of shrinkage management is the counting, investigation, and adjustment workflow that turns an unexplained gap into a documented, actioned cause.

On this page · 6 sections
How does Inventory Shrinkage apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What shrinkage is: the book-to-count gap
Every warehouse and manufacturing site keeps two numbers for each stock-keeping unit: a recorded quantity — the figure sitting in the WMS, ERP, or inventory ledger after every receipt, pick, move, and consumption transaction has been posted — and a physical quantity, which is what someone actually finds when they count the bin, pallet, or location. Shrinkage is the name given to a loss identified once that book-to-count difference has been investigated and supported as a real reduction in inventory, rather than left as an unexplained number on a report.
That distinction between a raw count difference and supported shrinkage matters more than it might first appear. A perpetual inventory record — one that updates with every purchase, sale, transfer, and adjustment as it happens — gives a business a book figure it can compare against a controlled count at any time. But the comparison on its own only produces a difference; it does not explain where that difference came from. Before a business can responsibly classify a gap as shrinkage, it needs to rule out or correct cutoff timing (a receipt or shipment booked in the wrong period), ownership issues (consigned or customer-owned stock mixed into the count), unit-of-measure errors, duplicate or omitted transactions, and genuine counting mistakes — because relabeling any of those as shrinkage hides the real problem instead of fixing it.
This also explains why periodic inventory systems — where stock is only counted and reconciled at set intervals such as month-end, quarter-end, or annually — struggle to isolate shrinkage at all. Under a periodic approach, cost of goods sold is calculated as beginning inventory plus purchases less ending inventory, so any unit that goes missing between counts simply disappears into that residual cost-of-goods-sold figure rather than surfacing as a separately visible shrinkage number. A perpetual system, because it carries a running quantity between counts, is what makes the gap visible in the first place — which is one of the reasons manufacturers and warehouse operators with meaningful inventory volumes tend to move toward perpetual tracking rather than relying solely on periodic stocktakes.
02Common causes of shrinkage
Shrinkage is rarely a single phenomenon — it is a catch-all label applied to several distinct failure modes that happen to produce the same symptom (a count lower than the book). Separating out the likely cause before writing off a variance is what makes an investigation useful rather than cosmetic.
- Counting and data-entry errors — a miscount during receiving, picking, or a cycle count, a transposed digit, or a transaction posted against the wrong SKU or location.
- Damage — product broken, crushed, contaminated, or otherwise rendered unsellable or unusable during handling, storage, or transport, and not formally written off at the time it happened.
- Theft — internal (employee-related) or external, ranging from opportunistic pilferage to coordinated diversion.
- Unrecorded consumption or usage — material pulled for production, samples, rework, or internal use without a corresponding issue transaction being posted, so the book quantity never reflects what actually left the shelf.
- Unit-of-measure errors — receiving or issuing in cases when the system expects eaches (or vice versa), or a conversion factor between a supplier's pack size and the stocking unit that was set up incorrectly.
- Evaporation, spoilage, and natural loss — for liquids, bulk powders, and perishable goods, a small and often expected volume or weight loss from drying, settling, or degradation over time, which is why some operations track an allowance for this separately from loss they treat as abnormal.
- Cutoff and ownership mismatches — a shipment or receipt recorded in the wrong period, or consigned, customer-owned, or in-transit stock that was never excluded from the count population.
The retail-focused 2023 Retail Security Survey from the National Retail Federation and the Loss Prevention Research Council, while framed around store-level loss rather than warehouse operations, illustrates how varied the causes behind a single shrink figure can be: participating retailers attributed their shrink to external theft (an average of 36% of total loss), internal theft (29%), and the remainder split across process and control failures, other causes, and unattributed loss — with the survey defining the broader “process, control failures and errors” category explicitly as losses from normal business operations such as cashier errors, inaccurate system pricing, accidental loss or damage, and other operational or systemic errors. 1 The same report put the average shrink rate reported by participating retailers at 1.6% of sales in FY2022, up from 1.4% the prior year, representing an estimated $112.1 billion in reported losses across the sector that year — figures useful as a point of reference for the scale shrinkage can reach, even though the survey's population, sector mix, and measurement conventions (it is explicit that most respondents exclude ecommerce loss, for example) are specific to US retail rather than manufacturing or wholesale warehousing. 2
03How cycle counting and full stock takes surface the gap
The two main mechanisms a warehouse or manufacturing site uses to compare its book record against physical reality are cycle counting and the full (or periodic) physical inventory, sometimes called a stocktake.
ASCM's inventory-control materials describe a periodic inventory approach as counting an organization's full physical inventory at a recurring interval — often monthly, quarterly, or annually — noting that this approach tends to suit organizations with smaller inventory volumes because a full count requires people to count items manually, which is a heavy, disruptive exercise at scale. A perpetual system, by contrast, uses continuous, automated tracking — purchases, sales, transfers, and other movements update the record as they occur, typically via barcode scanning, RFID, or point-of-sale and warehouse-management integrations — so the book quantity is always current without waiting for the next full count. 3
Cycle counting is the practice that lets a perpetual-inventory site validate its records without the disruption of a full stocktake. ASCM's own dictionary defines cycle counting as an inventory accuracy audit technique where inventory is counted on a cyclic schedule rather than once a year, usually on a regular, defined basis — more frequently for high-value or fast-moving items and less frequently for low-value or slow-moving ones — with the most effective programmes counting a certain number of items every working day, each item at a prescribed frequency. 4 Crucially, the dictionary definition frames the purpose of cycle counting as being to identify items in error, which then triggers research, identification, and elimination of the cause of the error — in other words, cycle counting exists to drive root-cause investigation, not just to produce a corrected number.
Which items get counted, and how often, is usually set by an ABC classification: ASCM describes ABC analysis as splitting a population of items into three groups based on sales or usage volume, with the A group — typically the top 10–20% of items by popularity, representing 50–70% of sales or usage volume — counted more often and controlled more tightly, the B group receiving a middle tier of attention, and the C group, covering the bulk of SKUs by count but a much smaller share of value, governed by looser, less frequent controls. 5 Applied to shrinkage, this means a warehouse's highest-value or fastest-moving lines get counted — and any variance investigated — far more frequently than slow-moving, low-value stock, so that the items where an undetected gap would do the most financial damage are the ones checked most often.
A full physical stock take still has a role even where cycle counting is well established. It gives a single, whole-site snapshot that cycle counting's rolling, item-by-item schedule does not — useful for year-end financial reporting, for catching errors that a narrow cycle-count sample might miss, or for resetting the book record after a system migration or a period of known disruption. Many operations run both: cycle counting as the continuous accuracy-maintenance routine, and a periodic full count as a less frequent, comprehensive check.
04Shrinkage KPIs and tolerance thresholds
The most common way shrinkage is expressed as a KPI is as a percentage: the value of the identified loss divided by the value of inventory, cost of goods sold, or sales over the same period, so that sites of different sizes can be compared on a like-for-like basis. The NRF/LPRC survey's own convention illustrates this: it reports average shrink as a percentage of sales and is explicit that this figure can vary significantly by sector — participants in grocery and supermarket reported average shrink rates over 2%, while specialty apparel participants reported a different range, underlining that a single universal "acceptable" shrinkage percentage does not exist across industries. 6
Alongside the headline shrinkage percentage, operations typically track a related but distinct measure: inventory record accuracy, usually expressed as the percentage of counted locations or SKUs that match the book record within an agreed tolerance. Cycle counting programmes are commonly built around a target accuracy level and a defined tolerance band for what counts as an acceptable variance versus one that must be investigated — a few units or a small percentage difference on a high-volume, low-value line might be tolerated, while the same absolute variance on a high-value or regulated item triggers a mandatory investigation regardless of its percentage size.
Because shrinkage KPIs are lagging indicators — they report what has already gone missing — many operations pair them with leading indicators drawn from the cycle-count programme itself: count completion rate against the scheduled cadence, the proportion of counts triggering a variance above tolerance, and the average time between a variance being flagged and the investigation being closed. A shrinkage percentage that looks acceptable but sits alongside a chronically late or incomplete cycle-count programme is a weaker signal than the same percentage backed by a programme that is actually running to schedule.
05Investigation and adjustment workflows, with an audit trail
The step between a raw count variance and a posted shrinkage adjustment is where the discipline actually lives. A workable investigation workflow typically moves through several checks before anything is written off: first, confirming the count itself was done correctly (wrong location, wrong unit of measure, or a missed sub-location can all produce a false variance); second, checking for cutoff and timing issues (a receipt or shipment recorded in the wrong period relative to when the count was taken); third, checking for ownership or consignment stock that should not have been included in the counted population; and fourth, checking for recording errors such as a duplicate receipt, an omitted return, or a transaction posted against the wrong SKU.
Only once those explanations have been ruled out, and a genuine loss is supported by the investigation, should the variance be posted as shrinkage — at which point the standard accounting treatment is to reduce the inventory account for the assigned cost of the missing units and record the corresponding expense. Understanding Accounting's glossary entry on the topic is explicit that an identified recording error should be corrected through its own account rather than being relabeled as shrinkage, precisely so the shrinkage figure reflects real loss rather than becoming a dumping ground for unreconciled administrative mistakes. 7
An audit trail matters at every stage of that workflow, not just at the point of adjustment. A defensible record typically captures: who performed the count and when; what was counted and at what location; the book quantity at the time of the count; the variance identified; what investigation steps were taken and what they found; who approved the resulting adjustment; and the date the adjustment was posted. This record serves several audiences at once — internal audit and finance need it to support the numbers in the accounts, operations needs it to spot recurring patterns (the same location, the same SKU, the same shift showing up repeatedly), and, in regulated sectors handling controlled substances, high-value components, or allergen-sensitive materials, an external auditor or regulator may expect to see it as part of a wider inventory-control review.
06How perpetual inventory systems reduce shrinkage
A perpetual inventory system does not eliminate shrinkage on its own — it is a recordkeeping approach, not a security control — but it changes how quickly and precisely a gap can be found and acted on. Investopedia's explainer on perpetual inventory systems notes that while these systems continuously record inventory changes in real time, they still require periodic physical counts to reconcile discrepancies from loss, theft, or damage, which the system itself cannot account for — the perpetual record shows what should be there; a physical count is still what confirms what actually is. 8
What the continuous, transaction-level record does provide is a far narrower window for an investigation to work with. Where a periodic system might only reveal that something went missing sometime in the last quarter, a perpetual system that logs every receipt, pick, transfer, and adjustment lets an investigator narrow a variance down to a specific location and a specific window of transactions, which both speeds up root-cause analysis and makes it far harder for a pattern of small, repeated losses to go unnoticed for long. Barcode and RFID scanning, in particular, reduce one specific cause of shrinkage directly — the manual data-entry and unit-of-measure errors that occur when staff key in quantities by hand rather than scanning a known pack or license-plate identifier.
Understanding Accounting's treatment of perpetual inventory makes a related point worth carrying into any shrinkage programme: a perpetual system is not self-validating. The reconciliation between the book record and the count-and-ownership ledger has to happen, and differences have to be investigated before they are posted, because calling every unexplained book-to-count difference "shrinkage" hides cutoff errors, consignment mistakes, unit-of-measure problems, duplicate receipts, and unrecorded movements rather than fixing them. 9 In other words, perpetual tracking and cycle counting work together, not as substitutes for each other: the perpetual record supplies the book figure that makes a gap visible sooner and more precisely, and cycle counting (plus periodic full stocktakes) supplies the independent physical evidence that confirms, or corrects, that figure.
Frequently asked questions
Q.Is every book-to-count difference the same as shrinkage?+
No. A book-to-count difference is only the starting point. Before it is classified as shrinkage, it needs to be checked for cutoff timing errors, ownership or consignment mismatches, unit-of-measure problems, and recording errors such as duplicate or omitted transactions. Shrinkage is the loss that remains once those other explanations have been ruled out or corrected through their own accounts.
Q.What's the difference between cycle counting and a full physical stock take?+
Cycle counting checks a rolling subset of items on a defined schedule — typically weighted so high-value or fast-moving items are counted more often than low-value, slow-moving ones — so the whole inventory is validated gradually without stopping operations. A full physical stock take counts everything at once, usually at a less frequent interval, and gives a single whole-site snapshot often used for year-end reporting or after a major disruption.
Q.Does a perpetual inventory system eliminate the need for physical counts?+
No. A perpetual system continuously updates the book record as transactions occur, but it still relies on periodic physical counts to catch and reconcile losses from theft, damage, or other causes that the transaction record itself cannot see. The two work together rather than one replacing the other.
Q.What causes unit-of-measure shrinkage?+
This happens when a receipt or issue transaction is recorded in the wrong unit relative to how the system tracks stock — for example receiving by the case when the system expects eaches, or using an incorrect pack-size conversion factor — which creates an apparent gain or loss that has nothing to do with physical loss of product.
Q.Is there an industry-standard acceptable shrinkage percentage?+
No single figure applies across industries. Reported shrink rates vary by sector — for example, the NRF/LPRC 2023 Retail Security Survey found participants in grocery and supermarket reporting average shrink rates above 2% of sales, with other sectors reporting different ranges — so tolerance thresholds are generally set by each organisation based on its own item values, margins, and risk profile rather than taken from an external benchmark.
Primary sources
- NRF / Loss Prevention Research Council — 2023 Retail Security Survey
- ASCM — Types of Inventory Control Systems (periodic vs perpetual)
- ASCM Insights — The XYZ's of Inventory Management (ABC classification)
- APICS San Antonio — Cycle Counting by the Probabilities (citing the APICS Dictionary definition of cycle counting)
- Investopedia — Perpetual Inventory System Explained
- Understanding Accounting — Inventory shrinkage
- Understanding Accounting — Perpetual inventory system
Further reading
- ABC Cycle CountHow ABC classification sets cycle-count frequency by item value and velocity.
- Cycle CountThe rolling count technique used to catch shrinkage between full stocktakes.
- Audit TrailThe record-keeping backbone behind a defensible shrinkage investigation.
- Perpetual InventoryThe continuous-tracking approach that makes book-to-count gaps visible faster.
- Inventory AccuracyThe related KPI that tracks how often counts match the book record.
Want to see how Inventory Shrinkage could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
