Here is a number that should make any retailer pause: U.S. retail shrinkage reached an estimated $112 billion in a single recent year — an $18 billion jump over the prior year. That is not a rounding error. It is the combined cost of theft, paperwork mistakes, spoilage, and miscounts quietly draining margin out of stores and warehouses everywhere.
The frustrating part is that most of this loss never shows up as a line item anyone watches. It hides inside cost of goods sold, gets discovered once a year during a chaotic full physical count, and then gets written off in one painful entry. By then it is far too late to learn anything from it.
This guide walks through what inventory shrinkage actually is, how to measure it with a shrink rate you can track over time, why cycle counting beats the annual freeze-and-count ritual, and how to record the adjustment cleanly so the loss is visible instead of buried.
What Inventory Shrinkage Actually Is
Inventory shrinkage is the gap between the inventory your records say you have and the inventory you physically have on the shelf or in the rack. If your system shows 500 units of a SKU and a physical count finds 485, you have 15 units of shrinkage. Those units were paid for, they sit on your balance sheet as an asset, but they cannot be sold to generate revenue. They are simply gone.
Shrinkage is not a single problem. It is a bundle of separate problems that happen to produce the same symptom. The major sources break down roughly like this:
- External theft. Shoplifting and organized retail crime are the single largest contributor, accounting for well over a third of total shrink at many retailers.
- Employee theft. Internal theft — staff taking merchandise, cash, or manipulating returns — typically runs close behind external theft, often around a quarter to a third of the total.
- Administrative and paperwork errors. Wrong receiving counts, pricing mistakes, untracked transfers between locations, and data entry slips. This is loss with no villain, just process gaps.
- Vendor and supplier shortages. You were billed for 100 cases and received 96, and nobody caught it at the dock.
- Damage and spoilage. Broken, expired, or unsellable goods that never got formally written off.
- Self-checkout and returns fraud. Newer loss channels that have grown sharply as stores automate checkout and liberalize return policies.
The reason this matters operationally is that each source has a different fix. Theft calls for security and access controls. Administrative error calls for better receiving and counting discipline. Spoilage calls for tighter rotation and ordering. If you only know your total shrink, you cannot tell which lever to pull. Measurement has to come before prevention.
How to Calculate Your Shrinkage Rate
A raw dollar figure for missing inventory is useful, but it is hard to compare across periods or locations. The shrinkage rate normalizes it. The formula is straightforward:
Shrinkage rate = (Recorded inventory value − Physically counted inventory value) ÷ Recorded inventory value × 100
Suppose your inventory system shows $250,000 of goods on hand. A physical count values the actual stock at $246,000. The calculation runs:
- Difference: $250,000 − $246,000 = $4,000
- Shrinkage rate: $4,000 ÷ $250,000 = 0.016, or 1.6%
Some businesses prefer to measure shrinkage against sales rather than inventory value, since that is how the National Retail Federation benchmarks the industry. Both are valid — what matters is picking one definition and applying it consistently so your trend line means something.
So is 1.6% good or bad? It sits right around the recent U.S. retail average. Most retailers consider an acceptable range to be roughly 0.5% to 2% of inventory value, though it varies by industry. Grocery and apparel, with their high theft exposure and perishability, tend to run higher. A low-traffic B2B distributor should expect a much smaller number. The single most useful benchmark, though, is your own history. A rate climbing from 0.8% to 1.4% over three quarters is a signal worth investigating, regardless of where the industry average sits.
Tracking shrinkage at the category or location level multiplies the value of the metric. A blended company-wide rate of 1.5% might be hiding a 4% problem in one department and near-zero everywhere else. That granularity is exactly what cycle counting is built to deliver.
Why Cycle Counting Beats the Annual Physical Count
The traditional approach to verifying inventory is the annual physical count: shut down operations, mobilize every available employee, and count everything in one exhausting push. It satisfies auditors, but as an operational tool it is close to useless.
The annual count tells you that shrinkage happened sometime in the last twelve months. It cannot tell you when, where, or why. By the time you discover a discrepancy in March, the receiving error or theft pattern that caused it is nine months cold. You also lose a day or more of selling or shipping, and a rushed all-hands count performed by tired, untrained staff introduces its own errors.
Cycle counting flips the model. Instead of counting everything once, you count a small slice of inventory continuously — every day or every week — so the entire catalog gets verified on a rolling schedule. The benefits compound:
- Discrepancies surface within days, not months. A miscount or theft pattern gets caught while the trail is still warm and the cause is still identifiable.
- No operational shutdown. Counting happens alongside normal business by a small, dedicated team.
- Better accuracy. A trained two-person crew counting 40 SKUs carefully beats 30 untrained employees rushing through 8,000 SKUs.
- Continuous record accuracy. Because errors are corrected as they appear, your perpetual inventory data stays trustworthy year-round, which improves purchasing, fulfillment, and forecasting.
Cycle counting does not always replace the annual physical entirely — some auditors and tax situations still call for a full count — but a strong cycle program makes that annual count a quick confirmation rather than a frantic discovery exercise.
Using ABC Analysis to Decide What to Count and How Often
You cannot count every SKU with equal frequency, and you should not try. ABC analysis solves the prioritization problem by sorting inventory into three tiers based on annual consumption value — roughly, unit cost multiplied by sales velocity:
- A items — the vital few. Around 20% of your SKUs that drive roughly 70–80% of inventory value. High-cost or fast-moving goods.
- B items — the middle. Around 30% of SKUs accounting for roughly 15–25% of value.
- C items — the trivial many. Around 50% of SKUs accounting for only about 5% of value. Low-cost, slow-moving, or both.
The counting cadence follows the tiers. A common schedule looks like this:
| Class | Share of SKUs | Share of value | Count frequency |
|---|---|---|---|
| A | ~20% | ~70–80% | Every 30 days |
| B | ~30% | ~15–25% | Every 60 days |
| C | ~50% | ~5% | Every 90–180 days |
The logic is simple: spend your counting hours where the dollars and the risk are. A counting error on a $0.40 component barely moves your books; the same percentage error on a $400 part is a real problem. ABC analysis ensures your high-value, high-theft, high-velocity items get the attention they deserve while low-value items get a lighter touch.
Some teams add a fourth practice: control-group counting, where a specific set of items is counted repeatedly over a short window to test the counting process itself. If the same items keep producing different numbers, the problem is your method, not your inventory.
Making a Cycle Count Program Stick
A few practices separate cycle counts that work from those that quietly get abandoned:
- Set a fixed schedule and protect it. Cycle counting fails when it becomes the first thing dropped on a busy day. Treat the schedule as non-negotiable.
- Count from the shelf to the record, not the record to the shelf. Counting blind — without seeing the expected quantity — prevents the unconscious bias of "confirming" the number the system shows.
- Use technology where you can. Barcode scanners, warehouse management systems, and RFID dramatically cut counting time and transcription errors.
- Investigate before you adjust. Every discrepancy is a clue. Before changing the record, check for a misplaced pallet, an untracked transfer, or a receiving error. Adjusting without investigating throws away the most valuable output of the count.
- Train a dedicated crew. Accuracy comes from people who understand why the count matters and know the tools, not from whoever happens to be free.
Recording the Shrinkage Adjustment
Once a count confirms a genuine loss, the books need to reflect it. The mechanics are simple. Lost inventory is an asset you no longer have, so its cost moves off the balance sheet and onto the income statement.
The journal entry debits an expense account and credits the inventory asset. Say a cycle count uncovers $500 of missing goods:
Debit: Inventory Shrinkage Expense $500
Credit: Inventory $500You have two reasonable choices for the debit. You can post it to Cost of Goods Sold, which is technically correct and keeps shrinkage inside your gross margin. Or — better for management visibility — you can post it to a dedicated Inventory Shrinkage Expense account. Burying the loss inside a generic COGS or "inventory adjustments" bucket makes it nearly impossible to see how big the problem is or whether it is improving. A dedicated account turns shrinkage into a number you can put on a dashboard and manage.
Either way, the effect on the financials is the same: shrinkage raises your cost of goods sold and reduces gross profit. Under a perpetual inventory system — where the inventory balance updates in real time with every transaction — this adjustment is what reconciles the system's running balance to the verified physical count. The cycle count is the reality check; the journal entry is how reality gets recorded.
One discipline worth enforcing: tie every shrinkage entry to documentation from the count that produced it. When an auditor or owner asks why $4,000 of inventory vanished, "see the March cycle count variance report" is a far better answer than a shrug.
Keep Your Inventory Numbers Honest from Day One
Shrinkage is easiest to control when your underlying records are clean and every adjustment leaves a clear trail. That is exactly where plain-text accounting earns its keep. Beancount.io gives you transparent, version-controlled books — every inventory adjustment is a readable entry you can trace, diff, and audit, with no black box hiding the loss. When your shrinkage entries live in plain text alongside the rest of your ledger, the margin leak stops being invisible. Get started for free and see why developers and finance-minded operators are switching to plain-text accounting. For dashboards and reporting on top of your data, take a look at Fava.
Shrinkage will never hit zero — some loss is the cost of doing business. But the gap between a business that measures it monthly and corrects course, and one that discovers it once a year and absorbs the hit, is the difference between a managed cost and a silent one. Start with a shrink rate you trust, count on a schedule that matches each item's value, and record every adjustment where you can see it.