About the Inventory Accuracy Calculator
This inventory accuracy calculator scores a cycle count or physical count. Paste one line per SKU or location with the system (book) quantity, the quantity actually counted and, optionally, the unit cost. It reports inventory record accuracy (IRA) — the share of records whose count matches the system within your tolerance — along with unit accuracy, dollar-value accuracy, net and gross variance and the value of shrink.
It is meant for warehouse managers, inventory control analysts, retailers and auditors who need a quick, consistent way to grade count results and see which items drive the error. Record accuracy is the metric most warehouses track because a single wrong record can cause a stock-out or a failed pick, even when total units look fine; the unit and value measures show how large the errors are.
Tolerance is a percentage of the system quantity: with 2% tolerance, a record of 120 units counts as accurate if the physical count is between 117.6 and 122.4. Set tolerance to 0 for strict, exact-match accuracy, which is what many companies use for high-value items. A record with zero on the books is only accurate if nothing is found.
With the default inputs, the inventory record accuracy is 70%. Change any value above to recalculate instantly.
How to use the inventory accuracy calculator
- 1Export or type your count sheet: SKU, system quantity, counted quantity and unit cost.
- 2Paste it into the box, one record per line, with “|” or tabs between fields.
- 3Set the tolerance your company uses (0 for exact match).
- 4Read record accuracy, then check the table and chart for the SKUs with the biggest variances.
- 5Recount and research large variances before posting inventory adjustments.
Formula and method
A record is accurate when |counted − system| ≤ tolerance × system quantity. Inventory record accuracy (IRA) is the number of accurate records divided by the number of records counted. It treats every record equally, so a small error on a cheap part hurts the score as much as a large one on an expensive part — which is deliberate, because either can cause a failed pick.
Unit accuracy compares the total absolute unit error with total system units, and value accuracy does the same in money using each line’s unit cost. Net variance lets overages and shortages cancel out (it is what hits the books after an adjustment), while gross variance adds their absolute values. Shrink is the value of units missing where the count came in below the system quantity, shown as a share of the book value counted.
- IRA
- Inventory record accuracy (%)
- system
- Book quantity in the WMS/ERP at count time
- counted
- Physical quantity counted
- tolerance
- Allowed difference as a share of the system quantity
Worked examples
10-SKU cycle count with 2% tolerance
A-100 (−2 of 120) and D-400 (−1 of 250) are inside 2% tolerance, but B-215 (+4 of 80), C-341 (−8 of 60) and E-500 (−1 of 12) are not, so 7 of 10 records are accurate — 70% IRA. Absolute unit error is 16 of 1,113 system units (98.56% unit accuracy). Shortages are worth $271.10, 3.68% of the $7,374 book value; netting the $25 overage on B-215 gives a net variance of −$246.10.
Strict exact-match accuracy
With no tolerance, A-100 and D-400 also fail, leaving only the five records whose counts exactly match the system: 50% record accuracy. Unit and value accuracy are unchanged because they measure the size of errors, not whether a record matches.
Small count without unit costs
At 5% tolerance, Bin 01-B (−1 of 25 = 4%) and Bin 01-C (−3 of 100 = 3%) both pass, so all four records are accurate. Absolute error is 4 of 173 units, giving 97.69% unit accuracy; the value measures are hidden because no unit costs were entered.
Frequently asked questions
How do you calculate inventory accuracy?+
The most common measure is inventory record accuracy: count a sample of SKUs or locations, mark each record accurate if the count matches the system within your tolerance, and divide accurate records by records counted. 92 accurate records out of 100 counted is 92% accuracy.
What is a good inventory accuracy percentage?+
Many distribution operations target 95% record accuracy or higher, and well-run warehouses with barcode scanning often aim for 99% or more. Retail stores frequently run lower. Track the trend over time and by product class rather than chasing a single number.
What tolerance should I use for cycle counts?+
Tolerances are usually set by item class: 0% for high-value or serialized items, and a small percentage (for example 1–5%) for low-cost, high-volume items counted by weight or estimated. Using a tolerance stops trivial differences on cheap items from hiding real problems.
What is the difference between net and gross variance?+
Net variance adds overages and shortages together, so they cancel out; it is the dollar adjustment that hits your books. Gross (absolute) variance adds the size of every error regardless of direction, which better reflects how inaccurate your records really are.
Is inventory shrink the same as variance?+
Shrink is the loss side of variance — inventory that should be there but is not, from theft, damage, miscounts, receiving errors or unrecorded scrap. Overages are not shrink, though they often point to the same process problems such as picks posted to the wrong SKU.