Inventory discrepancies are differences between the quantity, location, condition, or value recorded in an inventory system and what is physically available. Left unresolved, they can distort replenishment decisions, delay orders, create avoidable write-offs, and weaken confidence in financial records. This guide explains how to measure, investigate, correct, and prevent inventory discrepancies without relying on unsupported industry-wide percentages.
What Is an Inventory Discrepancy?
An inventory discrepancy exists when a verified physical count does not agree with the inventory record. A discrepancy may involve units, value, storage location, lot or serial number, ownership, or stock status such as available, damaged, quarantined, or expired. A count difference is a signal to investigate—not automatically proof of theft, counting error, or a system failure.
Common Types and Causes
- Receiving errors: Short shipments, over-receipts, duplicate receipts, or receipts posted against the wrong purchase order.
- Putaway and location errors: Stock is physically present but stored in an unrecorded, adjacent, staging, or exception location.
- Unposted movements: Transfers, picks, production issues, returns, or disposals occur physically but are not posted correctly or on time.
- Unit-of-measure errors: Cases, packs, kilograms, and individual units are converted or recorded inconsistently.
- Identification and master-data errors: Similar labels, duplicate item codes, incorrect barcodes, or inaccurate pack configurations cause the wrong SKU to be transacted.
- Condition or status errors: Damaged, expired, quarantined, or customer-owned stock is counted as available inventory.
- Counting errors: Poor count instructions, interrupted counts, double counting, inaccessible stock, or non-blind counts can affect results.
- Unauthorized removal or fraud: Theft and manipulated transactions are possible causes, but they should be concluded only after evidence is reviewed.
Inventory Discrepancy Formulas
Define the measurement method before comparing sites or periods. The same operation can report different “accuracy” results depending on whether it measures units, SKUs, locations, or value.
- Unit variance = Physical quantity − System quantity
- Absolute unit variance = |Physical quantity − System quantity|
- Value variance = Unit variance × Approved unit cost
- Unit-based inventory accuracy (%) = [1 − (Total absolute unit variance ÷ Total system quantity counted)] × 100
- SKU accuracy (%) = (SKUs with no variance ÷ Total SKUs counted) × 100
Use absolute variance when calculating unit-based accuracy so shortages and overages do not cancel each other out. The valuation basis—such as standard, average, or another approved cost—should follow the organization’s finance policy.
Worked Inventory Audit Example
| SKU | System qty. | Physical qty. | Unit variance | Unit cost | Value variance |
|---|---|---|---|---|---|
| A | 100 | 96 | −4 | $12 | −$48 |
| B | 60 | 63 | +3 | $8 | +$24 |
| C | 25 | 25 | 0 | $40 | $0 |
| D | 120 | 114 | −6 | $5 | −$30 |
Total system quantity is 305 units and total absolute variance is 13 units. Unit-based accuracy is therefore approximately 95.74%: [1 − (13 ÷ 305)] × 100. One of four SKUs has no variance, so SKU accuracy is 25%. The net value adjustment is −$54. These measures answer different questions and should be reported with their definitions.
How to Investigate an Inventory Discrepancy
1. Control movements and confirm the cutoff
Pause or tightly control transactions in the affected area. Confirm the count time and identify receipts, shipments, transfers, returns, or production postings that occurred around the cutoff.
2. Perform an independent blind recount
Ask a different counter to recount without seeing the system balance or first count. Verify the item, location, unit of measure, lot or serial number, and condition.
3. Check adjacent and exception locations
Search staging lanes, receiving docks, packing stations, quality-hold areas, returns, damaged-stock zones, and neighboring bins. Also check for mixed items or stock behind or above the expected location.
4. Reconcile the transaction trail
Review purchase orders, goods receipts, picks, shipments, transfers, returns, production consumption, write-offs, and manual adjustments. Compare timestamps, users, documents, and quantities.
5. Classify the root cause
Assign a specific cause code supported by evidence. Avoid broad labels such as “human error” when a process-level cause—unclear unit conversion, missing scan point, late posting, or poor location control—can be identified.
6. Approve and document the adjustment
Correct the record only under the organization’s approval policy. Retain the original count, recount, evidence, root-cause code, approver, adjustment reference, and financial impact.
7. Verify the corrective action
Assign an owner and due date, then review later counts or transactions to confirm the cause has not recurred.
Root-Cause Analysis Guide
| Process area | Evidence to review | Possible corrective action |
|---|---|---|
| Receiving | PO, delivery note, receipt time, dock records | Two-step quantity check; exception workflow |
| Putaway or transfer | Scan history, source and destination bins | Require confirmation at both locations |
| Picking or shipping | Pick list, pack record, shipment confirmation | Scan item and location at pick or pack |
| Returns | Return authorization, inspection and disposition | Separate returned, available and quarantined stock |
| Master data | Item code, barcode, pack size, unit conversion | Govern changes and validate duplicate records |
| Unauthorized activity | Access logs, overrides, adjustment history, CCTV where lawful | Segregate duties and review unusual adjustments |
How to Prevent Recurring Variances
- Use risk-based cycle counting: count high-value, fast-moving, regulated, or historically inaccurate items more often.
- Separate counting, adjustment approval, and custody where practical.
- Standardize item identifiers, units of measure, pack sizes, locations, and condition codes.
- Capture receipts, moves, picks, returns, and disposals at the point of activity with barcode or RFID controls where justified.
- Physically and digitally separate available, damaged, expired, returned, and quarantined stock.
- Review manual adjustments, negative balances, repeated recounts, and unusual user activity.
- Track root causes and recurrence, not only the total adjustment value.
For related planning controls, try the ABC Analysis Calculator to prioritize items and the Safety Stock and Reorder Point Calculator to test replenishment assumptions.
Inventory Discrepancy Audit Checklist
- Is the transaction cutoff defined and communicated?
- Are counters sufficiently independent from stock custody?
- Are initial counts blind, with recount rules defined?
- Does testing cover both floor-to-record completeness and record-to-floor existence?
- Are item, unit, lot or serial, location, ownership, and condition verified?
- Are adjustments supported, approved, and traceable?
- Is each material or repeated variance assigned an evidence-based root cause?
- Does every corrective action have an owner, deadline, and follow-up check?
Frequently Asked Questions
What is a good inventory accuracy percentage?
There is no universal target that fits every operation. The appropriate threshold depends on item value, customer and regulatory risk, process capability, and the metric definition. Set targets by risk class and publish the formula with the result.
Should every discrepancy be adjusted immediately?
Not necessarily. First confirm the cutoff, recount the item, and review open or late transactions. Once the difference is validated, follow the documented approval and accounting policy.
How often should inventory be cycle counted?
Use a risk-based schedule. High-value, fast-moving, regulated, or repeatedly inaccurate items generally deserve more frequent counts than stable, low-risk items. Adjust the schedule using observed causes and results.
Conclusion
Inventory discrepancies are best managed as control failures to understand, not merely balances to overwrite. Define the metric, protect the cutoff, recount independently, trace the transaction history, document the adjustment, and verify the corrective action. That approach improves both operational decisions and the reliability of inventory records.
Authoritative References
- U.S. Government Accountability Office: Executive Guide—Best Practices in Achieving Consistent, Accurate Physical Counts of Inventory and Related Property
- U.S. GAO: Inventory System Checklist
- GS1: Warehouse Management
- GS1 Barcode Standards
- PwC Global Economic Crime Survey 2024
- Association of Certified Fraud Examiners: Occupational Fraud Report
- IBM: Inventory Management Overview
