Physical Inventory: The Method for a Reliable Count

One Friday evening, in a warehouse, teams work until midnight to count every SKU by hand. The following Monday, inventory discrepancies disappear, and no one really knows where they came from. This scenario repeats itself every year at most French industrial and logistics companies. Physical inventory remains an accounting requirement and an operational necessity. If poorly prepared, it comes at a high cost in terms of time, reliability, and confidence in inventory figures. When done well, it becomes a management tool rather than an annual chore that everyone has to endure.

The difference between an inventory that produces actionable data and one that generates unexplained discrepancies is rarely a matter of chance. It comes down to a method: advance preparation, a freeze on inventory movements, clear zoning, double-counting, and rigorous analysis of discrepancies. This page details each of these steps and explains why more and more manufacturers are subsequently choosing to switch to cycle counting and RFID to eliminate their reliance on a large annual count.

Talk to a CIPAM expert

Why a Physical Inventory Fails

Before going into detail about the method, it’s helpful to understand why so many physical inventories yield disappointing results. The causes are rarely related to a lack of effort on the part of the teams. They stem from organizational decisions made well in advance—sometimes several weeks before the day of the count.

The primary cause is inventory movement that has not been frozen. If receipts, shipments, or internal transfers continue during the count, the book inventory and the physical inventory are no longer comparable at the same point in time. Even the slightest time lag between a scanned outbound transaction and an area that has already been counted is enough to create a phantom discrepancy.

The second cause is unclear zoning. When two teams count the same zone without realizing it, or when a peripheral zone is overlooked, duplicates and omissions quietly accumulate. This problem particularly affects warehouses with stock on the floor, quarantine zones, or untracked returns.

The third cause is the lack of double-counting for high-value or fast-moving items. A label-reading error, a misunderstood unit of measure, or confusion between two similar lots goes unnoticed if no one recounts the items. It is often these high-value or fast-moving items that account for the inventory discrepancies that are most costly to correct later.

Finally, the fourth cause is a sloppy reconciliation between the physical count and the theoretical inventory in the ERP or WMS. Many companies approve discrepancies without investigating their source. The inventory is automatically corrected, but the underlying cause remains, which guarantees that the same type of discrepancy will reoccur the following year.

The Method, Step by Step

A reliable physical inventory always follows the same process, regardless of the industry or the size of the warehouse. The process consists of five steps, from preparation to the final analysis of discrepancies.

Prepare and freeze the movements

It all begins several days before the inventory count date. You must physically clean the storage areas, organize the shelves, remove obsolete products, and ensure that each item has a legible barcode label. This often-overlooked preparation phase determines the speed and reliability of the inventory count.

The suspension of all movements is then non-negotiable. Receipts, shipments, and internal transfers must be halted for the entire duration of the inventory count, or, if that is not possible, must be isolated in a dedicated and clearly identified area. Without this freeze, the book inventory continues to change while the physical inventory is frozen, making any reconciliation only an approximation.

Zone and Assign Teams

The division into counting zones must be comprehensive and without overlap. Each zone is numbered, assigned to a specific team, and linked to a dedicated counting sheet or form. Non-standard areas—such as floor stock, quarantine, returns, and samples—must be explicitly listed to ensure they are never overlooked.

An up-to-date warehouse layout, distributed to each team before the start of the count, helps prevent errors related to the scope of the count. This is also the time to establish the sequence of areas to be counted so that the areas being counted remain off-limits to goods movements until the count is complete.

Counting in Pairs

Double counting involves having each area—or at least the high-value, fast-moving items—verified by a second person who is independent of the first counter. The two results are compared before being validated. If there is a discrepancy between the two counts, a third count is conducted to resolve the issue.

This step goes faster than it seems when teams use mobile devices instead of paper forms. Directly scanning the barcode or RFID tag reduces transcription errors and allows the two counts to be compared in real time rather than at the end of the day.

Adjust to the theoretical inventory

Once the physical count has been validated, it is compared with the book inventory recorded in the ERP or WMS. This reconciliation must be performed on a SKU-by-SKU basis, not just by total value. Two errors that offset each other in value may mask a real problem involving two different products.

The reconciliation generates a list of discrepancies—positive or negative—sorted by amount and frequency. This list will guide the final step and is then used to calculate the inventory reliability rate, an indicator that is tracked over time rather than merely assessed annually.

Analyze the discrepancies

Every significant discrepancy requires a documented explanation before it can be validated for accounting purposes. Data entry errors upon receipt, theft, unreported breakage, confusion regarding units of measure or reference, or unrecorded transfers: identifying the root cause allows you to correct the process that caused it, not just the inventory figure.

This analysis must be formalized and shared with the relevant teams. A discrepancy that recurs every year within the same product family almost always indicates a process issue rather than an isolated incident. The industrial traceability glossary provides useful definitions to establish a common vocabulary between field teams and accounting teams.

Switch to a rotating inventory system

A rigorous method makes the annual physical inventory more reliable, but it does not solve its structural problem. Shutting down an entire warehouse for several days, having teams work overtime, and halting the movement of goods comes at a real operational cost. Many industrial companies also find that the snapshot taken once a year quickly becomes outdated in light of the day-to-day reality of inventory flows.

The rolling inventory addresses this limitation by replacing the large annual count with partial, regular counts—zone by zone or family by family—spread out over the course of the year. Each cycle follows the same process described above—preparation, zoning, double-checking, and reconciliation—but on a smaller scale and without bringing the entire operation to a standstill.

RFID is a true game-changer in this area. A single RFID scan can identify dozens of items at once, without having to target each tag individually as with a traditional barcode scan. The comparison between RFID inventory and barcode inventory details the situations in which each technology delivers the best return, depending on the nature of the products, storage density, and desired counting frequency.

The Benefits of a Well-Managed Inventory Method

Organizing your physical inventory using this method yields benefits that go beyond simply meeting the annual accounting requirement. Teams feel more at ease, and a well-prepared, zone-based count proceeds without the usual stress associated with improvisation and overlooked areas.

The reliability of inventory data is improving steadily. Rigorous reconciliation and systematic analysis of discrepancies make it possible to address the root causes rather than repeating the same accounting adjustments every year without ever resolving the underlying problem.

Trust between field teams, logistics, and accounting is also growing. When everyone understands the method and sees discrepancies decrease from one cycle to the next, inventory stops being seen as an imposed constraint and becomes a shared management tool.

Finally, the transition to a continuous inventory system using RFID or barcodes frees up operational time. Regular, targeted counts replace the annual complete shutdown of operations, which reduces pressure on teams and improves actual inventory availability throughout the year.

Request an exchange with CIPAM Explore the RFID rotating inventory

Frequently Asked Questions About Physical Inventory

What is the difference between a physical inventory and a cyclical inventory?
A physical inventory involves counting the entire stock on a specific date, usually once a year. A cyclical inventory spreads this counting out over the course of the year, area by area or product category by product category, without tying up the entire warehouse.

Is it really necessary to freeze all inventory movements during the inventory count?
Yes, or, failing that, isolate the current transactions in a clearly separate area. Without freezing transactions, the book inventory continues to change while the physical count is frozen, which skews the final reconciliation.

Is double-counting necessary for all SKUs?
It is recommended, at a minimum, for high-value, high-turnover items, or those that have historically been sources of discrepancies. In a large warehouse, applying double-counting to the entire inventory may prove too burdensome without providing a proportional benefit.

How can we explain an inventory discrepancy that occurs every year for the same SKU?
A recurring discrepancy within the same product family generally indicates a process issue rather than an isolated incident, such as a repeated data entry error upon receipt, confusion over units of measure, or an incorrectly identified storage area. Root cause analysis allows this process to be corrected upstream.

Does RFID completely replace barcodes for inventory?
Not always. The choice depends on the nature of the products, the storage density, and the desired counting frequency. Some companies combine the two technologies based on product categories rather than switching entirely from one solution to the other.

How long does it take to set up a rotating inventory system?
Deployment depends on the number of SKUs, the division into zones, and the information system already in place. The process generally begins with an audit of the existing process, followed by a phased rollout in pilot zones before full-scale implementation.

Is inventory software essential for implementing this method?
It is not strictly required for an initial physical inventory conducted using inventory sheets. However, it becomes essential once the company switches to a cycle count system, as it allows for cycle planning, centralization of discrepancies, and integration of results with the ERP or WMS.