.jpg)
The ABC method derives from the Pareto principle (80/20), which in warehouse management allows goods to be classified by turnover. Usually, 20% of items with the highest turnover are selected and classified as A. Meanwhile, the XYZ classification distinguishes assortment items with more predictable demand from those with less regular demand, which results in forecasts with higher errors. There is extensive literature on this topic, and almost every practitioner has heard of the ABC/XYZ matrix.
When taking a closer look at their specifics and the needs in warehouse resource management, it turns out that it’s not so obvious. I had the opportunity to personally examine ABC/XYZ calculation methods in several ERP and WMS systems and noticed that in many solutions the Pareto principle is uncritically transferred to warehouse operation levels. In my opinion, this is a wrong approach. Why? Many years ago, I already had concrete doubts about the ABC method, including:
I have observed that while the classic approach to warehouse management based on the ABC/XYZ matrix works well for purchase planning, item, and supplier categorization, it is not always possible to use this method to optimize warehouse processes. To support this statement, here are some arguments:
➡️ CONCLUSION 1: From the perspective of operational warehouse management, the number of fast-moving goods (class A) is determined by the organization of the specific warehouse, and this number is CONSTANT because it is a direct derivative of the number of locations with specific parameters — such as the aforementioned access time. Thus, the 20% indicator for the highest turnover SKUs (A) does not apply here.
➡️ CONCLUSION 2: Every warehouse, differing in location structure, will have a different number of SKUs classified as fast-moving.
➡️ CONCLUSION 3: The WMS system should support multi-criteria descriptions of individual locations. For example, in our OPTIpromag WMS system, besides ABC/XYZ, we also have COI — but that is a topic for another time.
The answer seems simple: we must calculate the average quantity of SKUs issued per day. Basic.
However, analyzing this issue more broadly raises questions:
➡️ CONCLUSION 1: Different industries have goods with varying purchase and sales frequency. For example, in a standard market, FMCG items should have a short averaging period (e.g., 3–7 business days), also considering weekdays (related to stock replenishment files on Mondays and Fridays). Meanwhile, industrial goods in the same market may have much lower turnover, and the averaging period for this category could be 30 days. System reports are essential for analyzing optimal averaging periods.
➡️ CONCLUSION 2: The system should count only working days in turnover calculations. This seems obvious but is often overlooked.
➡️ CONCLUSION 3: Choosing the right unit of measure for turnover class calculation is crucial. Usually, this is the default issue unit or the basic unit of measure. In our WMS, we use the basic unit of measure. Therefore, every consultant must understand how to configure item and warehouse master data properly.
Here, we can fall into the “trap” of the unit of measure used in turnover calculation. This can happen if order structures change, e.g., low-price goods are promoted for ordering by cartons/sets instead of pieces. From the WMS perspective, what matters is the amount of time and work spent on a specific SKU — such as the number of picks, put-aways, and scans. These parameters determine time consumption and costs, hence the need and desire for their optimization.
For this reason, ABC metric calculation should be two-dimensional. In our OPTIpromag WMS system, we use two dimensions: turnover and the number of operations on the SKU. For both metrics, weights can be configured. This approach makes the ABC metric calculation more accurate, automatically responds to trend changes, and better reflects actual processes.
Calculating metrics for SKUs with a history of receipts and issues, considering the above points, should not be too difficult. The challenge lies with new SKUs. This is not straightforward because we may have new SKUs distributed according to a PUSH policy — when recipients do not order a SKU themselves but are obligated to accept it (e.g., planned promotional campaigns in retail chains). It differs when a SKU is distributed according to a PULL policy, where the distribution network and customers order the goods themselves. In such cases, the system should not wait for historical data to accumulate because it would disrupt optimal item placement, picking paths, and increase warehouse operation costs.
In the OPTIpromag WMS system, we use a mechanism to determine turnover levels for SKUs without issue history. This is useful when introducing product substitutes or goods expected to sell in large quantities. The mechanism relies on multiple parameters, including: average turnover rate in the product group over a comparison period (e.g., previous holidays), comparative goods, and sales forecasts. Thanks to this solution, the system can optimize allocations of new SKUs from the moment of their receipt, using standard control parameters.
The goods receipt operations are one of the three key processes in warehouse operational management. They are most commonly associated with deliveries, but also include transfers from other warehouses, commercial returns, and complaints.
The Invident system supports production-related areas through the integration of WMS with MES, eliminating errors in the processes of receiving, allocating, and labeling goods. Features such as automatic RFID identification, manual pallet registration, and quality control ensure accuracy, efficiency, and smooth warehouse operations.
.png){kind=logo}
