In issue 5/2008 of “Logistyka”, the author published the first article in the series “Between theory and practice of inventory management”, which discussed the importance of correctly determining the replenishment cycle time. Subsequent articles in this series covered:
– the consequences of errors made when determining the standard deviation of demand during the replenishment cycle (“Logistyka” 6/2008),
– the consequences of the erroneous assumption regarding the type of demand distribution during the replenishment cycle (“Logistyka” 1/2009),
– the impact of replenishment cycle time variability on the correctness of inferences about the relationship between safety stock and service level (“Logistyka” 2/2009).
This article is a continuation of the series, and its purpose is to discuss the fundamental differences between the two definitions of service level related to inventory management and to point out the potential consequences resulting from their misinterpretation.
First of all, it should be emphasized that the service level can be related to both a single stock-keeping unit and groups of items. In the first case, we define the availability of a specific material (product, good), while in the second, we determine the availability of groups (sets) of materials, for example for multi-item orders. The subject of consideration in this article will be the availability of individual stock-keeping units. The two basic indicators of customer service level (availability level) are: probability of satisfying demand (PSD) and fill rate (FR).
The difference between both definitions is exhaustively presented in the available literature [e.g., 1, 3], however, the author’s experience in contacts with enterprises indicates that these concepts are still very often not properly understood and distinguished. This often becomes the cause of incorrect shaping of safety stock and the generation of unnecessary costs.
The issue will be presented in two parts. In the first (this article), the foundations of both definitions and the principles of calculating both indicators will be presented in detail. This will be shown using the example of an item with low periodic demand, the distribution of which can be described by a Poisson distribution, because in this case, it is easier to demonstrate the essence of the calculations. The second part of the article will concern cases where the demand distribution corresponds to a normal distribution, and the results will be generalized.
The article comes from the magazine “Logistyka” 6/2010.