Acta Logistica | 2026
Authors: Zizka M.; Kustka M.
DOI: 10.22306/al.v13i1.746
Journal: Acta Logistica
Year: 2026
Publisher: 4S go, s.r.o
Document Type: Article
Open Access: All Open Access; Gold Open Access
Cited by: 0
The article deals with the very actual topic of inventory management with sporadic demand in small and medium-sized enterprises. The research aims to propose an easy-to-apply procedure for predicting dependent demand or consumption based on the analysis of the inventory assortment. The proposed methodology is particularly suitable for small and medium-sized enterprises and therefore does not require any special software products. The requirement is that it must be sufficiently reliable in terms of prediction accuracy. The methodology can be applied using commonly available spreadsheets. The solution is based on dividing the inventory assortment into nine groups based on the results of ABC/XYZ analyses. The sample consists of 60 inventory items with fluctuating demand predominating. Six forecasting methods were tested on this sample (naive forecasting, shorter and longer moving averages, linear regression with and without seasonal indices, and Croston's method). The solution was then supplemented with a bootstrap simulation. The reliability of each method's forecast was evaluated using the metrics of mean absolute error, mean absolute percentage error, and demand planning accuracy. Based on the empirical results, recommendations were formulated for selecting a suitable forecasting method for each inventory group. © 2026, 4S go, s.r.o. All rights reserved.
bootstrap method; Croston's method; custom production; demand forecasting; sporadic demand