Classification of products based on the uncertainty of supply chain demand: a case study of wineries in Chile

The wine industry faces distinctive supply chain challenges, including high product variety, export market fragmentation, and seasonal production, all of which contribute to demand uncertainty. Importantly, this uncertainty is not only externally driven but also amplified by tactical and operational decisions - such as labeling, bottling strategies, and product customization - that increase complexity. This study presents a product classification methodology based on demand behavior to improve decision-making in inventory management. Using a case study of three Chilean wineries located in the Central Valley, we compare the traditional ABC classification - commonly used in ERP systems - with a quantitative model that incorporates demand variability. The proposed approach enables segmenting products according to average demand and variability, offering clearer insights for setting differentiated service levels, inventory policies, and forecasting strategies. The findings show that the demand uncertainty-based classification provides more effective support for supply chain decision-making than conventional methods. The model has also demonstrated applicability beyond finished goods, such as in-process wine and critical inputs like corks and bottles. This research contributes empirical evidence to close the gap between theory and practice, providing a replicable tool for product segmentation in wine and other industries facing demand complexity.