Low inventory turnover represents a recurring problem in trading companies, as it leads to high holding costs, capital immobilization, and deficiencies in service levels. This study proposes an improvement process aimed at increasing inventory turnover through the integration of a Random Forest–based demand forecasting model and a slotting strategy applied to the warehouse. The study is conducted in a Peruvian company engaged in the commercialization of solar water heaters, whose products exhibit seasonal demand variability. The methodology includes an analysis of the current situation, the development of a demand forecasting model, the reorganization of the warehouse based on product turnover, and the validation of the proposed solution through discrete-event simulation using Arena software. The results show an improvement in forecasting accuracy, a reduction in days of inventory coverage, and an increase in inventory turnover for critical products. Furthermore, the simulation demonstrates enhanced operational efficiency under the proposed scenario. The developed approach proves to be applicable to trading companies facing inventory management challenges and logistical decision-making constraints.

Keywords

Inventory turnover, Demand forecasting, Random Forest, Slotting, Discrete-event simulation