Forecast demand in a pharmaceutical trading company using ABC classification, Holt Winters method and ERP for an efficient business model.

Abstract

Pharmaceutical product trading companies often face challenges in inventory management due to inaccurate demand forecasts. This is the case for the company under study, which lacks an appropriate demand forecasting method, as evidenced by its Mean Absolute Percentage Error (MAPE) of 25.65%. Given this issue, there are effective methods to improve demand forecasting precision, such as Holt Winters, Arima, and the naive method. Some authors suggest, based on their research, that the optimal and regular MAPE should not exceed 10%. Therefore, this research proposes a design that integrates the ABC classification technique, a demand forecasting method along with an ERP system to enhance the accuracy of demand forecasting for the company. The results allowed identifying the most relevant products, i.e., those generating the greatest economic impact. Additionally, using Holt Winters and the ERP system, a MAPE of 2.60% and 1.45%, respectively, was achieved. This demonstrates the feasibility of the proposed design, meeting the established metrics.