This thesis has the objective of developing and applying a methodology to measure the impact of demand forecast inaccuracy in the supply chain of a beverage industry, specifically in the inventory management, physical distribution and sales processes. The purpose is to create an awareness of the importance of forecasting area in the logistics planning and execution activities. To achieve these goals, a conceptual review of the major demand forecasting methods and of the three logistics processes under analysis has been made. After that, a methodology was defined and applied to three different warehouse data sets of the company analyzed. As a result of the methodology application, some opportunities for process improvement were identified and some changes were proposed for the current demand forecasting process. The results of methodology application and proposed actions implementation allowed the company to increase the demand forecasting accuracy for the major SKUs and to improve communication among the different links of the supply chain. Based on more accurate forecasts, the company will be able to better allocate physical and human resources, reduce operational costs and achieve the required customer service level.