Demand forecasting and planning has always been a very important issue for the efficiency of the entire supply chain. Sales forecasts are essential to formulate the sales budget, to plan production, to estimate the necessary production capacity and to plan promotions. Demand Planning therefore represents one of the most important strategic actions for the optimization of stocks, for an improvement in the level of customer service and for an overall reduction of costs. However, it is not possible to identify a single demand forecasting algorithm that can be applied to all types of companies and markets. The paper explores the problem, with the aim of identifying the forecasting process best suited to the business context under analysis and the most significant measurement metrics of forecasting accuracy.