Classificação de métodos de previsão de demanda para novos produtos: estudo no sistema brasileiro de franquias

Abstract

This research was concerned in discussing demand forecasting to new products and realized a study in order to propose an assortment of demand forecasting methods to new products considering the Brazilian franchising system features. For this purpose, was used the questionnaire type data collection technique and was realized a questionnaire validation in four areas: psychology, statistics, specialists and Portuguese language, beyond the application of the Cronbach alpha coefficient – that resulted in a high rating global coefficient, which demonstrates the questionnaire is reliable and consistent. In order to contribute to the purpose treatment, was used an approach based on the principles of the method of multicriteria decision support, ELECTRE TRI. To apply the method, was used the IRIS 2.0 software. The results analysis was realized in three franchising segments. To the segment of bars, restaurants, bakeries and pizzerias, on additions to existing product lines category, the results show the methods market research, historical analogy, sceneries simulation, sales team research, Box-Jenkins (ARIMA) and regression analysis, they fit in recommended methods category. To the cosmetics and perfumery segment, on additions to existing product lines category, results show market research, delphi method, historical analogy, sceneries simulation, sales team research, moving average and Box-Jenkins (ARIMA), they fit in recommended methods category. To the bookstores, graphics and signage segment, on new products for the company category, results show market research, delphi method, historical analogy, sceneries simulation, sales team research, they fit in recommended methods category. Was possible to note demand forecasting for new products should incorporate specific characteristics of the markets as diversity in consumer preference and socio-economic changes.