Multiple correspondence analysis for customer segmentation in large retail groups

The aim of the following work is to highlight the importance of categorical methods as applied to marketing strategies. In the past, multivariate statistical techniques were used for quantitative data in marketing decision support systems (MDSS), however, given the introduction of many categorical variables in present day marketing research, the elaboration of these variables requires the use of categorical statistical methods. In order to carry out this research, we have employed multiple correspondence analysis, which, by means of maps constructed on a limited number of latent dimentions, simplifies the reading of both the intricate relations among the numerous categorical variables observed and their categories. In market research on the purchasing behavior of consumers, these analyses have been used to determine the essential aspects of consumer behavior as a rational basis for adopting opportune marketing strategies.