SEGMENTATION OF SHERRY WINE CONSUMERS INTO 3 CLUSTERS ACCORDING TO THEIR CONSUMPTION FREQUENCY USING UNSUPERVISED MACHINE LEARNING

Resumen

Consumer segmentation is a fundamental for understanding the market and deciding appropriate marketing strategies for a product. Sherry wines, segmentation has been carried out for enotourism in the region and considering consumption preferences, but key elements such as consumer attitudes or the evaluation of intrinsic and extrinsic wine attributes have been overlooked. This study proposes a segmentation of sherry wine consumers based on their attitudes towards wine consumption and their evaluation of wine attributes. A sample of 1543 participants (46.9% men and 53.1% women; mean age = 29.71, sd = 12.212) was used to classify consumers using K-means cluster analysis into 3 clusters, using as input variables the normalized values of the 3 attitudinal dimensions of the Wine Attitudinal Scale for Consumer Research (Wine experience, Health consumption, and Social component) and the normalized mean scores of intrinsic attributes (Taste, aroma, color, alcohol content, year, and grape variety) and extrinsic attributes (price, Designation of Origin, labeling, awards, brand, and bottling). Segmentation was evaluated using one-way ANOVA analysis and shown to be effective in classifying the study sample according to consumption frequency through correspondence analysis (X2 = 360,240; p<,001). The results of this study represent a deeper approach to consumer psychology using an unsupervised machine learning approach.