Segmentation of Chinese consumer preference for wine extrinsic attributes based on stratification and weighted clustering algorithm

Abstract

Chinese consumers are unfamiliar with the sensory attributes of wine and often pay more attention to the extrinsic attributes of wine that can most directly present information. Therefore, it is crucial to study consumer preferences for extrinsic attributes. This paper proposes a consumer segmentation method based on stratification and weighted clustering algorithm, leveraging data from an anonymous online survey (N = 3179). The AFR model is constructed based on the RFM model to segment wine consumers from the perspective of customer value. This model removes Monetary indicator and adds a new indicator Amount to eliminate multicollinearity caused by the Frequency and Monetary indicators of RFM model. In addition, factor analysis is used to assign weights to clustering indicators, addressing the problems of unequal weighting and strong correlation among the indicators. This allows for a simple, efficient, and precise segmentation of wine consumers at different levels. The method proposed in this article segments consumers into two layers and six categories: the potential layer includes product-oriented type, amorphous type, and cheap-fine type, while the mature layer comprises the rational type, reputation-price type and random type. This research will provide wine producers and distributors with a simple and efficient basis for production and marketing decision-making. Additionally, it can serve as a reference for consumer segmentation in food and other fields.