Predicting Consumer Behavior Based on Big Data of User-Generated Online Content in Retail Marketing

The purpose of this study is to create, using big data from user content in retail marketing, a prediction approach to predicting consumer behavior. Based on an approach with two key components, prediction is achievable. The first step is accurate big data analytics of user-generated material, highlighting the essential information and monitoring changes in user behavior (posting and purchasing) as a result of shifting value proposition variables. Second, it involves the inclusion of specialists and seasoned marketers in sociological surveys that use large samples of respondents and the probabilistic method. The value proposition structure was broken down into ten components that influence the rhetoric of user content using the stratification approach. The competitive advantages or business objectives of stores, in turn, made clear the essential categories of user content. The study focuses on Russia’s most widely used digital trading platforms. The study developed an approach to the expert prediction of consumer behavior following changes in content quality and highlighted efficient digital tools for doing so using the sociological technique. The methodology for expert forecasting of consumer behavior amid changes in the quality of user content was developed using empirical, probabilistic, and sociological methods. The competitive advantage or goal of an online store was shown to be the most important element in altering consumer behavior. The proposed expert prediction methodology based on the likelihood matrix of a decline in customer conversion rates due to user content degradation is the study’s scientific contribution.