A matter of value? Predicting channel preference and multichannel behaviors in retail

• Published in: Technological forecasting & social change 2021-01, Vol.162, p.120401, Article 120401
• Main Authors: Acquila-Natale, Emiliano, Iglesias-Pradas, Santiago
• Format: Article
• Language: English
• Subjects: Algorithms; Behavior modification; Channel preference; Channels; Companies; Consumer behavior; Consumers; Demographic variables; Machine learning; Multi-channel retailing; Perceived value; Prediction models; Predictions; Preferences; Retail industry; Retailing; Sales; Shopping; Value creation
• ISSN: 0040-1625; 1873-5509
• DOI: 10.1016/j.techfore.2020.120401

•We use value components to predict channel preference and multi-channel behaviors.•We explore both prediction and classification algorithms.•Physical stores are still necessary to remain competitive.•We recommend the use of multiple but simplified delivery and returns options.•Value-based features perform good at classifying consumers’ channel preferences and behaviors. The Internet has changed consumer shopping behaviors and purchasing habits. Consumers use the channels that best suit their needs at any given time to have an enhanced shopping experience, forcing companies to innovate in their channel offerings and the ways they manage these channels. This makes it necessary to detect and predict shoppers’ multi-channel behaviors and channel preferences. Based on the idea of value creation and the different aspects of the value perceived by consumers, this study aims to identify the variables predicting channel preference and multi-channel behaviors in retailing. To build a predictive model, the research identifies five categories related to perceived value (perceived quality, monetary costs, non-monetary costs, hedonic elements, and brand knowledge), adding demographic characteristics and variables related to lock-in effects in multichannel shopping behavior. The theoretical predictive model is then empirically tested by comparing the results of traditional and machine learning techniques and algorithms with data from an online questionnaire answered by a representative sample of Spanish consumers of clothing and apparel products. The study contributes to a better understanding of the variables predicting multi-channel behaviors and channel preference, providing companies with actionable insight into how they should manage their multi-channel offering to increase sales.