Image Analysis in Recommender Systems and Customer Relationship Management

Images are becoming increasingly prevalent online, and as such, marketing research and customer relationship management (CRM) strategies need to start incorporating visual data. As an application example, companies should implement different types of recommenders and include image attributes in their algorithms, depending on the products, the generalizability of historical data, and the availability of their customers’ profiles. We propose that a company can enhance customer experience and manage customer relationships with recommender systems. Recommenders should be a hybrid supported by both knowledge- and collaborative-based intelligence that can provide the appropriate recommendations depending on the relationship goal during a customer’s website visit. As image analysis is still a nascent field, we call for future research that will further its methodologies, apply it in marketing research and recommender systems, and measure the effect of visual information on building customer relationships and gaining consumer insights.