Identifying Application Areas for Machine Learning in the Retail Sector

Abstract

Machine learning (ML) has the potential to take on a variety of routine and non-routine tasks in brick-and-mortar retail and e-commerce. Many tasks previously executed manually are amenable to computerization using ML. Although procedure models for the introduction of ML across industries exist, the tasks for which ML can be implemented in retail need to be determined. To identify these application areas, we followed a dual approach. First, we conducted a structured literature review of 225 research papers to identify possible ML application areas in retail, as well as develop the structure of a well-established information systems architecture. Second, we triangulated these preliminary application areas with the analysis of eight expert interviews. In total, we identified 21 application areas for ML in online and offline retail; these application areas mainly address decision-oriented and economic-operative tasks. We organized the application areas in a framework for practitioners and researchers to determine appropriate ML use in retail. As our interviewees provided information at the process level, we also explored the application of ML in two exemplary retail processes. Our analysis further reveals that, while ML applications in offline retail focus on the retail articles, in e-commerce the customer is central to the application areas of ML.