Enhancing decision-making on detractor-causing failures: an approach combining data mining and machine learning

The current retail landscape poses new challenges as consumers become increasingly discerning. To remain competitive, retail companies must analyze their databases to identify failures that lead to customer dissatisfaction—particularly detractors, those who report negative experiences. In this context, this study proposes a methodology that combines data mining (DM) and machine learning (ML) to analyze and predict the root causes of detractors using transactional data from the big data repository of a major sports retail company. The proposed approach employs Decision Tree (DT) algorithms to uncover patterns linked to service failures. The results highlight two primary issues: delivery discrepancies and product returns due to dissatisfaction, both of which directly affect customer loyalty. Based on these findings, targeted action plans were developed to support continuous improvement and operational effectiveness. This research advances the state of the art by introducing an explainable AI model capable of identifying the most impactful failures affecting customer satisfaction in large-scale retail operations. In addition to its scientific contributions, the study offers practical insights by demonstrating how big data can support faster, more informed decision-making in customer experience management.