Data-Driven Customer Segmentation and Campaign Optimization Using Predictive Analytics in the Retail Sector

Abstract

The digital era transformed how businesses created value, interacted with customers, and made strategic decisions. Retail companies, in particular, faced increasing pressure to adapt to rapidly evolving consumer preferences, competitive markets, and emerging technologies. This study examined the role of predictive analytics as an enabler of business transformation in the retail sector, using XYZ Retail as a case example. Leveraging sample transactional data from the SAS Viya library, the research applied statistical modeling—logistic regression and decision tree algorithms—combined with Recency, Frequency, and Monetary (RFM) metrics to predict customer responses to a targeted cashback campaign. Model performance was evaluated using misclassification rates, with the decision tree model outperforming logistic regression. Cluster analysis further segmented customers into actionable groups, enabling the development of precise marketing strategies. The findings demonstrated how predictive analytics supported broader digital transformation initiatives by enabling data-driven decision-making, improving customer engagement, and optimizing resource allocation. The study contributed both a methodological approach for predictive modeling in marketing and strategic insights for integrating analytics into long-term business transformation plans.