Consumer Analytics and Strategic Marketing Management: Developing an Intelligent Framework for Predicting Customer Loyalty in Digital Markets

Abstract

The increasing availability of consumer data creates opportunities for organizations to strengthen strategic marketing through intelligent prediction of customer loyalty. This study developed and evaluated a consumer analytics framework integrating demographic, transactional, behavioral, experiential, and digital engagement characteristics to distinguish loyalty from at-risk customers in digital markets. Data from 3,000 customers were analyzed using descriptive statistics, association tests, predictor screening, and supervised machine learning. Five classification algorithms were evaluated through stratified cross-validation, followed by hyperparameter tuning and independent test-set assessment. Logistic Regression provided the strongest overall predictive performance, achieving a test ROC-AUC of 0.8343, balanced accuracy of 0.7657, F1-score of 0.7184, and PR-AUC of 0.7555. Customer satisfaction and purchase frequency emerged as the strongest positive predictors of loyalty, while longer recency and greater discount usage were prominent negative predictors. Customer tenure, email engagement, NPS, social engagement, return behavior, and service responsiveness provided additional predictive information. The findings demonstrate that an interpretable analytics framework can combine heterogeneous consumer signals to identify loyalty patterns effectively. The proposed approach supports proactive customer segmentation, targeted retention strategies, relationship development, and evidence-based allocation of marketing resources, providing a practical foundation for intelligent strategic marketing management in increasingly data-driven digital environments.