Machine Learning–Driven Brand Loyalty Analytics Framework for Predicting Customer Retention and Enhancing Personalized Engagement in the Culinary Industry

Abstract

Due to the fast adoption of various practices including digital ordering, loyalty programs, mobile apps, and social media in the culinary industry, there is an abundance of customer data available to marketers but at the same time, there are difficulties in detecting customers' risks of retention and providing efficient personalized engagement. In this regard, the current research presents a Machine Learning-Driven Brand Loyalty Analytics Framework, which involves the analysis of transactional behavior, loyalty program engagement, digital engagement, customer sentiment, and customer profile for the prediction of retention and provision of relevant engagement approaches. For this purpose, the current research uses the dataset, which includes 5,000 customer records gathered within one year. Data preprocessing includes the removal of duplicates, dealing with missing values, detecting outliers, data normalization, and feature transformation, followed by the development of five machine learning models: Random Forest, XGBoost, LightGBM, CatBoost, and Deep Neural Network. XGBoost outperformed other algorithms by achieving an accuracy of 94.3%, precision of 93.7%, recall of 92.8%, F1 score of 93.2%, and AUC of 0.96. Customer segmentation revealed 1,250 loyal customers with 96.5% retention probability, whereas 750 customers at risk had just 38.4% retention probability. Personalized engagement was linked to enhancements in retention rate from 72.4% to 87.8%, repeat purchase rate from 58.3% to 79.5%, customer satisfaction from 74.6% to 90.3%, and customer lifetime value from $420 to $615. These results highlight that the use of predictive analytics in combination with personalization can enhance customer retention and loyalty management. This framework serves as an effective base for culinary marketing using data analytics, while further research is needed on multi-restaurant data, real-time analytics, multimodal customer data, AI agents, and privacy-preserving learning.