Predicting Cross-Selling Health Insurance
\nProducts Using Machine-Learning
\nTechniques

This study delves into the utilization of Machine Learning (ML) techniques for predicting health
\ninsurance cross-selling behavior in South African consumers. The main goal is to create a robust
\nML model that assists health insurance companies in pinpointing potential customers with higher
\nprobabilities of purchasing additional health insurance products. Employing quantitative
\nmethodology, the study extracted consumer data and applied various ML algorithms such as
\nrandom forest, K-nearest neighbors, XGBoost classifier, and logistic regression using Python.
\nTailored feature engineering techniques were employed to enhance predictive accuracy.
\nAnalyzing 1,000,000 customer records with 16 features, Random Forest emerged as the topperforming
\nmodel, achieving an accuracy score of 0.99 and F1 score of 1.00. The study reveals
\nthat customers aged 25–70, with prior insurance and longer service history, are more inclined to
\npurchase additional health insurance products. These findings provide actionable insights for
\nrefining marketing strategies, boosting customer acquisition, and increasing revenue.