Comparison of classical RFM models and Machine learning models in CLV prediction

Abstract

This study analyzes the difference between classical RFM models and Machine Leaning (ML) models when calculating CLV with transactional data. In this paper we run different programs to analyze the CLV value by using both methods. Based on the results, the researchers found out that Pareto/NBD model have better predictive power of performing CLV predictions than ML models. Lastly, the findings proved the effectiveness of the Pareto/NBD method of calculating CLV.