Research work · 2026
Predicting Customer Lifetime Value Using Machine Learning in CRM Predicting Customer Lifetime Value Using Machine Learning in CRM
Hemadri Ravilla
Frontiers in Emerging Artificial Intelligence and Machine Learning
Abstract
Customer Lifetime Value (CLV) forecasting plays a vital role in optimizing Customer Relationship Management (CRM) strategies. This research examines the application of machine learning to forecast CLV using the publicly available "Online Retail II Dataset" from the UCI Machine Learning Repository containing two years' worth of transactional data from an online retail company based in the UK. The salient purchasing frequency, monetary value, and recency characteristics were employed to train machine learning models such as Random Forest, Gradient Boosting, and Neural Networks. The method indicated that the best accuracy was achieved through the use of Gradient Boosting with the mean absolute error (MAE) of 7.5%, 15% higher than that achieved using standard techniques. The research indicates the efficacy of machine learning to forecast CLV accurately and thereby enable the optimization of retention strategies and forecasting revenues. The future will witness the application of deep learning to further optimize performance.