This research uses the e-commerce transaction database to forecast customer's future behavior and active probability based on the BG/NBD model. The individual-level forecasting results help to identify customized needs, which help company to re-formulate marketing strategies, to retain existing customer base, and to increase efficiency of customer churn management. The calibrated data set includes non-contractual customers who made purchases on particular hepatic capsule with the repeat purchase frequency and purchase timing. Results indicate the proposed model is reliable in individualized forecasting to differentiate significantly the active probability and transaction total monetary. The results suggest the company should design adequate customer "awakening" programs for those inactive customers and to efficacious customer relationship management to integrate enterprise resources to the targeted individuals.