E-book platforms have gained popularity due to their affordability, portability, and extensive title selection. This study analyzes customer behavior based on their platform activities before the launch of a subscription service and investigates strategies to boost subscription plan adoption. To achieve this, we employ customer segmentation through a clustering algorithm to examine customer behavior. Subsequently, we track customer behavior in two distinct groups: RFM and proposed features, regarding their engagement with purchased plans via A/B testing. The results indicate that the conversion rates for the proposed features outperform the RFM model in experiments by 0.42% (2072 customers) and 0.54% (27 customers).