This study investigates how consumer behavior in the e-book market is influenced by electronic word-of-mouth (eWOM) and online customer reviews (OCRs). Addressing the limitations of traditional sales forecasting methods, we propose a novel framework leveraging network-based metrics and Ordered Weighted Averaging (OWA) scores. By integrating helpfulness scores, review quality, spelling error checks, clustering, clustering coefficients, and PageRank, our model offers comprehensive insights into customer sentiments. This research contributes to e-commerce analytics by providing a more accurate methodology for predicting sales performance and enhancing decision-making processes. The proposed model is validated with empirical data, demonstrating its effectiveness in capturing the complex dynamics of online consumer behavior. We can also recommend suitable items with more exclusive offer to the particular users who gain high Network Promoter Score (NePS) value with positive sign, because they are a reliable and positive reviewer. They will purchase those items and influence others to buy those items with ratings and reviews. The company should also give the same focus to the users, who gain high NePS value with negative sign, because they are also reliable users. They may give negative ratings due to some dissatisfaction about the quality of the items. A company should recommend good quality items to those users based on their preferred areas. Rating based recommendation systems ignore these negative users. Ignoring detracted users is not at all good for a company’s financial health.