This research proposal outlines a novel mechanisms designed to optimize the ranking order of digital products, primarily focusing on movies, TV series, and video games. The aim here is to develop a more objective quality measurement for digital products, which are typically more challenging to assess compared to physical products. The proposed mechanism, named the “Double Feedback Digital Product Ranking System”, uses product’s quality and customer’s preferences to optimize products’ listed ranking. This research relies on semantic analysis and label classification to process text reviews, and on Large Language Models (LLMs) to represent customer preferences. By refining the estimation of user preferences and product quality, online marketplaces can rank products in a more justified order for each customer, and customers will have more chances to purchase suitable products, potentially enhancing collective customer satisfaction. The main hypothesis will be tested through a multiagent system simulation that models the dynamics of the marketplace, sellers, and customers, providing a comprehensive evaluation of the system’s impact on online digital product platforms.