Understanding public opinions and sentiments related to blood donation is crucial for improvement of the services and experiences provided by the blood donation centers in Japan. Social media platforms such as ‘X’ (formerly Twitter) offer a wealth of real-time data that can be analyzed to gain insight into these behaviors. However, analyzing massive amounts of data manually require unfeasible resources. This study explored the feasibility of using BERTopic modeling to process Japanese ‘X’ data and identify relevant blood donation topics. Building on prior research that created sentiment- and role-classified datasets of Japanese tweets on blood donation and compared various topic modeling methods, this paper delves into implementing the most promising analysis method. The goal of this study was to optimize the performance of the BERTopic model by tuning the parameters and customizing the BERTopic layers to better capture the nuances of Japanese ‘X’ data. Key topics identified included “donation requests”, “festivity campaigns”, and “health considerations”, which are similar to topics commonly found in blood donation in other countries. We also identified country-specific topics, such as “technology benefits”, “prefecture campaigns”, and “collaboration controversy”. Data categorization also allowed validation of more specific results when processing the data by specific groups, such as “food and beverage incentives” for donors, “enjoyment of participation” for potential donors, and “hospital interaction” for deferred donors. This study highlights the potential of advanced topic modeling techniques, particularly BERTopic, for analyzing Japanese ‘X’ data to gain public health insight. The proposed implementation closely captured the public sentiment and provided insight into factors influencing blood donation behavior, highlighting both common concerns and unique cultural aspects of the Japanese context. This approach reduces the amount of manual work required for analyses, and proposes ‘X’ data as a potential source of information for targeted interventions and improvements in the reception of blood donation campaigns. The methodology developed can be extended to other applications of social media data analysis, providing a robust framework for understanding public sentiments in various contexts.