Accurately grasping the rules of user behavior and market changes and timely adjusting decisions and strategies are the ways for brand development and innovation. In this paper, we proposed a brand user attention model based on online text review analysis. First of all, we collected and preprocessed the user comment text from the online forum. Secondly, through the LDA topic model and LDAvis visual analysis, the potential topics of user reviews were extracted, and a multi-dimensional feature analysis model was constructed to reveal the users' attention features of brand products. Finally, took the new energy automobile brands as an example, the users' attention features for the different new energy automobile brands were explored and the empirical study was carried out. This study found that the brand user attention model based on online text analysis can effectively extract the characteristics that brand users care about, obtain valuable business insight, and provide support for managers' decision-making.