Cinema, as one of the most influential cultural and artistic products, plays a key role in shaping cultural tastes and fulfilling leisure needs. Leveraging data mining and analyzing customer touch points throughout the purchase journey provides an effective tool for understanding customer intelligence, identifying behavioral patterns, and predicting audience return rates. This study aimed to develop an analytical model based on customer intelligence to extract cinema audiences’ decision-making patterns. The research was descriptive–analytical, cross-sectional, and conducted with an exploratory mixed-methods approach. In the first stage, semi-structured interviews and content analysis identified customer retention indicators, categorized into eight themes. In the second stage, quantitative data were analyzed using a decision tree model. The qualitative sample consisted of 18 adult cinema-goers, and the quantitative sample included audiences in Gorgan cinemas. Using random sampling, 450 questionnaires were distributed, of which 419 valid responses were analyzed. Findings revealed that repeat purchase decisions result from the interaction of multiple indicators rather than a single factor. Among the 32 identified indicators, snack product pricing emerged as the most significant predictor and served as the root node of the decision tree. The model achieved an accuracy of 0.896 and generated clear if–then rules for both repeat and non-repeat purchase scenarios. Price satisfaction influenced decision-making differently depending on factors such as prior experience, film quality, payment convenience, online advertising, and staff behavior.Results confirm that predictive data mining approaches, particularly decision tree models, offer deep insights into customer intelligence, enabling targeted marketing and retention strategies that enhance prediction accuracy.