Resumo. Artificial Intelligence (AI) has expanded the analytical capabilities of marketing by enabling organisations to process large volumes of consumer data, identify behavioural patterns, and support predictive decision-making. This study examines the application of AI and machine-learning techniques to marketing analysis through a critical review of the literature and an empirical case study based on transactional consumer data. The empirical analysis compares Logistic Regression, Random Forest, Support Vector Machine (SVM), and XGBoost using accuracy, precision, recall, F1-score, and ROC-AUC as evaluation metrics. Among the evaluated models, XGBoost achieved the strongest overall performance, with an accuracy of 83.87%, precision of 81.79%, F1-score of 79.19%, and ROC-AUC of 0.9126, while Logistic Regression achieved the highest recall, at 80.33%. The findings indicate that machine-learning methods can identify predictive associations in transactional data and support consumer classification, although such associations should not be interpreted as causal effects. The study also discusses methodological limitations related to the behavioural proxy adopted, the available predictors, and the generalisability of the findings. Overall, the results indicate the potential contribution of AI-based predictive analytics to marketing decision-making while emphasising the need for transparent, interpretable, and ethically responsible applications.