Analysis of the use of machine learning algorithms for predicting consumer purchasing behavior and their impact on the effectiveness of marketing campaigns

Abstract

Abstract. Traditional approaches to analyzing consumer purchasing behavior prove to be insufficiently effective in the context of high market dynamics, demand personalization, and the multifactorial nature of influencing factors, which necessitates the application of modern analytical tools, particularly machine learning algorithms. The purpose of the study is to generalize approaches to forecasting consumer purchasing behavior based on machine learning algorithms, as well as to assess their impact on the effectiveness of marketing campaigns. The methodological basis of the study includes methods of big data analysis, machine learning, and economic and statistical modeling. The study utilizes open e-commerce datasets containing information on user behavioral characteristics, transactional operations, and marketing interactions. To achieve the research objective, classification, clustering, and association rule mining algorithms are applied. The results of the study indicate that behavioral and marketing factors, such as time spent on a website, frequency of interaction with the platform, and responsiveness to advertising stimuli, have the most significant impact on purchasing behavior, while demographic characteristics demonstrate limited explanatory power. A comparative analysis of models shows that ensemble methods provide higher predictive accuracy: the XGBoost model achieved an ROC-AUC value of 0.95, while Random Forest reached 0.90 and Logistic Regression 0.82. The evaluation of the practical effects of applying machine learning algorithms demonstrates an increase in conversion rates by 40–70%, growth in return on advertising spend by 30–60%, improvement in customer retention rates by 15–25%, and an increase in customer lifetime value by 20–40%. The obtained results confirm that the use of machine learning algorithms enhances the effectiveness of marketing strategies through the personalization of offers, optimization of communications, and more accurate targeting of consumers. The practical significance of the study lies in the possibility of applying the results for managerial decision-making in digital marketing and improving the competitiveness of enterprises.