As marketing ecosystems grow increasingly complex, organizations are under pressure to move beyond descriptive analytics and embrace predictive models to enhance strategic marketing decisions. This study investigates the application of predictive analytics, powered by machine learning, in optimizing marketing decision-making processes. Employing a quantitative research approach, we analyzed historical and behavioral data from a digital retail platform over a 12-month period. Several machine learning techniques logistic regression, random forest, and XGBoost were used to build predictive models that estimate customer conversion probabilities and forecast campaign outcomes. The results indicate that predictive models can significantly improve the precision of strategic marketing initiatives, enabling marketers to identify high-value customer segments and allocate resources more efficiently. Compared to traditional methods, the predictive approach led to a measurable uplift in campaign effectiveness and ROI. From a managerial perspective, the study highlights how data-driven strategy and real-time insights can support agile, evidence-based decision-making in competitive markets. Academically, this research contributes to the growing field of predictive marketing analytics by demonstrating the strategic utility of machine learning techniques. In an era where consumer behavior evolves rapidly, leveraging predictive tools is no longer optional it is essential.