Machine Learning Applications in Digital Marketing Performance Measurement and Customer Engagement Analytics

This study examined the application of machine learning–oriented analytics in digital marketing performance measurement and customer engagement analytics using a quantitative, explanatory research design. The analysis was conducted within a multi-platform digital marketing environment integrating web analytics, paid media, email systems, and conversion-tracking data. A final sample of 420 valid analytical units was retained after applying inclusion and exclusion criteria. Customer engagement was operationalized as a multidimensional construct comprising engagement frequency, duration, depth, and retention behavior, while digital marketing performance was measured using continuous outcomes including revenue-based performance indicators and customer lifetime value. Analytics capability was included as an organizational-level explanatory construct reflecting the intensity and maturity of analytical tool usage. Descriptive results demonstrated substantial variability across engagement and performance constructs, with engagement frequency averaging 18.9 events per week and engagement duration averaging 4.6 minutes per session, while customer lifetime value showed a wide range with a mean of USD 1,846.3 and a standard deviation exceeding USD 1,100, indicating strong heterogeneity suitable for predictive modeling. Reliability analysis confirmed acceptable internal consistency across all multi-item constructs, with Cronbach’s alpha values ranging from 0.80 to 0.90, supporting aggregation into composite scores. Multiple regression models revealed strong explanatory power, with the revenue performance model achieving an adjusted R² of 0.46 and the customer lifetime value model achieving an adjusted R² of 0.50. Engagement depth and retention behavior emerged as the most influential predictors across models, with standardized coefficients exceeding β = 0.26 and β = 0.31 for revenue outcomes and β = 0.28 and β = 0.36 for customer lifetime value, respectively. Analytics capability demonstrated a consistent positive effect across outcomes, while engagement frequency showed significance for revenue but not for long-term value. Overall, the findings demonstrated that digital marketing performance was best explained through disaggregated engagement measurement combined with analytics capability, highlighting clear differences between drivers of short-term revenue and drivers of long-term customer value within data-driven marketing environments.