Multiple linear regression, random forest, customer engagement, conversion, digital marketing
Abstract
This study explores the impact of customer engagement on conversion rates in the context of digital marketing scenarios. Different customer engagement indicators are studied in depth by integrating multiple linear regression and random forest model. A dataset containing demographic information, marketing variables, and customer engagement variables was obtained from Kaggle, and the model was constructed with customer engagement variables as explanatory variables and conversion as response variable. The results of the multiple linear regression model showed that variables other than social sharing had a significant positive effect on transformation, passed the F-test, and had no covariance or auto-correlation problems, but data normality was not fully satisfied. The random forest model was accurate and fitted well on the test set. The study shows that there are differences in the order of importance of customer engagement indicators in different models, and more accurate conclusions need to be analyzed in combination with practical application scenarios, so as to provide guidance for marketers to understand the relationship between customer engagement and conversion and formulate relevant strategies.
