This project analyzes non-login visitors on the Labelmaster Sana website from May 2023 to January 2025 by integrating web analytics (GA4), e-commerce transactions, and Sana behavioral data. The goal is to identify visitor personas and determine behavioral and channel factors that drive conversion and revenue uplift. Using K-Means clustering, distinct visitor segments are profiled based on engagement, recency, and acquisition channels. A Decision Tree and optional XGBoost regression model are then applied to predict conversion likelihood and revenue drivers. Findings aim to inform targeted marketing strategies, optimize ad spend, and enhance website conversion effectiveness.
Watch the team present this project at 02:27:34 in the session recording here.
Keywords: Non-login visitors; web analytics; K-Means clustering; conversion prediction; behavioral segmentation; XGBoost regression; customer journey analysis; digital marketing optimization; revenue uplift; GA4 integration
Faculty Advisor
I am a full-stack data scientist, software developer, and educator with diverse backgrounds. I have thirty years of experience in applying data science in commercial software development, property and casualty insurance, and retail financial service industries. I specialize in algorithm development with an emphasis on statistical learning, machine learning, and artificial intelligence.
I co-published the book “Using Data Analysis to Improve Student Learning: Toward 100% Proficiency” in 2006. I published the e-book “A Practitioner’s Guide to Machine Learning” in 2020.
I received my Ph.D. in statistics from the University of Chicago.