Abstract
The aim of this research was to develop a machine learning model that can predict customer churn in the consulting industry. The goal was to help businesses identify customers who might leave, allowing them to take action before it happens. To achieve this, a dataset was created by combining different data sources to reflect real-world customer behaviors. After cleaning the data and addressing issues like missing values and data imbalance, four machine learning algorithms—XGBoost, Neural Network, Random Forest, Logistic Regression, and Support Vector Machines (SVM)—were used to build the prediction models. The findings showed that XGBoost performed the best, achieving the highest accuracy (95.7%) and recall (0.99), which meant it was the most effective at predicting customers at risk of churning. The study also found that the RFM (Recency, Frequency, Monetary) score was the most significant factor influencing the churn prediction. These results demonstrate the potential of machine learning in improving customer retention by proactively identifying at-risk customers. In conclusion, this research highlights the value of using machine learning for customer churn prediction and suggests that consulting companies can use this approach to better understand customer behavior and take proactive steps to retain clients.
Library of Congress Subject Headings
Customer relations--Management--Data processing; Turnover (Business); Consulting firms--Management; Machine learning
Publication Date
1-20-2025
Document Type
Thesis
Student Type
Graduate
Degree Name
Professional Studies (MS)
Department, Program, or Center
Graduate Programs & Research
Advisor
Sanjay Modak
Advisor/Committee Member
Hammou Messatfa
Recommended Citation
Alfalasi, Mediya, "Predicting Customer Churn in B2B Consulting Services" (2025). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/12055
Campus
RIT Dubai
Plan Codes
PROFST-MS
