This study utilized Machine Learning (ML) models to analyze and predict customer churns in the context of webbased collaboration platform. Secondary data was used in the study, containing 109,740 observations and 12 predictors. This was divided into test dataset and out-of-time (OOT) dataset where the prior was used for model fitting and the latter is to test performance stability in unseen data. Furthermore, Synthetic Minority Over-sampling Technique (SMOTE) was performed to resolve the data imbalance, hence, preventing bias and distortion in the models' performance. These 3 ML models were assessed based on Accuracy, ROC-AUC, Precision, Recall and F1-Score. Given the business context's applicability, F1-score and Accuracy were used as bases for performance, leading to the selection of Decision Tree Classifier as the ML model in this study, with Accuracy of 92.1% and F1-Score of 63%. Furthermore, hyper-parameter tuning was performed on Decision Tree Classifier to prevent overfitting. To reinforce the model selected, Survival Analysis was implemented, specifically, Kaplan-Meier (KM) Estimator and Cox Proportional Hazard (CPH) were utilized to analyze the rate and timeframe of disengagement to the platform, revealing that beyond 72 months, it was projected to retain only 60% of its user base. Hence, these multidimensional results and insights derived from both Decision Tree Classifier and Survival Analysis were anchored in the formulation of customer retention strategy, proactively target customers who are predicted to churn.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.