Prediksi Pelanggan Churn pada Perusahaan Ekspedisi Menggunakan Algoritma Regresi Logistik

Abstract

The freight forwarding industry in Indonesia has experienced rapid growth, driven by the rise of online shopping and technological advancement. The increasingly fierce competition in this sector pushes logistics companies such as PT. Tiki Jalur Nugraha Ekakurir (JNE) to continuously innovate and improve service quality in order to maintain customer loyalty. One of the main challenges faced is customer churn, where customers stop using the service. This study aims to predict customers who are likely to churn using the Logistic Regression algorithm. Customer transaction data collected from a JNE branch office serves as the foundation for developing the predictive model. The research process includes data collection, preprocessing, data splitting into training and testing sets, model training, and evaluation using metrics such as accuracy, precision, recall, and F1-score. The results of this study are expected to assist the company in identifying churn-prone customers early, enabling timely preventive actions to retain them. Thus, this research contributes to supporting data-driven business strategies to enhance competitiveness and ensure company sustainability.