Analisis Customer Lifetime Value (Clv) Dengan Clustering Untuk Mendukung Strategi Customer Relationship Management

darvesh, muhammad arkan karindra (2025) Analisis Customer Lifetime Value (Clv) Dengan Clustering Untuk Mendukung Strategi Customer Relationship Management. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025211236-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB)

Abstract

Persaingan ketat di industri distribusi menuntut strategi yang berpusat pada pelanggan untuk meningkatkan retensi. Penelitian ini bertujuan untuk melakukan segmentasi pelanggan pada sebuah perusahaan distribusi di Surabaya berdasarkan nilai dan perilaku mereka guna merancang strategi Customer Relationship Management (CRM) yang efektif. Metode yang digunakan adalah klasterisasi berbasis model LRFM (Length, Recency, Frequency, Monetary) dari data transaksi 580 pelanggan aktif, dengan membandingkan secara komprehensif algoritma K-Means, DBSCAN, dan Hierarchical Clustering. Kualitas setiap model dievaluasi menggunakan Koefisien Siluet dan analisis distribusi klaster, sementara Fuzzy AHP digunakan untuk pembobotan kriteria. Hasil penelitian menunjukkan adanya trade-off antara performa teknis dan kegunaan bisnis; model K-Means dengan K=4 (Skor Siluet = 0.5303) dipilih sebagai model final karena menghasilkan klaster yang paling seimbang dan dapat diinterpretasikan, yang berhasil mengidentifikasi empat segmen pelanggan: Pelanggan Setia, Pelanggan Berisiko, dan Pelanggan Potensial/Baru. Berdasarkan perankingan segmen menggunakan skor berbobot, penelitian ini menghasilkan rekomendasi strategi CRM yang terdiferensiasi, memungkinkan perusahaan untuk mengalokasikan sumber daya secara lebih efisien guna meningkatkan loyalitas dan nilai seumur hidup pelanggan.
=====================================================================================================================================
Intense competition in the distribution industry demands customer-centric strategies to improve retention. This research aims to segment customers of a distribution company in Surabaya based on their value and behavior to design effective Customer Relationship Management (CRM) strategies. The methodology involves clustering based on the LRFM (Length, Recency, Frequency, Monetary) model derived from the transaction data of 580 active customers, with a comprehensive comparison of K-Means, DBSCAN, and Hierarchical Clustering algorithms. Each model's quality was evaluated using the Silhouette Coefficient and cluster distribution analysis, while Fuzzy AHP was used for criteria weighting. The results revealed a trade-off between technical performance and business utility; the K-Means model with K=4 (Silhouette Score = 0.5303) was selected as the final model for producing the most balanced and interpretable clusters, successfully identifying four customer segments: Loyal Customers, At-Risk Customers, and Potential/New Customers. Based on the ranking of segments using a weighted score, this research provides differentiated CRM strategy recommendations, enabling the company to allocate resources more efficiently to increase loyalty and customer lifetime value.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Kata kunci: Segmentasi Pelanggan, Customer Relationship Management (CRM), LRFM, K-Means, Hierarchical Clustering, DBSCAN, Fuzzy AHP. Keywords: Customer Segmentation, Customer Relationship Management (CRM), LRFM, K-Means, Hierarchical Clustering, DBSCAN, Fuzzy AHP.
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Arkan Karindra Darvesh
Date Deposited:	30 Jul 2025 09:37
Last Modified:	30 Jul 2025 09:37
URI:	http://repository.its.ac.id/id/eprint/124403

Actions (login required)

View Item