Segmentasi Pelanggan Produk Digital Service Indihome Menggunakan Algoritma K-Means Berbasis Python

Abstract

Telekomunikasi Indonesia is one of the companies that prioritize customers, but there is no information about customer characteristics. In this research, an analysis of customer characteristics used as a basis for determining customer segmentation and customer profiling for digital products add on Indihome services using the K-Means Algorithm. Determination of the best number of clusters done using the Elbow method and a value of K = 3 obtained, so that customer data grouped into three segments. Customer data processing is divided into 3 simulations with the percentage of train data and test data 80% - 20%, 70% - 30% and 50% - 50%. The data used totaled 1392 records as a population where the data will used to find the characteristics of each data. Cluster evaluations carried out using the Silhouette Index, Davies Bouldin Index, and Calinski Harabasz Index methods. The results of the study show that the third simulation is the best based on cluster evaluation with 50% data train presentation and 50% data test where customer profiling is seen by analyzing the members of each cluster from the third simulation where cluster 0 has 396 customer members with a customer category that provides the biggest profit for the company, cluster 1 has members of 286 customers in the category of customers who unwittingly have great potential in providing benefits for the company, and cluster 2 has a member of 14 customers in the customer category that provides fewer benefits than the cost of providing services.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 14 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.