Customer Segmentation by using RFM Model and K-Mean Clustering in PT XYZ R. Hafidhullah Zakariyya
Universitas Telkom
Abstract
This research was aimed at conducting customer segmentation by applying Recency, Frequency, Monetary (RFM) model. Segmentation was administered by K-Mean Clustering, and clustering was carried out by utilizing Elbow Method. The data, it is about 15.545 customer data, was obtained from PT XYZ which running accommodation service business. Through Elbow method, the finding of the study revealed that there were three clusters. First cluster was marked as R↑ F↓ M↓, meaning that its recency was considered high, while frequency and monetary were considered low. This first cluster was translated as a ‘new customer’. Cluster two was marked as R↓ F↓ M↓, meaning that all attributes from RFM is considered low. Hence, this cluster can be categorized as ‘lost customer’. Meanwhile, Cluster three was the best cluster and categorised as ‘loyal customer’. This cluster was marked as R↑ F↑ M↑.
