An LRFM Model to Analyze Outpatient Loyalty From a Medical Center in Taiwan

Abstract

This research is intended to study the behaviors of outpatients in a medical center and constructs a set of data exploration procedures such that hospital management can deal with patient relationship management more effectively. This study adopts LRFM (length, recency, frequency, and monetary) model and cluster analysis, including self-organizing maps and K-means method, to categorize 321,908 outpatients of the medical center into 12 groups and then uses the multidimensional customer clustering philosophy to classify the outpatients. Outpatients can be categorized into five different types of groups, namely, core customer groups, potential customer groups, new customer groups, lost customer groups, and resource-consuming customer groups. In addition, seven types of outpatients based on five types of categories are identified. The similarities and differences of each group based on the patients’ characteristics are analyzed to give differentiation strategy advices for hospital management. Hospital management thus can design the optimal service strategies, provide the best care services, enhance hospital’s performance, and reduce the overall cost to establish quality relationships with outpatients.

A data examination on a supermarket dataset in Taiwan

“…The total amount is a continuous variable, whereas merchandize items belong to a categorical variable. To test if the customers with different demographic and behavioral variables, such as gender, age group, day of the week, date of transactions and time of purchase as shown in Table 1, would have different merchandize items purchased as well as total amount of money spent statistically, Chi-square test and ANOVA are performed, respectively (Chao et al. , 2021).…”

Perbandingan K-Means K-Medoids dan Fuzzy C-Means untuk Pengelompokan Data Pelanggan dengan Model LRFM

“…Mengetahui lama transaksi nasabah selama periode waktu tertentu (L/Length) penting untuk menghitung loyalitas pelanggan. Fitur model LRFM meliputi Transaction Length (L) yang menunjukkan jarak dari waktu ke waktu antara transaksi awal pelanggan dan transaksi terakhir (hari) Recent Transaction (R) yang mengukur interval antara transaksi terakhir yang dilakukan konsumen dan waktu analisis saat pengguna melakukan segmentasi (hari), Annual Frequency (F) yang menunjukkan volume khas transaksi yang dilakukan oleh klien selama periode waktu tertentu, Average Monetary Value (M) yaitu jumlah rata-rata uang yang dipertukarkan oleh konsumen selama periode tertentu [15][16] [17].…”