Kerangka Pengambilan Keputusan Untuk Pemasaran Presisi Menggunakan Metode Rfm, Algoritma K-Means Dan Decision Tree

Abstract

Precision marketing provides the ability for companies to offer products made specifically to customers and provide the ability for companies to attract customers with specially made marketing messages. This paper presents a new decisionmaking framework using data mining techniques. First, this study presents a trend model for accurately predicting monthly supply quantities; second, using the RFM (Recency, Frequency, Monetary) model to choose attributes to group customers into different groups; third, using the K-Means Algorithm to create customer clustering based on RFM data of each customer, fourth, using Decision Tree to identify attribute values that are important for distinguishing different customer groups; and finally, from the process of data mining that researchers do create various bidding strategies that target each customer cluster. Sales data from Syifamart in Subang, West Java, were collected and used in a case study to illustrate how to implement the proposed framework. The research shows that the process of data mining from sales transaction history is 351,158 rows, with customer-based aggregation using the RFM method and extracted using the k-means clustering algorithm to form 4 (four) optimal clusters. The four (four) clusters are classified using a decision tree algorithm so Syifamart can find out which potential customers and potential customers. For the availability of stock supply, management predicts the need for product inventory using a trend method where the stock in the following month is predicted by using the sales history in the previous month.Kata Kunci: Pemasaran Presisi, Metode Tren, RFM, K-Means, Decision Tree. PENDAHULUANDalam menjalankan roda bisnis, perusahaan membutuhkan strategi pemasaran yang cocok dan efektif dengan bisnis yang sedang mereka jalankan. Hal itu bertujuan untuk menarik pelanggan lebih banyak lagi serta meningkatkan penjualan

Marketing Maps Pada Lembaga Amil Zakat Menggunakan Algoritma Clustering Dan Association Rules

“…Pada penelitian tersebut menunjukan bahwa proses data mining dari history transaksi penjualan sebanyak 315.158 row dengan agresi berdasarkan pelanggan menggunakan metode RFM dan diekstrasi dengan menggunakan algoritma kmeans clustering membentuk 4 (empat) cluster yang optimal. Dari keempat (empat) cluster tersebut diklasifikasikan dengan menggunakan algoritma decision tree untuk mengetahui mana pelanggan potensial dan mana pelanggan yang tidak potensial [3].…”