Designing Customer Target Recommendation System Using K-Means Clustering Method

Abstract

UD Swiss is a company engaged in the field of goods distribution located in Cirebon. In achieving sales targets, customer marketing department sets customer targets to be visited based on the type and location of outlets. However, the method of targeting customers does not achieve the sales target yet due to the differences in the characteristics of purchases per product category for each type of outlet. The research in this paper focuses on the analysis and implementation of management information system to help the company gain knowledge in targeting customers based on the profile and characteristics of each customer group in doing transactions. The information system is made to load each of the knowledge generated by the analysis of customers’ characteristic using the k-means clustering. The system is designed to use the programming language “Groovy and Grails” and is built using the .NET Framework that can run on the Java platform with support of PostgreSQL as a database. Grouping customers using k-means clustering method generates groups of potential customers who are considered to be the target in the process of product sales. Customers who have an average purchase at least Rp 2,028,813.00 per transaction with the minimum purchase frequency of 25 transactions a year is a potential customer.