DECISION MAKING WITH TOPSIS AND K-MEDOIDS: AN APPLICATION ON CUSTOMER RECOVERY WITH R PROGRAMMING LANGUAGE

Large companies prefer the most profitable customers when they try to recover their old customers who left the company and acquired by the competitors. This choice requires many variables to be evaluated simultaneously under the constraints of cost and time. TOPSIS method is a decision- making technique for multi-criteria problems and is used in a wide range of applications such as: supplier and facility location selection, production systems, enterprise resource planning and marketing, management, health, safety and environmental management problems etc. In this study, data of 1145 customers such as usage of voice, data service and other value added services usage are analyzed by TOPSIS method. The customers analyzed in this study are churned by a competitor of a telecommunication company. In order to determine the most valuable customers to company’s profit,TOPSIS scores are clustered into four segments by k-medoids clustering algorithm. As a result of the analysis, customers are segmented into four groups and totally 37 customers which were placed in the third segment identified as golden customers, thus the company should start churn activities with the customers placing in this segmen