Telecom Customer’s Segmentation Using Decision Tree to Increase Active Electronic Money Subscribers

Abstract

The ABC telecommunication company as one of electronic money providers has more than 100 million customers. If it is compared to the number of electronic money customers which have growing potential. In December 2017, the number of electronic customers owned by ABC was on the third rank of electronic money ownership. Some efforts to increase the number of electronic money customers have been conducted but it has not achieved the expected target. Based on the aforementioned problem so that identifying future customers from potential telecommunication customers to be active electronic customers thus campaign activity can be done effectively and controllably. Therefore, a customer predicting model is needed to predict potential customers to be active electronic customers. This research creates model which can be used to predict future customers using telecommunication transaction act at ABC Company. The analysis used was telecommunication transaction data for all electronic money customers with 32 variables. Those variables were formed from variables such as voice, SMS and internet usage including other forming transaction such as customer's dominant location, operating system from device and device type used by customers. Forming method model used decision tree with accuracy (ACC) measuring evaluation, positive prediction value (PPV), negative prediction value (NPV), true positive rate (TPR) and true negative rate (TNR). Based on the evaluation result, this model can predict the future customers who will be the active electronic customers for 54,09%.

Analysis of Demography, Psychograph and Behavioral Aspects of Telecom Customers Using Predictive Analytics to Increase Voice Package Sales

“…This is in line with the research conducted by Fudin (2018) which produced 79.7% accuracy in the purchasing prediction model using data mining to predict Telkomsel's prepaid customers based on five significant variables that influence customers to purchase data packages on My Telkomsel. Meanwhile, Noor (2018) found accuracy of 54.09%. It can be inferred that buying it or not, the calculation generated a higher value compared to previous research and the accuracy in this research shows the precision in predicting customers who would purchase voice packages.…”