Abstract

This research aims to determine customer segmentation based on characteristics of credit status customers in BMT EL Bummi 372. The aim of this segmentation to identify a potential customer by loan status stalledwhich can be minimized. This research used secondary data, which consist of 94 customers withunpaid status and 851 customers withpaid status. One of method market segmentation is Improved CHAID (Chi-Squared Automatic Interaction Detection analysis) which improved contingency table by Tschuprow’s T. Improved CHAID will produce the decision tree which provide information on degrees relations between variables dependent on variables independent and information about characteristic segments. In BMT El Bummi 372 case, there are six segments were identified. Relationship segment which haseducation (elementary school, junior high school, and senior high school), less than equal 50 years, family size less than equal 3 persons, and monthly income more than Rp 1.500.000,00 is the lowest ratio of unpaid credit segment.

Authors and Affiliations