Segmentation Applications in Banking - Data Mining Techniques in CRM: Inside Customer Segmentation - page 230

Table 6.1 Credit card segmentation: summary of implementation methodology.

Critical point Decision

Segmentation level Customer level

Segmentation population Active customers with at least one open and active credit

card and with monthly average purchases of at least $50
during the last six months

Data sources Mining data mart and MCIF

Time frame of used data Last six months

Segmentation fields Purchase KPIs contained in the MCIF (percentages,

averages, etc.) summarizing purchases of the last six
months

Profiling fields All fields of interest: usage KPIs of the last six months,

demographics, etc.

Technique used Clustering technique to reveal segments

THE SEGMENTATION FIELDS

The list in Table 6.2 presents the clustering inputs which were used for model
training. It also presents the profiling fields that did not participate in the model
building but only in the profiling of the revealed clusters.

Table 6.2 Fields used for the segmentation of credit card holders.

Field name Description

Purchases by merchant category - clustering fields:

PRC_ACCESSORIES The percentage of total purchase amount spent at accessory

stores (includes shoes/jewelry/cosmetics)

PRC_APPLIANCES The percentage of total purchase amount spent at electrical

appliances/electronics stores

PRC_CULTURE The percentage of total purchase amount spent at cinemas/

theaters/museums etc.

PRC_GAS The percentage of total purchase amount spent at gas stations

PRC_BOOKS_MUSIC The percentage of total purchase amount spent at bookstores and

music/DVD/video/computer games retailers

PRC_APPAREL The percentage of total purchase amount spent at clothing stores