Segmentation Applications in Banking - Data Mining Techniques in CRM: Inside Customer Segmentation - page 281

Table 6.18 Clusters and product ownership.

Product ownership

Clusters Total Time Investments Mortgage Consumer Credit

savings deposits (%) loans (%) loans (%) cards

(%) (%) (%)

Cluster 1 70 12 19 4 8 18

Cluster 2 97 1 5 1 9 26

Cluster 3 61 1 4 3 89 64

Cluster 4 97 8 9 27 30 40

Total 80 4 8 8 38 39

Table 6.19 Clusters and transactional patterns.

Transactional patterns

Clusters Number of Credit Number of Debit Total credit

credit transactions debit transactions card spending

transactions amount transactions amount amount

Cluster 1 1 16.828 1 472 177

Cluster 2 1 1.552 19 7.472 240

Cluster 3 3 4.357 10 4.381 1.001

Cluster 4 6 13.314 5 6.514 288

Total 3 7.977 9 4.985 470

Nevertheless, their penetration in consumer loans is the lowest. On the other
hand, cluster 2 is associated with savings. Almost all customers assigned to cluster
2 hold a savings account but few of them hold any other product. Cluster 4 is
also characterized by savings accounts. But this cluster also presents an increased
percentage of mortgage loans and time deposit owners. Finally, cluster 3 has the
highest proportion of borrowers. Almost 90% of its members have a consumer loan
with the bank and almost 70% are credit card holders.
The revealed clusters also show diverse transactional patterns, as given in
Table 6.19.
Cluster 1 has the highest average volume (amount) of credit transactions,
although the respective frequency is rather low. On the contrary, cluster 2
customers are the heaviest and most frequent debit transactions customers. As we
will see in the next table, most of them have their salary deposited in the bank,
thus they probably have to make regular withdrawals from their payroll savings