By exploring a cluster analysis, a company can segment their customers to create more effective marketing strategies. They can also design products to fit each segment of customers. Cluster analysis involves placing similar objects into groups. In this project, a cluster analysis was completed on the customer portfolio of a community bank. Community banks tend to have fewer resources than large national banks, and they achieve goals by creating relationships with their customers to help spur economic growth within their community. This cluster analysis examined both demographic and account activity information of active customers throughout the year 2013. Using the bank’s database, a query was written to obtain 22 variables portraying 31,108 individual customers. Once the data was gathered, necessary variables were transformed and normalized in order to bring the variables closer to normal distribution, as well as provide equal weight to each variable during the analysis. After exploring multiple options, the clustering method applied was hierarchical clustering using Ward’s minimum variance criterion with Euclidean distances. As a result, five clusters were formed. Key variables in the analysis were found to be average balance and number of transactions for different account types. Cluster A was characterized by customers with high savings balances and number of transactions. Customers in Cluster B can be described as having high loan balances and number of transactions. If a customer has high CD balance and number of transactions, that customer tends to belong in Cluster C. Cluster D can be distinguished by customers who participate in online banking and receive e-statements. Lastly, Cluster E consists of customers with characteristics of a checking account, but little to no other products. These five clusters were validated by comparing multiple random selections of the dataset. It was seen that this cluster analysis provides insight into the different customer types found within a community bank. Because product types characterize the clusters, the customer segments are product driven. From this analysis, the marketing and retail departments in the bank can better align their strategies based upon the customer segments they are working with.
","abstract_html":"<p>By exploring a cluster analysis, a company can segment their customers to create more effective marketing strategies. They can also design products to fit each segment of customers. Cluster analysis involves placing similar objects into groups. In this project, a cluster analysis was completed on the customer portfolio of a community bank. Community banks tend to have fewer resources than large national banks, and they achieve goals by creating relationships with their customers to help spur economic growth within their community. This cluster analysis examined both demographic and account activity information of active customers throughout the year 2013. Using the bank’s database, a query was written to obtain 22 variables portraying 31,108 individual customers. Once the data was gathered, necessary variables were transformed and normalized in order to bring the variables closer to normal distribution, as well as provide equal weight to each variable during the analysis. After exploring multiple options, the clustering method applied was hierarchical clustering using Ward’s minimum variance criterion with Euclidean distances. As a result, five clusters were formed. Key variables in the analysis were found to be average balance and number of transactions for different account types. Cluster A was characterized by customers with high savings balances and number of transactions. Customers in Cluster B can be described as having high loan balances and number of transactions. If a customer has high CD balance and number of transactions, that customer tends to belong in Cluster C. Cluster D can be distinguished by customers who participate in online banking and receive e-statements. Lastly, Cluster E consists of customers with characteristics of a checking account, but little to no other products. These five clusters were validated by comparing multiple random selections of the dataset. It was seen that this cluster analysis provides insight into the different customer types found within a community bank. Because product types characterize the clusters, the customer segments are product driven. From this analysis, the marketing and retail departments in the bank can better align their strategies based upon the customer segments they are working with.</p>","abstract_has_math":false,"creators":["Cundy, Lance"],"institution":null,"degree_name":"Master of Science (MS)","degree_level":"Thesis - University Access Only","degree_discipline":"Mathematics and Statistics","degree_department":null,"school":null,"contributors":["Xijin Ge"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2014,"date_issued":"2014-01-01T08:00:00Z","date_published":"2014-01-01T08:00:00Z","updated_at":"2026-07-24T04:29:15Z","subjects":["Mathematics"],"languages":["en"],"rights":["In Copyright - Non-Commercial Use Permitted
http://rightsstatements.org/vocab/InC-NC/1.0/
In Copyright - Non-Commercial Use Permitted
http://rightsstatements.org/vocab/InC-NC/1.0/
By exploring a cluster analysis, a company can segment their customers to create more effective marketing strategies. They can also design products to fit each segment of customers. Cluster analysis involves placing similar objects into groups. In this project, a cluster analysis was completed on the customer portfolio of a community bank. Community banks tend to have fewer resources than large national banks, and they achieve goals by creating relationships with their customers to help spur economic growth within their community. This cluster analysis examined both demographic and account activity information of active customers throughout the year 2013. Using the bank’s database, a query was written to obtain 22 variables portraying 31,108 individual customers. Once the data was gathered, necessary variables were transformed and normalized in order to bring the variables closer to normal distribution, as well as provide equal weight to each variable during the analysis. After exploring multiple options, the clustering method applied was hierarchical clustering using Ward’s minimum variance criterion with Euclidean distances. As a result, five clusters were formed. Key variables in the analysis were found to be average balance and number of transactions for different account types. Cluster A was characterized by customers with high savings balances and number of transactions. Customers in Cluster B can be described as having high loan balances and number of transactions. If a customer has high CD balance and number of transactions, that customer tends to belong in Cluster C. Cluster D can be distinguished by customers who participate in online banking and receive e-statements. Lastly, Cluster E consists of customers with characteristics of a checking account, but little to no other products. These five clusters were validated by comparing multiple random selections of the dataset. It was seen that this cluster analysis provides insight into the different customer types found within a community bank. Because product types characterize the clusters, the customer segments are product driven. From this analysis, the marketing and retail departments in the bank can better align their strategies based upon the customer segments they are working with.
"]},{"key":"dc:title","label":"Title","values":["Customer Portfolio Cluster Analysis in a Community Bank"]}]}],"canonical_facts":{"dc:contributor":["Xijin Ge"],"dc:creator":["Cundy, Lance"],"dc:date.available":["2017-08-14T07:00:00Z"],"dc:description.abstract":["By exploring a cluster analysis, a company can segment their customers to create more effective marketing strategies. They can also design products to fit each segment of customers. Cluster analysis involves placing similar objects into groups. In this project, a cluster analysis was completed on the customer portfolio of a community bank. Community banks tend to have fewer resources than large national banks, and they achieve goals by creating relationships with their customers to help spur economic growth within their community. This cluster analysis examined both demographic and account activity information of active customers throughout the year 2013. Using the bank’s database, a query was written to obtain 22 variables portraying 31,108 individual customers. Once the data was gathered, necessary variables were transformed and normalized in order to bring the variables closer to normal distribution, as well as provide equal weight to each variable during the analysis. After exploring multiple options, the clustering method applied was hierarchical clustering using Ward’s minimum variance criterion with Euclidean distances. As a result, five clusters were formed. Key variables in the analysis were found to be average balance and number of transactions for different account types. Cluster A was characterized by customers with high savings balances and number of transactions. Customers in Cluster B can be described as having high loan balances and number of transactions. If a customer has high CD balance and number of transactions, that customer tends to belong in Cluster C. Cluster D can be distinguished by customers who participate in online banking and receive e-statements. Lastly, Cluster E consists of customers with characteristics of a checking account, but little to no other products. These five clusters were validated by comparing multiple random selections of the dataset. It was seen that this cluster analysis provides insight into the different customer types found within a community bank. Because product types characterize the clusters, the customer segments are product driven. From this analysis, the marketing and retail departments in the bank can better align their strategies based upon the customer segments they are working with.
"],"dc:identifier":["https://openprairie.sdstate.edu/etd/1557"],"dc:language":["en"],"dc:rights":["In Copyright - Non-Commercial Use Permitted
http://rightsstatements.org/vocab/InC-NC/1.0/