Data Mining, OLAP and Warehousing in the Banking Industry

With the realization that data mining and warehousing have become such integral and important issues for the development of banking organizations, there is a clear impetus to examine the impact of this technology in the industry. Using this as a basis for research, this investigation considers the application of data mining and warehousing technologies in the banking industry. Through a careful consideration of what has been written on this process, a more integral understanding of how this technology is impacting the banking industry will be elucidated.

Data Mining

Hormozi and Giles (2004) in their examination of the application of data mining to the banking industry note that in many instances, data mining can provide a notable advantage when it comes to marketing. According to these authors, "Data mining enables the organization to sort through vast amounts of customer data to target the right customers. [...] Substantial amounts of time and money can be saved if an organization knows who their customers are and are able to predict what their spending patterns will be" (p. 65). Hormozi and Giles go on to note that in the banking industry where competition can be quite tight, targeted marketing can provide the organization with a notable competitive advantage. As such, data mining clearly has implications for the development and financial success of the banking organization.

In an effort to elucidate the myriad of ways in which data mining in banking can be used for marketing, Wallace (1997) notes the case of San Diego-based Advanta Mortgage Corp. As noted by this author, Advanta Mortgage has been able to incorporate data mining for the purposes of cross-selling other products and services to its credit card customers. By identifying specific trends in customer behavior, Advanta has been able to better identify specific customers that might be interested in other programs offered by the organization. As spokesman for the organization notes, "Our strategy in leveraging and mining the data is to help customers consolidate debt and to offer them a suite of products that help them lower payments and make them more financially stable" (p. 60). Through the process the organization can offer more services to the customer while providing the best financial outcomes for both the customer and the organization.

Chye and Gerry (2002) in their examination of the manner in which data mining can be applied to the banking organization argue that that this practice can improve customer relationships between the organization and consumers. According to these authors, customer relationship management can be better managed through the application of data mining. "Data mining uses sophisticated statistical processing or artificial intelligence algorithms to discover useful trends and patterns from the extracted data. Datamining can yield important insights including prediction models and associations that can help companies understand their customers better" (p. 3). These authors go on to further argue that these objectives can be met through the application of three data mining tools: description and visualization; association and clustering; and classification and estimation (prediction)" (p. 4). Overall, data mining can enhance the relationships between the customer and the organization by providing the organization with a more integral understanding of the customer.

In order to demonstrate how data mining can be used by banks in the process of building customer relationships, Adams (2005) examines the initiatives undertaken by Provident Bank in it efforts to improve its CRM operations. As noted by Adams, Provident Bank has been able to augment its CRM operations by drawing on data from a wide range of sources. "The overall goal is streamlining what the institution does for customers when trying to win or expand relationships" (p. 31). Overall the technology employed by the organization allows for a more comprehensive picture of the customer's relationship to the organization. By acquiring this picture, the organization can improve or enhance the products and services that it offers to individual clients. Thus, customization clearly becomes an issue for the development of banking services when data mining is used from improving customer relations.

Other applications of data mining in the banking industry include the development of the profitability in the banking organization. Pallatto (2003) in his examination of Sumitomo Mitsui Bank Corporation's application of data mining technologies notes that the bank was able to "to determine what the cost of dealing with different classes of customers" (p. 14). As a result of this data, the organization was "formulate standard prices for a range of services that the bank provides" (p. 14). This data also provided the organization with a clear understanding of the overall degree of profitability that each group of the bank's customers provided. As such, the organization found that some groups that were supposed to be highly profitable for the organization were not profitable at all. In other instances, groups that had been targeted as low or no profit to the organization were found to be quite profitable overall. With this data, the organization was able to improve customer service for those groups whose value to the organization had been previously reported as low. Through this process the organization was able to extrapolate more value from its existing customer base.

Huber (2004) also reports on how banks have used data mining as a means to improve the profitability of the organization. As noted by this author, Lloyds TSB was able to save more than £20m (approximately $35.6 million USD) in 2003 by using data mining as a means to identify fraudulent credit card applications and users. Huber argues that increases in credit card fraud have prompted Lloyds TSB to develop data mining procedures that identify specific characteristics of "typical card fraudsters." Using this data, the organization was able to compare transaction patterns and activity of credit card holders to determine if their activities were indeed fraudulent. Through this process the organization was able to save millions of dollars over the course of just one year.

OLAP-Online Analytical Processing

Data on the application of data mining to the banking industry clearly shows notable uses of data mining technologies. While each of these areas is of critical concern for the development and success of the organization, researchers have recently focused on OPAL or online analytical processing. Defining this process, O'Sullivan (1996) notes: "Data mining is a type of On-Line Analytical Processes, a technique which allows multipart questions to be posed of the database, Instead of a report on revenue by branch (think of a spreadsheet grid) OLAP, also known as multidimensional processing, might report revenue by branch, subdivided by product lines and region" (p. 45). This method of data mining provides managers with different methods for looking at data that is collected by the organization. This process can provide different insights into how data can be used to improve operations in the organization.

Although data mining and OLAP appear to be somewhat similar in nature, researchers examining the differences between data mining and OLAP have noted that there are significant contrasts when the two technologies are compared. For instance Alexander (1997) in his examination of applying both data mining procedures and OLAP to the banking organization notes that, " Using data mining, you may come up with a model to find who are the most profitable customers. Then you may do more traditional OLAP analysis of that subset of data to see what the impact would be if you lost those customers, how it would affect your bottom line" (p. 61). What this effectively demonstrates is that OLAP provides the necessary tools to further analyze data for the purposes of extracting critical information about operations and customers. Through the use of OLAP the organization can further refine data collected through data mining to provide a more integral insight into how data trends will impact the organization.

In an effort to provide a clear understanding of how OLAP can impact banking operations, Alexander notes the use of OLAP at First Union Bank in Charlotte, North Carolina. According to a spokesman for the organization "we're trying to sift out customer behavior" based on the data collected from data mining protocols (p. 63). Alexander goes on to note that:

The company is shifting analysis away from products, from using predictive models to determine who might buy a product in the future based on who bought the product in the past. Now it analyzed consumer's transaction patters at an automated teller machine (ATM), for example, so it can offer that customer the products that may suit their needs and preferences (p. 63).

Through the use of OLAP, the organization has been able to identify key issues that are best suited for the examination of data. This has enabled the organization to alter the specific methods that it uses for marketing its products. The end result should be an increase in the overall success of marketing efforts undertaken by the bank.

Data Warehousing

Tully (2001) examines the use of data warehousing by the Royal Bank of Canada. As noted by this author, the Royal Bank of Canada began using data warehousing for operations and customer data beginning in 1978. Although the organization had considerable success with this system overall, Tully notes that there were some drawbacks. Specifically, this author reports that, "While this innovation provided front-line staff with segment codes, the problem was that these lent themselves to subjective interpretation-resulting in the lack of a strategic or consistent approach at a corporate level. A further drawback was that the system did not provide any opportunity for pro-activity, performance measurement or predictive modeling" (p. 12). Using the historical data that had been collected by the organization, further analysis of the data showed that problems had developed in the context of customer service. "This analysis revealed that bank clients really wanted a banking relationship in which they were well-understood, their needs were anticipated, and their business was valued-no matter where, when or how they interacted with their financial institutions" (p. 12).

Through the application of data mining and warehousing technologies, the Royal Bank of Canada was able to take existing data and extrapolate new information from the data that had previously be collected by the organization. Through the application of this technology, the Royal Bank of Canada did not have to implement new technologies or data warehousing protocols. Rather, all the organization had to do was examine the data that had been previously stored and aggregated by the organization to find the specific trends in data that would allow the organization to make necessary changes and improvements. Overall, Tully notes that the Royal Bank of Canada had considerable success with its efforts to utilize historical data from the database of the organization.

While it is quite evident that data warehouses can provide organizations with invaluable information for understanding both clients and operations, research on the development of data warehouses seems to suggest that this process can be quite daunting for many banking organizations. For instance, Khirallah (1999) in his examination of the complexity warehouses notes the challenges that faced Chase Bank in their efforts to build a data warehouse. According to this author, when then organization began efforts to build a data warehouse in the mid 1990s, it had more than 32 million customers. Further, the organization had been created as a result of a number of mergers including those with Manufacturers Hanover Chemical Bank and Chase Manhattan. As such the records and data, which had been part of these organizations had to be aggregated and complied in a data warehouse that would be usable and manageable for employees in the organization.

Overall, the process used by Chase Bank in the development of its data warehouse provided notable insights into the issues involved with development. First, Khirallah notes that the organization's experiences demonstrate the need for justifying the costs of undertaking such an endeavor. Second, the experiences of Chase Bank illuminate the risk involved with developing a data warehouse. Finally, the experiences of the organization demonstrate what insights can be gained from the use of data warehouse. In the end, it seems reasonable to argue that the information that can be collected from the data warehouse is critical for the development and success of the bank over the long-term. As such, banks that do not have this technology in place will face notable challenges when it comes to retaining a competitive edge in the banking industry.

Conclusion

When all of the information from this investigation is synthesized, it becomes evident that data mining, OLAP and data warehousing have notable implications for the development and success of the banking organization. While this research clearly demonstrates that there are a host of areas in which these technologies can be used, the current research seems to suggest that banks use this technology in the areas of operations and customer service management. Through the application of these technologies to these areas, banks can effectively improve customer relationships offer targeted marketing to increase customer use of the organization's services and track operations in the organization. For banks that are using this technology for these purposes, the insights garnered can provide the competitive edge necessary for success in an industry that is otherwise highly competitive.

Although it is quite evident that data mining, warehousing and OLAP technologies can have innumerable benefits for the development of the banking organization, it is clear that these technologies carry with them a considerable price for the organization. In the case of Chase Bank, the benefits of the data warehousing were quite significant; however the costs and challenges associated with the task of developing this technology was quite substantial for the organization. For banking organizations making the decision to adopt these technologies, the issues and challenges of technology development and implementation are of paramount concern. While current literature highlights these issues, it does not provide a comprehensive glimpse of the process and problems that can arise in the context of adopting this new technology. For this reason, it is clear that more data and research on these issues needs to be presented. If banking organizations are to successfully adopt these new technologies, all of the issues of development and implementation must be addressed.

References

Adams, J. (2005). Provident opens new mines for data. Bank Technology News, 18(9), 31. This article examines the use of data mining to improve CRM at Providence Bank.

Alexander, S. (1997). Users find tangible rewards digging into data mines. InfoWorld, 19(27), 61-62. This article provides an overview of data mining and some specific examples of its use.

Chye, K.H., & Gerry, C.K.L. (2002). Data mining and customer relationship marketing in the banking industry. Singapore Management Review, 24(2), 1-27. This article provides a review of data mining through the use of CRM to improve outcomes for banking organizations.

Hormozi, A.M., & Giles, S. (2004). Data mining: A competitive weapon for banking and retail industries. Information Systems Management, 21(2), 62-71. This article provides a general overview of data mining and the ways that is applied in both the banking and retail industries.

Huber, N. (2004, March 23). Lloyds TSB saves £20m by using monitoring software to cut fraud. Computer Weekly, 5. This article examines the methods used by Lloyds TSB to reduce credit card fraud.

Khirallah, K. (1999). Building a data warehouse at Chase Manhattan Bank. Bank Accounting & Finance, 12(3), 40-46. This article details the development of a data warehouse at Chase Bank.

O'Sullivan, O. (1996). Data warehousing-without the warehouse. ABA Banking Journal, 88(12), 42-45. This article examines the specific tools that are being used for developing data mining protocols.

Pallatto, J. (2003). Data tools expose sales opportunities. Internet World, 9(1), 12-13. This article provides a review of how various organizations are using data mining as a means to develop and grow their businesses.

Tulley, J. (2001). Establishing unique customer relations using data warehousing. Canadian Manager, 26(1), 12-13. This article explores the use of data warehousing at the Royal Bank of Canada.

Wallace. D.J. (1997). Data mining and the new marketing. Mortgage Banking, 57(5), 58-63. This article focuses on the specific marketing strategies that can be used by banks that employ data mining.