The topic of data mining in the banking industry is based on my previous 12 years experience at a large local bank in Syracuse, NY. Over those years I was exposed to some of the industries most advanced technologies as well as some of its least. Notoriously slow to keep up with technology advances, banks have been forced to address this issue more for their survival than productivity. Only recently have banks been making the investment in technology to replace systems that in some cases are well over 20 years old.

Banking, or some variation, is one of the oldest and simplest businesses in the world. They take in deposits and give out loans. Their income has generally been made from the spread in interest rates received on loans minus rates paid to depositors. The past ten years or so have seen an enormous change in the way banks have been forced to do business. Bank consolidation, the advent of the Automated Teller Machine, and now the internet are all issues that banks must concentrate in order to survive. What used to be one of the most secure places to be employed is now becoming one the most insecure.

In order to stay competitive, banks have realized that they must increase market share and income to remain independent and not be gobbled up by a major bank holding company. To this end they have increased expenditures over the last five years to improve their computer systems. The process began by first implementing the use of PCs to facilitate routine and redundant office tasks and some financial forecasting and planning. This led the way towards networking of these PCs to allow shared information more easily accessable. These advances have increased productivity of the individual employee, but have not had a large impact on increasing business opportunities.

To increase business banks have decided that they must finally more thoroughly understand the needs and wants of their customers. While they have been able to review statistics on previous earnings performance and market penetration, they(majority of banks) do not use or use effectively any tools of discovery or prediction. This is changing in some of the largest banks in the country(ie Nations Bank, Chase, Bank America).

Recently there has been a trend to try to better understand the spending habits of customers through the use of credit cards. These transactions hold an extremely large amount of detailed information about each customer. Databases containing this information are extremely large and require a considerable amount of time to process traditional SQL queries. The use of parallel systems in banking is again something very rare, if at all(unable to locate any information on banks using this technology). With the consolidation of more and more banks producing larger and larger databases of customer information, it appears that parallel systems would be the only option to accomodate these databases with any kind of effectiveness and timely results.

With the exception of the technology industry itself, banking has shown to be one of the most competitive and volatile industries in the US and over the long term, in the world. You either keep up and lead the way, or you perish. Parallel processing seems to be one of the few ways banks will be able to compete going forward.