Bankers are coming to realize what marriage counselors long have known: relationships are not easily understood.
According to the American Banker/Tower Group 1995 Survey of Technology in Banking, fewer than one in five financial institutions surveyed had the ability to quickly identify all of the products and services used by members of a particular household.
And as nonbank invaders competitors threaten to occupy even more of banks' turf, the industry is scrambling to fix situations like this one.
Observers say most institutions - and particularly the largest players - are moving aggressively to acquire the technology to know their customers better.
The systems that can help banks attain such knowledge - including relational data base management software and parallel processing computers - are expensive. But the cost of not investing is much greater, many bankers contend.
"I think you're history if you don't" invest in customer-knowledge technologies, said Jonathan J. Palmer, chief technologist and head of retail banking at Barnett Banks Inc., Jacksonville, Fla.
Mr. Palmer said that bankers clearly recognize the need for such systems.
At a recent conference, he asked a crowd of bankers how many of them worked for institutions investing in so-called "data warehouses," where vast amounts of client and demographic data are stored and mined. Virtually everyone raised a hand.
Unfortunately, the solution to the customer knowledge problem requires much more than merely recognizing the need for such systems. And bankers historically have been slow to act on this front.
"In the 15 or 20 years I've been working with banks, they've been wringing their hands over this same set of issues without making much progress," said Fred Reichheld, a director with the consulting firm Bain & Co.
"Most banks have not put this at the absolute top of the priority list. It's always (been) third or fourth or fifth, and it never seems to get done."
Mr. Palmer and others insist that is changing. Part of the change is being driven by technology advances. Data storage, processing, and communications tools are getting ever-cheaper, making it more cost effective for banks to mine the lode of available customer data.
A more significant catalyst, however, is the crush of nonbank competition in recent years.
As has been well documented, nonbanks are garnering an increasing "share of wallet" from consumers at banks' expense. The main reason for this is clear, according to Mr. Reichheld: "Banks are a bad relative value," he said. Bank response to nonbank competition thus far has been to consolidate and reduce costs, but "that's not likely to cut it much longer."
In order to please customers, many banks, including Barnett, BankAmerica Corp., and Banc One Corp., are moving customer data to a single repository - a data warehouse - that can provide branch and customer service employees with comprehensive customer relationship information.
The goal, from a customer service perspective, is to be able to satisfy inquiries about various accounts through a single point of contact.
At many institutions today, a customer seeking information about a mortgage loan, a credit card account, or a checking account has to contact different business units that handle these products.
Giving customers a single point of contact would, at the very least, allow banks to equal the level of service that consumers have come to expect from their contact with insurance and investment companies.
Consolidating customer account information has benefits beyond customer satisfaction. One of these is improved marketing and sales.
With a single silo of data, banks can identify cross-selling opportunities more readily than if information is housed in "islands of information," observers said. For financial institutions with such islands, pulling information from a variety of applications is time consuming and typically allows for only broad market segmentation and product pricing.
On top of the challenge of building data warehouses, banks are faced with the task of putting their mountains of transaction data to better use.
According to the survey, most institutions are failing in this regard. For instance, only about one-third of the institutions surveyed have the ability to quickly react to a customer closing one of several accounts.
A system that responds well to such an event would tell the bank representative closing the account whether steps should be taken to ensure that the customer was satisfied with the rest of his or her bank relationships.
By tracking transaction activity more closely, banks can also determine which offices are the most important to particular customers. When this information is combined with analysis of customer profitability, a bank can determine the offices or delivery channels that should get the most attention and investment.
"Understanding your customer is going to be the key to attracting, retaining, and getting the biggest share of wallet from profitable customers," said Rockwell F. Clancy, managing director of retail banking at the Bank Administration Institute in Chicago.
Such information is also necessary for banks that want "to prioritize and allocate resources in some sort of logical manner."
Though the benefits of technologies that allow banks to know their customers better are many, bank investment in these technologies is hardly a foregone conclusion for all institutions.
Small and midsize banks may feel that the benefits of such systems are too intangible to cost-justify. And even the largest institutions are likely to think twice before committing the millions of dollars necessary to install data warehouse systems and the related parallel processing computer technology.
"The payoff of these systems has not been proven to a lot of senior bankers, and I think that's the nut to crack for many institutions," said Terence B. Freeman, a managing consultant in Towers Perrin's Atlanta office.
But once a commitment to customer knowledge systems has been made, banks must train employees how to use the information.
On the corporate side of the banking business, bank representatives already are well attuned to picking out and coddling their best customers. But on the retail level - particularly in the branches - looking for the most profitable customers and selling them new products are largely foreign concepts.
Consultants and bankers alike emphasized that in order for the technology to have its desired effect, retail personnel need to be given incentives, such as sales commissions.
"You are asking a person to change from a customer service/transaction problem solver to a seller," said Mr. Freeman. "This transition does not just happen."
In addition, it is important not to overwhelm employees who deal with the public with customer information. A good system will not simply spew every bit of data on a particular customer. Rather it should provide bank representatives with a few suggestions for new product sales and a thumbnail sketch of a customer's relationship.
Deep historical data should be available, if needed.
"We need to get a whole lot more precise in the way we target our offerings to the market," said Barnett's Mr. Palmer. "We need to understand more than intuitively what our customers are doing with us."