CUSTOMER SERVICE IN banking has changed a lot over the years.

Before World War II, customer service was personal. The teller would greet customers by name.

Of course, not all banking went on at the account owner's branch. Travel, moving, long-distance business, new needs -- a lot could happen. When it did, customer service was still handled personally. Correspondent banking was a big deal and would allow little banks to split up a larger corporate credit or arrange for distant services in a far-away city.

This connection between location and personalization has been totally broken in the modern world. Employees can only be in one place at one time. But bank customers want service delivered in numerous places, at any time they choose, and through a channel of their own choosing. So, how can a bank combine good service - when, where, and how the customer wants it - with personalization?

The on-line terminal solved this problem temporarily in the 1970s. We've already forgotten how revolutionary on-line access first was. The original computer systems of the 1950s were all batch-type. All questions and calculations required complex programming and at least 24 hours turnaround time. Static data, like an account balance, existed on a giant printout somewhere, and the provider of customer service had to locate that printout.

The first wave of computerization made the customer service problem even worse. So much extra data were created in the back office that accounts and transactions got far, far more complicated than ever before. As the cost of transactions dropped, and as the economy and living standards grew, the gross volumes of transactions per acount and per person grew.

New investment options like mutual funds soaked up money but were hundreds of times more complicated in terms of how much data they required or how much information the customer needed to know about them. (No prospectus required for a passbook savings account.) The net result: escalating customer service concerns.

The banking industry's current customer service solution is the customer service workstation. These workstations are generally defined and funded by a specific line of business, which in turn is a combination of products and customer sets.

Credit card, commercial lending, cash management, trust, and investment management businesses are typical places where customer service workstations are found in banking.

The application software is usually specific to the need and highly tailored to the bank's own situation. The user of the workstation is usually a customer service representative (for retail) or a relationship manager (for wholesale.) Customer communication can be by phone, mail, electronic mail or in person, but phone predominates today.

Technologically, a workstation "system" usually includes a number of microcomputers; a variety of local servers, mainframe access, and one or more local area networks to tie it all together. The workstation generally has direct access only to the servers on the LAN. It will almost always have a graphical user interface, windows, a mouse, icons, and pull-down menus.

There's typically a data base server containing a relational data base with account and customer data extracts from the mainframe. It may also contain data from external sources or data created by the customer service department.

There would probably be a mainframe server to allow direct inquiries. There may also be a fax server, an mage server, a print server, and servers to reach other electronic networks. Anywhere from dozens to thousands of workstations might be on the network, depending on the business unit's size.

Very interesting capabilities1 can be enabled by combining these technologies:

* The latest product/service bulletins can be displayed each time a customer rep signs on. Much - possibly all - of the paper involved in customer service can be eliminated.

* Various windows can be kept open simultaneously: one for the caller's existing accounts, for example, and second for detailed data on a selected account. A third could list previous transactions.

* Other windows could contain customer information such as investment profile, demographic data, stock market data, or credit data.

* Windows with images can contain copies of any correspondence. Of course, 6-month 12-month, or 18-month statement histories plus current statement data are available. The objective is to eliminate call-backs.

* Icons can be used to generate actions that finish the transaction. There can be a fax and/or mail icon to send standard messages with text options selected by the customer service representative. A product information icon will open up to a menu of product literature options. Point, click, and the literature is automatically sent.

* Free-form comments can be kept in yet another window. Example: "Mr. and Mrs. Taylor are divorced. Give no information to spouse."

* Dynamic management information is available on service-rep productivity, queue lengths, transaction volume changes and trends, response times, adherence to qualify standards, type of product data requested, leading causes of calls and more.

* Work can be shifted electronically between different customer service rep units by usiong workflow software, which is becoming a standard component of customer service workstations.

* Training can be considerably easier. It can be built into the system, with extensive help screens, and the work load can even be dynamically linked to evolving employee skills.

Do these customer service workstations degrade customer service? Far from it. Given the complexity of modern financial products, the sheer volume of transactional data, and the need ro ubiquitous access, there really is no other way to handle it.

For example, many deposits are leaving bank accounts to enter mutual funds. As this happens, the customer service shifts from branch teller-platform environment that probably lacks a customer service workstation to a mail-telephone fund distributor whose representatives probably do have a customer service workstation. The net result: Many customers feel that companies like Fidelity Investments have better customer service than banks ever did.

Want some typical "rules of the road" for customer service workstation solutions?

* Make minimal changes to the mainframe systems. Surround those legacy systems with a steel wall and throw all the new functionality onto the new client-server system. Augment those walls with a data-permeable gate that has entry and egress.

* Reengineer the entire customer service function, including job descriptions, compensation, training, back-office fulfillment procedures, measurement systems, budgets, and so forth.

* Rebalance the channel strategy for reaching the customers. Phone access to customer service reps with workstations mow accounts for about 25% of all retail banking interactions. Too many banks are now paying for multiple channels - and their costs go up, not down.

* Be careful about how many different systems are accessed. Technologists love the aid of interoperability - the idea that any user can access any data anywhere in the bank. But the fact is, it takes training and authority to use bank data. Credit card customer reps aren't going to start magically handling stock transfer data.

Nevertheless, customer service workstations, and the technology that supports them, are here to stay. The 1970s online paradigm is still common but is inadequate for the long run. No more than a quarter of bank customer service employees have customer service workstations today. But with time, they all will.

Banks that wait another five years to implement workstations are going to be industry leggards. Exactly when to move in a strategic decision that depends on the specific line of business, the competition and the implementer's financial situation. The bottom line: Take it seriously, think ahead, and, most of all, build for the future.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.