Client-server computing is spreading like wildfire.

If this is news to the people at the top of financial institutions who don't pay day-to-day attention to computer trends, it is only a matter of time before they will have to start making sense of client-server.

It has all the appearances of a revolution going far beyond information processing and to the very core of the way they run their businesses.

A product of the continuing miniaturization and flexibility of computer power, client-server represents a simple but radical change in the organization and processing of information. The top-down transmission of data from central mainframes to outlying peripherals, the historical model on which most back-office banking functions were designed, gives way to networking and sharing among clusters of processors.

The central "servers" of a network, and the "client" computers on it, can pool their "intelligence" resources to accomplish more tasks faster and more efficiently than in a rigid mainframe configuration. The improvements, some have found, can be by many orders of magnitude, in areas ranging from basic transaction processing to management information and accounting.

This prospect was enough for historically conservative institutions, like some of those covered in this special report, to take notice of client-server and begin investing in its possibilities.

In some cases, a complex financial service organization like Household International will already have made a corporatewide commitment to client- server. More often, companies are taking a piecemeal or building-block approach. As the conversions' complexity and strategic importance come to the fore, vendors like Alltel Information Services, AT&T Global Information Solutions, and Electronic Data Systems Corp. have developed staffs and expertise to help their customers along.

There is enough such activity in the information systems departments of banks and other corporations to suggest that client-server is only beginning what could be the longest run of any post-mainframe-era trend.

It may not be a trend at all, but a lasting change in the way organizations work with data.

Even people who are skeptical of any sort of computer-industry hype embrace the principles of client-server.

Steve Williams of M One Inc. in Phoenix, a consulting firm that says it looks after the interests of chief executive officers in their technology purchase decisions, contends there is "not a lot of beef" in some vendors' recent product announcements.

"We see prototypes that create the illusion of a nice system that can accommodate important banking tasks, but they are for the most part just that - prototypes," Mr. Williams said. "These are very complex environments in which to make this kind of system change, and people will be working long and hard to make it happen."

Still, Mr. Williams said, "There is no reason to resist the move to client-server. Real productivity gains will come from the dramatic shift to these types of networks.

"This comes at an especially important time for banks," he added. "After all the cost-cutting they've done, the only way left for them to improve profitability is through productivity and customer service."

What is being said to promote or justify client-server might sound familiar in light of past computer-industry touts:

* Computers are getting smaller and cheaper.

* Their increased horsepower is compounded through networking.

* The increasing openness and interconnectivity of computer architectures will not only transform data processing, but support the more fundamental requirements of reorganization and reengineering.

Because many earlier computer-age promises never panned out, those who are pushing client-server - technology companies and their converts in the user community - bear a burden of proof.

Bank chief executive officers and their business-sector chiefs have a long history of mistrusting their information systems colleagues. The latter were often viewed as out of touch with business issues and out of line when presenting their cost-benefit analyses.

Many of those tensions subsided as top executives gained computer- literacy and admitted chief information officers into senior management councils. But doubts linger.

CEOs are well aware that their huge investments in personal computers before client-server were less than perfectly coordinated. And amid all the talk about client-server, they raise logical and legitimate questions about the ultimate returns on their multimillion-dollar mainframe expenditures.

John Haley, vice president of Alltel Information Services (formerly Systematics) in Little Rock, Ark., is one of the client-server experts taking responsibility for industry education as well as some of the implementation.

He takes a measured view of this putative revolution, pointing out there is "no conceptual definition separating client-server from distributed processing," an earlier approach that turned out to be less permanent and widespread than salespeople led buyers to expect.

Mr. Haley also would not force client-server into mainframe realms that do fine without it.

"Mainframes still do some things very well that are difficult to do in other ways, such as check processing," Mr. Haley said.

"The benefit of client-server will be greatest not in the back-office factories" where cost-reduction is the principal aim, he said.

"The real future is in automating processes that were seen as messy in the past - work flow applications that might involve breaking down organizational walls," Mr. Haley said.

In other words, another buzzphrase of the moment: "business process reengineering."

"The key is no longer to look at automation as a series of computer applications, but at the actual business processes," Mr. Haley said.

In addition to reegineering, he cited information access and organizational integration and communication as areas ripe for client- server.

"Just having E-mail in some cases will pay the freight for client-server implementation," Mr. Haley said.

But will that kind of assertion sell the top level of a financial institution on client-server? And has client-server achieved the status of a strategic issue worthy of, and getting, the CEO's attention and direction?

"Not many managers understand (client-server), and the technology people who do understand it don't understand the business implications," said Charles Wendel, a New York-based consultant, repeating the familiar refrain.

"The heads of commercial banking, middle-market, and small-business banking should be aware of it and establish liaisons with the technical people to make sure it serves the needs of the business," Mr. Wendel said.

"Interest is at a high level in banks," Mr. Haley said, but the resulting commitments are mixed.

"Some make a major commitment to client-server because they see it as a strategic advantage," he said. "But some make client-server investments without a clear and detailed idea of what they want to achieve.

"Some of the people we talk to are clearly focused on the benefits, and we try to help those who are not."

"Client-server is a fact of life and it is becoming more powerful," said Mr. Wendel of Mercer Management Consulting, who has witnessed it mainly in corporate and mortgage banking applications, including officers' doing business remotely on laptop computers that they hook up periodically to servers.

"What they are doing is elementary, in terms of accessing the server," Mr. Wendel said. "The frustration comes in when the server goes down, and that's still a frequent occurrence."

Mr. Williams of M One said demand is high at "operating levels" of banks, which are confident they can overcome the inherent reliability and security problems of client-server networks.

He also pointed to at least two factors that get CEO level attention: the status of their legacy mainframe investments, and the fact that client- server - with all its benefits and pitfalls - moves the personal computer phenomenon in the direction of "mission-critical" applications.

The mission-critical aspect, he said, is what differentiates client- server from the proliferation of PCs over the last decade.

"Branch automation and mortgage-loan origination are two of the areas seen as critical," Mr. Williams said.

"We have to get the top-level managers to recognize the need for strategy and getting all this to work together. It's a very complex process.

"People seem to be attracted to Windows software and point-and-click interfaces, but they don't see the whole picture," Mr. Williams added. "Networks go down, and people have to think about what that means for mission-critical banking applications."

He also warned of "system anarchy" if operating divisions are not well coordinated in their choice of client-server tools.

The spread of open systems and enterprise networks has spawned useful support products. Sun Microsystems Inc.'s SunSoft unit earlier this year introduced Solstice, designed to manage networked computers with the ease of traditional mainframe "glass houses."

"It puts client-server in perspective to note that it has been hyped for five years, and companies are still making leading-edge, multivendor announcements," Mr. Haley said. "We think it's time to move aggressively into client-server, but this underlines how necessary it is to be realistic."

Also putting client-server in perspective was a recent comment by Jack Wagner, chief information officer of Dime Bancorp in New York. A merger waylaid plans to migrate off the mainframe, and Mr. Wagner concluded, "There is no way to have a client-server system maintain a $20 billion bank."

He said the bank may be ready for the technology, or vice versa, in about five years.

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.