SPECIAL SECTION: HARNESSING THE INETERNET.

With the initial waves of hype and giddiness worn off, established businesses in every sector are discovering that the Internet and related technologies are here to stay.

While the dot-coms and their financial backers were wrong in projecting infinite growth curves -- the familiar hockey sticks -- the core belief that information networks would transform business and society still holds true. What's different now is both a willingness of enterprise managers to see this dynamic and their insistence on tangible, if unconventional, evidence of business value.

As e-commerce "comes of age" in enterprises seeking to utilize its particular advantages, several things are happening. Most importantly, businesses are moving Internet-related technologies farther and farther into the back office. As managers recognize the power of network effects and efficient information flows, customer-facing processes rely more than ever on accurate feedback loops for backup in providing customer value. If there's a Web site with on-line loan application, for example, hand-carrying the completed application form for entry into the rating engine wastes effort, introduces delays and the potential for error and generally undermines the promise of customer self-service.

Another reason why Internet technologies will include the back office reflects the benefits of real-time information, delivered via a standardized, easy-to-learn interface -- an easy sell in contrast to klunky proprietary applications, frequently run in overnight batch mode. Examples include both wireless technologies and increasingly sophisticated business-to-business commerce tools designed to address cross-enterprise coordination issues.

A case in point brings these two messages home. A number of months ago, a major bank announced a two-fold commitment to wireless technologies, targeting both internal constituencies and consumer applications. The executive in charge was surprised at the outcome: despite nearly unanimous "expert" predictions of wildfire adoption, consumer wireless services have been almost universally ignored, and the bank's customers were no exception. In contrast, and in defiance of the predictions, "wireless has been hailed by our corporate customers because it enables quick-hit connectivity for CFOs and treasurers who are always on the move. Cash management, it seems, demands considerable real-time attention (for obvious reasons)." The executive goes on to assert that "these two experiences go to show that adoption (read human behavior) often comes for the least generally expected reasons, and the most obvious opportunities often turn out to be underwhelming."

Overall, the changes wrought by the Net will continue to be substantial but rarely where we expect them to be. The evidence shows that business value often accrues first to customers, who then insist on its expansion: If I can have a catalog, what about ordering? After ordering, tell me order status and other customer service updates. After transactions, I want order history and time series data to analyze my transactional patterns. Businesses that focus only on internal cost-saving and similar payback measures tend to avoid these customer-centric benefits, often to the detriment of their external relationships.

Therefore, the best possible stance is one of flexibility and adaptability rather than tightly-controlled optimization. As the wireless example showed, most predictions have been wrong. Rather than trying to pick winners and tuning processes based on a projection, winning organizations are developing portfolios that cover a range of potential scenarios. And more importantly, they nurture cultures that accept change as opportunity, preparing for the inevitable surprises.

Also, companies that think beyond the zero-sum mentality have found that Internet technologies with apparent external benefits pay back in subtle but important ways, such as increasing switching costs and otherwise building loyalty.

In the end, taking advantage of the Internet should not be the article of faith that its most zealous advocates proclaimed, but neither is it a new stage in the history of computation as the automation of existing processes. Companies that have succeeded thus far have combined a willingness to explore unintended consequences with hard-headed insistence -- often at the project definition phase -- on incremental milestones and objectives. These have the effect of compartmentalizing the cost of learning into manageable steps rather than aiming for the unfortunately named "big bang" of predefined grand ambition.

Dr. John M. Jordan is a principal with Cap Gemini Ernst & Young Center for Business Innovation. Email him at [e-mail] or telephone [phone].