Too Much of a Good Thing

The integration breather is essential, but it's gone too far.

Why doesnt this stuff work together?" I first heard this back in the early 80s as a technology analyst in Chase Manhattans back office. Digital was selling VAXes to translate among IBM mainframes, and I discovered that connecting mainframes to PCs required something called IRMA, as well as cookies for the sysop.

"Why doesnt all this stuff work together?" The late 80s produced islands of incompatible PC databases, with names such as R:base, DataEase and Revelation. "Why doesnt all this stuff work together?" In the 90s, this question referred to IPX, NetBEUI, DLC, Ethernet, Token-Ring and incompatible e-mail. I hear it again today as corporations struggle to combine dissimilar systems into a unified XML and Web services framework, accessible via browser, PDA or cell phone.

Art Gillis, a veteran of gritty bank systems and their unofficial spokesman in American Banker and on the Web, calls this the decade for "making it all work together and better." "Theres nothing new on the horizon," Gillis said. I agree but only partially. Most businesses acquired an overabundance of technology over the past five years, and this integration breather is essential. But weve gone too far. The focus has shifted almost completely to integration, not innovation.

Back in the 80s, microcomputers helped Frito-Lay create tremendous competitive advantage by empowering its delivery people to superserve routes. More recently, FedEx parlayed wireless, the Net and bar coding into a system that tracks packages anywhere. Even GM clawed its way back to competitiveness via a system that integrated suppliers into its internal systems. Take a lesson from these world-class companies. Integration is important, but dont miss the opportunity created by new technology.

Whats on the horizon? Jim Louderback, editor in chief of Ziff Davis Internet, can be reached at