Get ready for the re-centralization of information management.

The triumph of the Internet doesn't mean that the decentralization of business computing is here to stay. Just the opposite will occur: The Internet's weaknesses will give birth to a new IT era of central network management and configuration control. When every computer is connected to everyone and employees are totally dependent on networks to get work done, executives will insist on strong central controls to contain costs and deliver expected results.

Precipitating the crisis

The year 2000 fiasco is one of them. Corporate management is also learning that tighter operational discipline will lower the total cost of ownership of PCs. If the current estimates of a savings of up to $4,000 per PC per year are roughly correct, CEOs will wonder who was minding the store while 200 million PCs were installed.

Other lessons are yet to come. Companies are now paying to replace just-installed client/server systems with supposedly faster, cheaper Internet solutions.

The infatuation with the Internet won't last. Loss of control over data, incompatible applications and the threat of security corruption will gradually end business' blind faith in the Internet as an information management cure-all. Billions spent on Web pages and Internet business ventures will eventually be wiped out.

These expensive lessons won't sit easily with top executives. They will ask why so many technology projects abort during implementation or have such a short lifespan. They will question why IS managers waste money by abandoning existing investments in hardware, software, data and training with each generation of new technology.

This time, top management won't be content with more presentations from the CIO. Short-term thinking and build-and-junk habits have become an unaffordable luxury. The 1996 IT spending for 3,110 of the largest U.S. firms now equals their total annual economic value-added. Companies don't have the money to start yet another massive replacement of systems investments inherited from prior computer technology cycles.

A backlash against wasteful IT spending will shape the coming technology cycle. CEOs will impose centralized constraints on all future IT investments, to make sure they get lasting value.

The Coming of the "Network Computer" Era

The hallmark of this era won't be a particular hardware platform or operating system. Any microprocessor will be able to run any application, as long as non-proprietary codes, languages and standards are followed. With "Pure Java," platform-independent computing, universal network connectivity and the doctrine of "write once, run anywhere and anytime," I see for the first time a glimpse that "open systems" technologies could become a compelling reality.

The economic model of Network Computing will also differ from what we have now. Vendors, consultants and CIOs concentrate almost exclusively on IT costs which are defined as what is included in the IT budget. That is a misjudgment. When one analyzes the Total Costs of Ownership of computers one finds that most of the costs that affect the productivity of people show up as operating expenses, not as IT. (see CW, July 14, 1997, Page 62). The new economic model must encompass all of the costs of information-handling, including user costs as well as gains in their productivity.

The old disputes that drove arguments favoring centralization vs. decentralization -- whether the central IS or the end-use customers possess the hardware -- will no longer be relevant when everyone has mainframe MIPS in their pocket. The principal question will be who manages the rules of network control. The CIO may not be the key player in that round. Chances are that the technological complexity, economies of scale and scarcity of talent will lead companies to outsource most of the network management tasks. But CIOs will have to make sure that outsourcing doesn't hand too much power to the contractor, and that the outsourcing agreement doesn't become a trap.

CIO Implications

The job of a CIO will not be limited to management of data processing. In the new era the scope of that job expands to cover all information-related business processes. That includes not only the automation of computationally limited business functions but most information-handling and information coordination activities. In the network computer world, IT leaders must cultivate and conserve the most costly and valuable elements of any system: data, business processes, workflow, work-enlargement, the enhancement of the employees skills and the protection of knowledge assets.

The squandering of computer resources, which has characterized the history of computing so far, is finally coming to an end. Long-term economic benefits, not short-term technology costs, will rule from now on. Whenever that prevails, the forces of centralization will reassert themselves.