Then Y2K hit and the dot-com bubble burst soon after, followed by 9/11 and a worldwide economic slowdown that forced companies to rein in what many saw as out-of-control IT spending.
Today, we are coming full circle again. More and more, with the advent of service orientated architectures (SOA) and broadband networks, companies are looking to re-centralize IT.
“Everything runs in a cycle,” said Sun’s CIO Bill Vass. “If you want to predict the future, just look at the past. I mean how many times are we going to invade Russia in the winter before we learn?”
As such, today re-centralization efforts are not an attempt to go back to the “good old days” of green screens and massive computer rooms, but an effort to get the most flexibility and cost-effectiveness out of the old and existing technologies while incorporating the best of the new.
IT is a much different animal today than it was even five years ago, when most people were just learning about the Internet and the World Wide Web.
The build-out of broadband networks that span the globe, better server designs (blades, for example), virtualization technologies that separate applications from the hardware that runs them, network attached storage and storage area networks, continuously decreasing hardware costs, etc. are all making it possible to run IT from a centralized location yet let people feel they are having a local experience.
The reasons are many, but the main driver is cost savings. Better governance, forced on many IT departments because of government regulations, also is a factor. So is the age-old challenge of IT/business alignment. Both of these efforts are aided by centralized IT.
And, unlike the old days, centralizing IT doesn’t mean making it more rigid. New, automated middleware monitoring and provisioning tools, methodologies (like ITSM), and architectures (i.e., SOA) are enabling IT to become more and more flexible even as it is consolidated into fewer and fewer data centers catering to more users.
“Catering to the internal customer in the future won’t be about doing what he wants,” said Rich Melnicoff, a managing partner in Accenture’s Strategic IT Effectiveness practice, “It will be about balancing flexibility and autonomy with some centralization; (the enterprise will) get more value in the long run.”
One of Melnicoff’s clients, for example, is saving $1 million per month in operating costs by going from more than 4,000 servers in 150 locations to fewer servers in just 10 or 12 locations, he said. And that’s just on the hardware side.
Consolidation lends itself well to the ability to implement, say, a business process management (BPM) suite, so that everyone enterprise-wide can start doing things in the same way using the same technology, thus saving time, money and giving customers a single view of the company — something that is very hard to do if every business unit is running its own IT department, said AMR Research’s Dennis Gaughan.
“What companies found was they were not presenting a consistent face to the customer; that different parts of the organization were interacting with the same customer in a different way,” said Gaughan, a research director. “It was creating confusion and lack of good service on the part of the customer.”
Consolidation also has the benefit of taking advantage of technologies that just don’t work well in a distributed system. SaaS (software as a service), for example, is a technology designed specifically to push applications out to users in a one-to-many fashion. This means companies now can pay for and maintain a single instance of an application for a worldwide group of users.
This is not a new idea, of course. But with the ability of broadband, 3G and WiFi networks to handle massive amounts of data today, it is one whose time may have come. And a centralized infrastructure only further enables this effort, said Doug Neal, a research fellow at Computer Sciences Corp.
“I think we’re looking at an era where the IT organization has a lot different set of responsibilities,” he said. “And … we think IT is no longer a monolithic kind of structure, but rather IT consists of a layered set of services, some of which are developed internally, some of which are developed externally. But it’s the job of IT to put a wrapper around the whole thing and to orchestrate and manage the whole shebang.”