Everyone’s a critic. All manner of people have an opinion about what’s wrong with IT departments. They are too big, too decentralized, too unstrategic — too something.
Like the fable of the six blind men exploring the elephant, different people focus on different parts. Arriving at the truth means assembling different perceptions of what ails IT.
Some believe a major sea-change is under way.This is not the death of IT organizations — we would rather call it a renaissance.John Mahoney>
At a recent symposium on the future of IT organizations, Gartner Inc. predicted that by 2011, at least 10 per cent of IT organizations will be disbanded, 10 per cent will be relegated to commodity status, and at least 75 per cent will change their role, based on extrapolations from current economic and technology trends. “This is not the death of IT organizations — we would rather call it a renaissance,” said John Mahoney, chief of research at Stamford, Conn.-based Gartner. “A new organization type is emerging, one that will take the lead on information and process.”
As technology grows more critical to the routine operations and strategic goals of most businesses in the future, he said, the IT organization’s contribution will come under greater scrutiny. Does it produce good results or bad? Is it worth the investment? The acid test will be the overall performance of the business.
If the IT organization’s operational and strategic contribution is weak, enterprises will cast a baleful eye and likely disband them. “We expect one in 10 enterprises that “own” their IT organizations will conclude that it is not strategic to their business to the extent they need to have a dedicated IT area,” said Mahoney. Outsourcing IT will serve their needs better than an in-house group, he said.
Other enterprises that only need technology for routine operations will also re-evaluate the IT organization’s role. “One in 10 will decide not to upgrade their use of technology to something more strategic,” said Mahoney. “These IT organizations will be treated like a utility that doesn’t require specific and separate management.”
The need to master technology to achieve loftier strategic goals is what will reshape IT organizations for the majority of enterprises, said Mahoney. “The drivers we see have to do with business processes, business relationships and business information. These three are forcing change away from pure delivery of technology to management frameworks for these three things.”
Mahoney warns that a major challenge for the renaissance IT organization will be developing a system framework that unifies all elements. “A framework will be needed that links lateral business processes, their relationships and so on together, and that defines the connections between those processes and regulates the ways they play together — not just within the business, but with its external partners.”
The utility of IT Even Nicholas Carr, author of Does IT Matter?, a controversial book that made waves in technology circles in 2003, thinks Gartner’s predictions go too far. “They are making a provocative forecast; it sounds a bit aggressive to me,” he said. “I would think one in 10 IT organizations would at least be moving towards being disbanded by 2011, but whether they’ll actually be gone is another thing.”
However, Carr does sniff change in the air. “We’re at the start of a fundamental shift,” he said. “It’s fair to say, by 2011, we will have reached the tipping point. In five to 10 years, the direction of IT organizations won’t be theoretical anymore.”
Carr is looking at historical precedents these days for insights about the future of IT. In his 2005 essay, The End of Corporate Computing, published in MIT Sloan Management Review, he looks at power generation during the early years of the 20th century to diagnose impediments to the emergence of a true utility model for IT in the 21st.
In the early years of the Industrial Age, American manufacturers maintained their own private power plants to reliably generate the electricity they needed to run their factories. But small industrial concerns couldn’t afford to do this, which opened the door to fledgling electricity producers who produced power as a utility.
These small players were soon consolidated into a giant monopolistic utility, Commonwealth Edison, by a visionary character, Samuel Insull. He was the first to realize that by consolidating and centralizing generating capacity, a utility could fulfill the power demands of the largest factories while also distributing power more efficiently across many customers.
Carr points out that IT is in a similarly ripe position today for the emergence of a true utility model, and trots out some compelling statistics to support his argument. In his essay, he cites a recent study of six corporate data centres that revealed that most of their 1,000 servers were using just 10 to 35 per cent of their available processing power. And desktop computers fare even worse, with IBM Corp. estimating average capacity utilization rates of just five per cent.
“The creation of tens of thousands of independent data centres, all using virtually the same hardware and for the most part running similar software, has imposed severe penalties on individual firms as well as the broader economy,” said Carr.
The pace of technology growth at the turn of the century was in many ways faster than today’s. “Real change in corporate electricity power supply happened during the1905-1925 period,” said Carr. “When the capacity of outside utilities was clearly good enough that factory owners could trust them, then it happened quickly.”
IT has been around for 50 years, and the concept of providing IT services as a utility has been around almost as long, so what is the holdup? Carr cites a few impediments.
No visionary like Insull has yet to emerge to put all the pieces together into a pure utility model, he said, where integration of hardware, software and data occurs in the utility itself instead of the client’s operations.
“A lot hinges on how to meter and charge usage. In traditional IT outsourcing, charges are not based on actual usage, but on the number of servers, users, chips and so on. You want to get to a model where you pay for what you actually use, not a proxy for use.”
But major IT outsourcers and enterprise hardware vendors aren’t motivated to change, he said, because they’ve benefited from the inefficiencies in the client server model. “If you can consolidate IT into a utility, you can increase utilization a lot. But this means customers won’t have to buy as much hardware and software, and that can make vendors nervous. That undermines their traditional model, which is supplying a heck of a lot more goods and services than the customer will actually use.”
Other impediments are political, said Carr. CIOs must cooperate in the demise of their empires, and it is human nature to resist this. “The big challenge CIOs face is that 70 to 90 per cent of their budgets and resources go towards maintaining their IT infrastructure. And they know they will be relegated to a peripheral technology maintenance role if they don’t reduce that,” he said. “The only way out of the trap is to get rid of the infrastructure.”