Nearly every sizable company has its own horror stories about IT projects that went wildly over budget or off schedule,...
that never came close to delivering the promised benefits, or that were simply abandoned. A study conducted by the Standish Group in 1995 uncovered a striking record of failure in IT projects. Of the more than 8,000 systems projects Standish examined, only 16% were considered successes -- completed on time and on budget and fulfilling the original specifications. Nearly a third were canceled outright, and the remainder all went over budget, off schedule, and out-of-spec. Large companies -- those with more than $500 million in annual sales -- did even worse than the average: Only 9% of their IT projects succeeded.
And when IT projects fail, Standish found, they fail in a big way. Most of the cost overruns amounted to more than 50% of the original budget, and nearly a quarter of the over-budget projects exceeded estimated costs by 100% or more. Of the projects that went off schedule, 48% took more than twice as long as originally planned, and 12% took at least three times as long. Of the projects that were completed but fell short of initial expectations, more than 30% failed to deliver even half of the originally specified features and functions. Standish also found that almost all the projects -- 94% -- had to be restarted at some point, and some had to be restarted several times over.
In a follow-up survey in 1998, Standish found some improvement, but the overall picture remained bleak. Although the percentage of successful projects had risen to 26%, this was still less than the percentage that had been canceled (28%) or that failed to achieve their intended results (46%). Another 1998 study, by accountants KPMG, delivered even worse news. Of the 1,450 companies surveyed, three-quarters said that their IT projects exceeded their deadlines, and more than half said that the projects went substantially over budget. When KPMG analyzed 100 of the failed initiatives, it found that 87% of them had gone over budget by more than 50%. Bob Napier, the chief technology officer of Hewlett-Packard, summed up the situation well in a 2003 interview: "The number of projects that fail is scary."
Many of the failures were, in retrospect, inevitable, a natural consequence of the process of trial and error that goes on as any new technology is adopted by business. Trying to place blame on any one group -- vendors, consultants, CEOs, CIOs -- would at this point be a futile exercise. The challenge now is to bring the failure rate down -- quickly and sharply. Given the high risks inherent in IT projects and the declining likelihood that they will lead to the durable advantages necessary for higher profits, both users and vendors need to concentrate on such mundane but essential requirements as efficiency, predictability, reliability and security. The time has come, in other words, for a more conservative approach to IT management. As the infrastructure matures, the companies that succeed will not be those that reflexively pursue innovation, that seek to push the proverbial envelope, but rather those that are pragmatic in planning and competent execution.
IT management presents companies with many risks, but at the moment the greatest of them all is overspending. Information technology may be a commodity, and its costs may fall rapidly enough to ensure that any new capabilities are quickly shared, but the very fact that it is entwined with so many business functions means that it will go on consuming a large portion of corporate spending for the foreseeable future. Information technology will continue to be "an insatiable economic sump," as author James McKenney memorably described computing during the mainframe era. What's important -- and this holds true for any commodity input -- is to be able to separate essential investments and activities from ones that are discretionary, unnecessary or even counterproductive.
The first challenge facing managers is to put their IT house in order. Most companies can reap significant savings by simply cutting out waste. Personal computers provide a good example. Businesses purchase more than one hundred million PCs every year, most of which replace older models. Yet the vast majority of workers who use PCs rely on only a few simple applications -- word processing, spreadsheets, e-mail and Web browsing. These applications have been technologically mature for years; they require only a fraction of the computing power provided by today's microprocessors. Nevertheless, companies have continued to roll out across-the-board hardware and software upgrades, often every two or three years.
Much of the spending, if truth be told, is driven not by the interests of the buyers but by the strategies of the sellers. Big hardware and software suppliers have become very good at parceling out new features and capabilities in ways that force companies to buy new computers and applications much more frequently than they need to. Intel and Microsoft in particular have created a very lucrative cycle of releasing faster microprocessors and more complex software; buying one often leaves companies with no choice but to upgrade the other. Some vendors of expensive enterprise systems such as ERP software even require clients to upgrade to new versions in order to continue to receive maintenance. Since vendor support is critical to keeping the complex systems running, businesses have little choice but to pay the tab.
But if there's a silver lining in the commoditization of IT, it's that the balance of power is tipping away from the vendor and toward the user. With competition among suppliers intensifying, IT buyers are now in a position to throw their weight around much more aggressively -- negotiating contracts, for example, that ensure the long-term viability of their PC investments, tie payments to actual usage, incorporate tough service-level agreements, and impose hard limits on upgrade costs. And if vendors balk, companies should be willing to explore cheaper solutions, including open-source applications and barebones network PCs, even if it means sacrificing features. If a company needs evidence of the kind of money that might be saved, it need only look at Microsoft's huge profit margins on PC software.
But PCs are just one example. Wasteful IT spending has long been endemic in corporations, and it reached plague-like proportions during the Internet boom of the late 1990s, when, as one computer industry executive put it, "servers were growing like bacteria." Today, in the wake of all the excess spending, "considerably less than half of [installed] IT capacity is used," reports the Financial Times. Needless to say, much of the superfluous hardware and software will never be used -- it's already out of date. But the lesson is clear: Companies need to make sure they get the value out of past investments before making new ones.
Businesses also have opportunities to impose tighter controls on IT usage. That's particularly true with data storage, which has come to account for more than half of many firms' capital spending on IT. The bulk of what's being stored on corporate networks has little to do with making products or serving customers -- it consists of employees' saved e-mails and files, including terabytes worth of spam, MP3s, and video clips. Computerworld estimates that as much of 70% of the storage capacity of a typical Windows network is being wasted -- an enormous unnecessary expense. Restricting employees' ability to save files indiscriminately and indefinitely may seem distasteful to many managers, but even such a simple step can have a real impact on the bottom line. Now that IT has become the dominant capital expense for most businesses, there's no excuse for waste and sloppiness.
Excerpt reprinted with permission from Does IT Matter? Information Technology and the Corrosion of Competitive Advantage, by Nicholas G. Carr. Harvard Business School Publishing, 2004. All rights reserved.
To read more articles like this one, visit HBS Working Knowledge, an online source for business analysis, information and research.
© 2003 President and Fellows of Harvard College