Dell CIO Randy Mott thinks that CIOs are dropping the ball if they don’t spend a lot of time planning for obsolescence. Mott, who sounded the message in late January at LinuxWorld in New York City and elaborated later in an e-mail response to questions from TechRepublic, said that enterprises in all industries must do a better job of preparing for the future. As it stands, he said, they’re putting their companies—and by extension their industry segments—in danger by propping up antiquated technology too long. He cited the airline, automotive, and railroad industries as good examples of how not to proceed.
Mott contends that the increasing power of technology hasn’t been well leveraged by IT professionals. On one hand, he said, the United States—using computers that would be called primitive by today’s standards—successfully met what appeared to be an unrealistic challenge by President John Kennedy in 1961—to land a man on the moon by the end of the decade. He compared that with a world in which a vast majority of IT dollars—he pegs it at 85 percent—are spent “keeping the lights on,” Mott’s euphemism for maintaining the status quo.
He said people reluctant to actively court the future end up as history. He cited three airlines—Eastern, Pan Am, and Swiss Air—as prime examples. “The lack of planning for obsolescence meant that they didn’t take steps to more forward, so there was a lot of decay,” he said. Mott said that the airline industry achieved a tremendous amount, but that stopped once the era of innovation ended.
Great entrepreneurs—he mentioned Sam Walton, Bill Gates, George Westinghouse, and others—accept this, he said. It demands a conscious effort. George Westerman, a research scientist in the Center for Information Systems Research at MIT Sloan, is familiar with the issue.
Focus on innovation
“One problem is that, in most mature organizations, the organizational culture you need for innovation can be very different from the culture that keeps the rest of the business going,” Westerman said.
It can be a struggle, Westerman said. “Without a focus on innovation, innovators tend to get squashed by people who want to use money for doing more of what they are already doing. That's why innovation often takes conscious effort. The top of the company needs to foster an environment that appreciates innovation and improvement. Then, the middle and lower levels of the organization need to be creative—trying new experiments to see what happens.”
Mott is a big fan of open source and Linux. Last year, about half of the operating systems in the field were UNIX-based. The rest were divided among Windows, Linux, OS 390, and OS 400 running on mainframes and NetWare. By 2010, Mott—who emphasized that he wasn’t speaking on Dell’s behalf—said that the OS world could be segmented quite differently. For instance, he suggested, it could be almost half Linux, perhaps 30 percent Windows—he called Gates a “great obsolescence planner”—and perhaps 25 percent the “next OS.” If this or something like it happened, Mott suggested, it would mean that people were taking obsolescence planning seriously.
Moving from the proprietary model
Dell, according to Mott, is working to move from the proprietary model, which he says is a great hindrance to the type of innovation that he sees as necessary. He said that in fiscal year 2004, 55 percent of Dell’s “people resources” will be dedicated to innovation. The plan is to move that percentage up to 75 percent by fiscal year 2006. Last year, he said, the company retired 14 systems running proprietary OSs in favor of clusters running Linux. The clusters are 89 percent faster and 41 percent cheaper to run, he said.
How to prepare for obsolescence
Mott suggested ways in which IT professionals could better prepare for obsolescence. His advice, however, appeared short of specifics. He counseled employees to know the business they are in, enable—not disable—the enterprise, innovate and not stagnate, and avoid using band-aids to cosmetically update legacy gear.
Other things can be done, Mott said. “One obvious and often overlooked step is to consistently conduct systems and version upgrades, which enables organizations to fully leverage new capabilities,” he says. “If the application development organizations and infrastructure teams are closely aligned, they communicate regularly, and they govern this with good processes, the end result can be very positive and they can increase their development significantly.”
Westerman said it’s a case-by-case scenario. “Unfortunately, it's hard to put a clear number on how much effort is too much or too little for innovation, until it's too late and you're playing catch-up. You need to stay aware of what's going on so that you can take the right action at the right time for your firm.”
In general, Mott said, IT managers should know what percentage of their budget is tied up in legacy systems and processes, what percentage is tied up in keeping the enterprise running on a day-to-day basis, whether the enterprise’s technology roadmap plans for obsolescence, and whether the company’s business partners are using systems that are aligned with the IT organization.
The near-term goals of the organization need not be sacrificed, he said. “IT shops should focus on delivering projects in six months or less and identifying annualized benefits that can be accounted for in the corporate profit and loss statement and reflected in the individual internal line of business budgets,” Mott said. “This approach ensures they will deliver the tools and results that their business partners need and expect, and measurable returns that complement Wall Street expectations. The key is...the approach must be balanced, as this isn't an either-or scenario, and the near-term results must be coupled with the strategic roadmap.”
If the industry takes heed, he said, by 2010, IT organizations will be able to apply 85 percent of resources to development and increase productivity from 20 percent in the 1990s to 50 percent for the combined 2000–2010 time period.