Architected IT is the only way to achieve the most return on your technology investment, and a radical makeover is the only way to get to that rationalized approach to running a company's technology infrastructure, Hewlett-Packard CIO Mott told a group of rapt technology managers yesterday in Austin, Texas.

Randy Mott is something of a CIO superstar, having made his bones as the engineer of Wal-Mart's highly regarded data warehouse (full disclosure: he's also on InformationWeek's editorial advisory board). Since late 2005, Mott has been working on a "transformation" of Hewlett-Packard's approach to internal IT. Newly ensconced HP CEO Mark Hurd coaxed Mott from his job at the time as the CIO of Dell for just that purpose.

Transformation is Mott's own word, and he doesn't use it lightly. He and Hurd have set as a goal the consolidation of 85 HP data centers worldwide down to six in the United States; 6,000 legacy applications down to 1,500; 1,240 IT projects down to 700; and 19,000 internal IT and contract workers down to 8,000. They're also looking to flip the ratio of IT workers supporting old applications versus those working on new apps from 70%-30% to 20%-80%, and increase the percent of IT projects being delivered on time from 81% to 98%. And all this in just three years, which means the deadline for this ambitious IT overhaul is fast approaching.

Another value Hurd wanted out of the IT makeover: to use the story of the transformation, and the resulting hyper-efficient IT operation, as a showcase for customers and potential customers on the most effective way to use computers in business. To that end, Mott and Hurd have been traveling the country giving quarterly updates to conference-room-sized crowds of technology managers on the company's progress, and the attendant benefits to business when the makeover is complete. Thus the meeting in Austin.

The first step, Mott told the Austin crowd, is that technology managers must realize that a radical IT transformation is the only way to achieve significant and lasting results. Trying to pick and choose among various and equally pressing IT priorities -- server consolidation, application portfolio management, rationalizing IT resources -- is a recipe for failure. "Choosing is losing," he said. "You're going to guarantee you'll never get finished. The incremental fashion just doesn't work."

A deep dive into a company's technology infrastructure will help open a technology manager's eyes. Some of the 1,240 IT projects Mott cataloged at HP during a 30-day review weren't slated to be finished for 20 years. "If it takes 20 years to finish, it can't be that important," he jokingly pointed out.

Next, CIOs have to get a commitment to IT transformation from upper management in terms of cold, hard cash. Hurd & Co. ponied up a cash investment of 2% of HP's 2005 revenue for Mott's three-year IT transformation project. According to Mott, citing "various indices," the average ratio of cash to revenue in most companies is in the area of 10% to 25%. "Most of you have the cash to do a transformation," he told the crowd.

But no company hands over cash easily. The trick is making the case for an IT transformation in terms of return on investment. And speed is a very important factor. Each company must determine the right time frame for such an ambitious project, but the shorter the period of return, the more likely the suits will buy in -- and vice versa. "Being conservative on how long this [transformation] takes will ensure you won't get the cash," Mott said.

The benefits of HP's IT transformation -- and therefore the business case -- include:

A radical IT transformation has its downsides, Mott said, such as higher up-front investment and higher risk. But the upside is much greater. Mott anticipates a 30%-plus improvement in IT efficiency, as old servers are swapped out for fewer, better-performing ones, and storage and network costs are lowered while increasing capacity and bandwidth. It also leads to more innovative IT solutions and better "synergy between initiatives," Mott said.

All this can add up to some "fascinating results," Mott told the Austin crowd. If Mott and Hurd are able to achieve their ambitious goals for rebuilding HP's IT architecture and performance, those results will indeed be fascinating -- and perhaps worth emulating. That's HP's selling point. And a point of pride for Randy Mott.