Durham, NC - An analysis of computer purchases during the fourth quarter of fiscal year 2010 reveals that Duke is not yet taking full advantage of the savings promised by the Computer Purchasing Program introduced eight months ago.

The report shows that 81 percent of the computers purchased during the three-month period ended June 30 were "in program," according to Assistant VP and Controller Tim Walsh.

"Computers bought through the program represent enormous savings for Duke," Walsh says. "When departments use the established Duke standards, they can enjoy a 50 percent discount off the manufacturer's suggested pricing on business-grade machines with extended warranties."

Even so, as EVP Tallman Trask told the Academic Council at its Sept. 23 meeting, the percentage of high-end devices purchased through the program exceeds initial estimates.

"Some 72 percent of computers purchased for individual use through the program in April, May and June were high-end, customized machines-that's 58 percent of all purchases across the university, in or out of the program," Walsh says. "It's clear people are still leaving a lot of potential savings on the table," he says. "If we can eliminate the out-of-program purchases and rebalance the in-program purchases to take advantage of the first- and second-tier offerings, we will have achieved some real cost savings for Duke."

In fairness, we may not have seen the full effect of the program yet, simply because of the natural cycle of computer acquisition. To arrive at these calculations, Walsh has removed machines destined for computer labs, which must be high-end. In addition, a check of Finance and Human Resources, two large administrative departments with many users of first-tier machines, shows that these departments have purchased very few computers since the introduction of the program.

Developed in conjunction with the work of the Duke Administrative Reform Team (learn more on the Enduring a Troubled Economy website), the program sells computers and peripherals from Duke's three preferred vendors: Dell, Apple, and Lenovo. Each vendor offers equipment in tiers to address the needs of faculty and staff:

- who use e-mail and calendaring software, office programs, web applications and enterprise client applications like SAP, PeopleSoft and Business Warehouse (about 60 percent of the total population)

- who run enterprise applications and Crystal Reports together (about 30 percent of the population)

- who do complex data collection and analysis, visualization, web-publishing and systems administration work (the remaining 10 percent of users).

The group that developed the standards, including Ed Gomes, head of IT for Trinity College, and Duke Computer Store IT manager Jim Rigney, acknowledged that most staff would need machines from the first and second tiers, while faculty would fall mostly into the second and third tiers. With many faculty in the departments Gomes supports engaged in data-intensive research-and moving increasingly into multimedia and visualization-he believes that 50 to 60 percent of his computer purchases will be in second tier machines, and purchases in the third tier may go as high as 40 percent of the total.

"There certainly are programs where level 2 machines are going to be adequate to meet the needs of faculty," Gomes says, "but there are growing segments where more robust machines are necessary." Still, he agrees that we can reduce the cost of computer equipment, especially in administrative departments-even IT groups, where a large majority of users can use PCs instead of Macs, and where only those running virtual servers or doing programming, multimedia and web development work need machines from the second or third tiers.

Walsh says that, with the price of machines in the first tier roughly half what Duke departments have paid in the past for computers, "buying these models, when appropriate, can double the savings for the university community."

The original scope statement for the computer purchasing program identified the risk of extending the useful life of computers to four years for laptops or five years for desktops, since the inclination would be for people to buy more powerful machines at the outset. Ginny Cake, assistant CIO for the university and a member of the team that designed the new computer standards, says that, to reduce Duke's overall computer expenditures, we should commit to purchasing computers from the first tier when they meet the business needs of the users.

"The program guidelines take into account the risk of obsolescence," she says. "If a machine is no longer adequate to meet a user's needs-or it requires a costly repair out of warranty-it can be replaced before the end of the extended life cycle." As long as these machines are in warranty, maintenance is handled by Duke Computer Repair.

Wayne Miller, who runs IT for Duke Law, has always relied on the standard offers of the Duke Computer Store. He gave feedback on the program design as it developed, and he is appreciative of the changes as the process moved along.

"We asked that the program include minitowers," he says, "because it's much easier to open up a minitower and upgrade equipment or add memory, something we may have to do to extend the useful life of the equipment. We were also pleased with the price and specifications of the wide-screen monitors that the program makes available."

To learn more about the program, visit the Duke Computer Store site. When you're ready to buy a new computer, contact your local IT support team.