Rush to Judgment

U.S. firms are adopting the 'Colin Powell' approach to IT buying: short lists, fast tracks and magic quadrants.

The days when nobody got fired for buying IBM are long gone. But these days, selecting IT products is becoming ever more complex as vendors and their offerings proliferate. Buyers are responding to the difficulty in two ways: by limiting choices to a short list of preferred suppliers, and by finding ways to fast-track the product selection process.

But shortening what was once a painstaking (and often painful) evaluation process introduces risks that must be carefully managed, acknowledge fast-trackers.

Since 2000, The Boeing Co. in Chicago has whacked its list of preferred IT suppliers in half, to 40 companies. Greg Farmer, director of IT procurement at the $52 billion company, says that has helped Boeing deal with the complexity of the supplier landscape, but the complexity of the stuff it buys has not decreased.

"In the past, we'd go out and do a request for information, learn about the companies, read about them, and after this exhaustive study we'd say, 'OK, now we'll go to these five suppliers with a request for proposals,'" Farmer says. "More and more now, we look at the Gartner Magic Quadrant and say, 'Who are the top players?' and then send the requests for proposals out to them."

It isn't just the complexity of vendors and products that's driving Boeing's buying practices. "There is so much pressure at Boeing to reduce costs and be lean and efficient," Farmer says. "We don't have the time, energy or resources to duplicate what [firms like Gartner] have already done."

Another trend, he says, is to see products through the eyes of the business rather than the lens of technology. Indeed, his 60-person IT buying team is not populated with technologists and does not report to the CIO. They are business people with IT buying experience, and they report to a vice president for supplier management.

Different Strokes

When MIT recently looked to buy business intelligence software, CIO Jerry Grochow resisted advice to do a very detailed analysis of the many potential vendors and their BI offerings. Instead, he just picked Hyperion Solutions Corp. "We had some people around MIT who had used Hyperion elsewhere, and they were very satisfied with it. And it was clearly one of the market leaders," he says.

MIT limited its risk by bringing in the BI product for a "quick prototype," Grochow says. "Rather than spending $20,000 on analysis of the feature sets of alternatives, I said, 'Let's bring one in and spend $20,000 to do something with it.' We learned more by doing that."

Still, Grochow acknowledges that such a rapid-fire "try it then buy it" approach is not always possible or appropriate. MIT will soon spend up to nine months and $500,000 evaluating student information systems -- about as mission-critical as IT gets on campus. "Other schools have spent $30 million to $40 million on them, so we are going to go through a significant study, lots of review of alternatives, detailed visits to other places that are using them and so on," Grochow says.

ING Group NV has offices and CIOs all over the world, and the CIOs sometimes come to Raymond Karrenbauer, group chief architect, for help with their IT purchases. He says doing in-depth analyses of alternatives is common practice in the European and Asian business units. "They spend time very diligently looking at accuracy, predictability, all the features and requirements, until they have a near-perfect set of details," he says. Then the buying decision must be approved at multiple levels in the company.

But North American companies are more likely to apply the "Colin Powell" philosophy to IT buys. The former secretary of state, according to Karrenbauer, "used to say that as long as he had 70% of the data, that's good enough to make a decision and just move forward." The difference reflects a U.S. business culture of moving very quickly, but both paths work well for their respective regions, he says.

Karrenbauer says he can select vendors and products quickly because of the homework he does beforehand. He starts with an internal system architecture for the corporation that factors in business requirements and forecasted needs. The resulting framework of business needs against architectural standards makes the actual product selection pretty straightforward, he says. Certain vendor "footprints" will be readily seen as good fits, while others obviously are not, he says.

Dennis L'Hereux, CIO at Rockford Health System in Rockford, Ill., says his company has gone from a philosophy adopted more than 10 years ago of buying best-of-breed products for each function to one in which preference is given to buying suites of integrated products from just a few vendors. Best-of-breed products are likely to better address the needs in question, but they often require the use of integration tools, "and that requires a lot of effort," he says.

For example, when L'Hereux recently looked for a medication safety software package, he considered just the four offerings from preferred vendors, not the 15 to 20 that were available. But those four contenders each received a detailed evaluation.

Managing Expectations

The detailed analysis helped L'Hereux choose the right package, of course, but it had a side benefit as well, he says. "It's an exercise for the people who will use the system to understand what it will do and what they want it to do," he says.

That exercise can head off unpleasant user scenarios. "They all want everything, but when they get it, they don't want to use it because it doesn't work like the old way," L'Hereux says. Or another unwanted user reaction: "Bring it in, and then I'll tell you why I don't want it," he says.

Like Boeing, Rockford Health relies on market research to provide some guidance. L'Hereux taps into the analyses of Klas Enterprises LLC, an IT product research firm for the health care industry, to help him prescreen products and vendors.

L'Hereux says it was once standard practice for IT shops to put together "very detailed requests for proposals in three-ring binders with everything we wanted in a system." But in addition to being time-consuming and costly, he says, that process "doesn't materially result in any better decision."

Indeed, says MIT's Grochow, spending a lot of time on an IT buy is no guarantee of success. "People will say, 'Let's examine the details of all 30 alternatives.' Do they come up with a significantly better decision than the person who says, 'Here are the two market leaders; let's bring one of them in here and try it'?"

But doesn't studying every alternative help cover your rear? "It's a hallmark of the more senior manager: You have to be willing to take some risks," Grochow says. "Sometimes the risk is explicit, and sometimes it's implicit. The people doing the detailed analyses are still taking some risks; they just don't think they are."

A 2005 'Magic Quadrant' for Backup/Recovery Software
A 2005 'Magic Quadrant' for Backup/Recovery Software

Source: Gartner Inc., July 2005

Computerworld's IT Salary Survey 2017 results
Shop Tech Products at Amazon