Backdoor I.T.

Tom Brailsford had a good idea for a technology project, but he couldn't get it done. Not without a little backdoor help from his IS department.

Brailsford, who is manager of knowledge leadership for the consumer research division of greeting-card company Hallmark Cards Inc., faced a challenge that haunts businesspeople everywhere these days: How do you get a technology project up and running when IS is too busy to help you?

People like Brailsford are getting the work done by slipping in through the back doors of their IS departments and enticing the techies to help them in off-hours, or nights, or weekends. The trick is to make the project relevant and exciting enough to crack open the door and catch somebody's eye. Brailsford tied his project-basically a catalog of Hallmark's research on customers' and retailers' spending habits-to something IS was already working on and let IS people get creative with the Web technologies they were eager to learn about.

Brailsford was convinced the company would gain by getting a peek at research that had remained holed up in his research department for years. But his proposed intranet project didn't show well, to use a real-estate term. He couldn't attach a hard ROI number to it, and in 1998 that was the kiss of death at Hallmark. Proposals requiring more than 10 man-days of IS's time had to get past a tough prioritization review board that scrutinized each for business value and ROI.

Brailsford didn't even bother going before the review board. His project would not survive that kind of scrutiny, and he knew it. It wasn't a flashy enterprise resource planning (ERP) or electronic-commerce project with software vendors and consultants on board to help shake money and resources out of the corporate chieftains. His was the kind of project that bubbles up out of a department in response to problems people have getting their work done. It was something that would have a direct impact on the way people did their jobs, but he just couldn't prove it. "What's the ROI on knowing what you know and sharing it around the organization?" he asks. "People ask you, 'What's that worth?' You have to say, 'I don't know.'" But that wasn't the end of Brailsford's project. Knowing that IS had officially set aside resources to do four intranet pilots in the company, Brailsford signed on his research group for one of them. The deal was pretty simple-even a little stark. You got room on Hallmark's network to run the intranet, software to develop the site and a server to run it on. "IS said, 'If you want to do it, more power to you,'" recalls Brailsford. The only catch was that the intranet projects would receive no budget and only a wisp of IS time to be split among the four projects.

And that's where backdoor IS people like Kurt Stackelbeck can help. IT projects that are caught in a limbo of constrained resources need assistance from IS people who don't have time to lend a hand. Stackelbeck, an IT specialist at Kansas City, Mo.-based Hallmark, is one of the people who keeps grassroots, no-budget, no-people IT projects like Brailsford's alive. He finds ways to lend spare minutes here and there to projects that lack the funding and the people to survive without some backdoor cajoling.

You need people like Stackelbeck, too. Your IT infrastructure will look like a mess without him. Businesspeople who try to go it alone or hire a consultant to help with their small projects are at a loss unless they have a backdoor IS person who understands the complex IT guts of the company-its networks, its application infrastructure, the weird and unique quirks in the corporate flow of information. Without the Stackelbecks of the world, there would be a proliferation of amateurish, poorly designed applications that are unstable, unscalable and a nightmare for IS to support.

But Stackelbeck can sympathize with the businesspeople who do try to go it alone out of frustration. "It's prudent to prioritize when you have a shortage of resources and need to maximize those resources," says Stackelbeck. "But IS always looks like the bad guy because we have to say no more than we can say yes, particularly when you're trying to clamp down hard on costs as we were."

The tragedy of the prioritization process in any company is that many projects that could show demonstrable ROI in practice are never implemented because that ROI can't be projected accurately at the start and captured in a spreadsheet. In the cost-cutting climate that existed at Hallmark in 1998, you just didn't go around crowing about how knowledge management or intranets were going to change the world. "It wasn't, 'Hey, let's do an intranet. It costs a million, but it's worth it-you'll see,'" recalls Brailsford. "You'll make one of those pitches in your career [at Hallmark]."

Although Brailsford's pilot status got him on IS's radar screen, he was still at the very bottom of IS's priority list. If Brailsford, who was the only person committed to his intranet pilot full-time, wanted to spend any money on people or resources, something would have to give somewhere else. "The directive was that you will not add budget or people. Take it from existing resources," he recalls.

To help these grassroots intranet efforts get off the ground, Hallmark's IS group, along with the public affairs department, set up the Intranet Steering Committee-a support group really-with representatives from the four pilots, IS and public affairs. In the beginning of the process, the group met monthly to share ideas about website design, troubleshoot content and help each other find volunteers to get the work done. The people part was the hardest. To enlist help for the research group's intranet, dubbed Voice of the Marketplace (VOM), Brailsford had to find people able to balance their spare-time participation in the effort with his need to keep the intranet moving at a somewhat reasonable, rational pace. He tried to lure people in his research group by playing to their personal interests and to the glory of hooking up research with the rest of the company online. "The promise of getting a new skill and sharing information made it worthwhile for people to spend nights and weekends on this," he says.

But Brailsford didn't limit himself to schmoozing up insiders for help. He also threw himself at the feet of other companies that didn't mind showing off their intranet accomplishments, including Levi Strauss & Co., Qualcomm Inc., Sprint Corp. and Sun Microsystems Inc. By pushing his background in research, Brailsford managed to get appointed to The Conference Board's all-volunteer council on knowledge management and learning organizations in late 1997. As he and his colleagues on the council began developing their own intranets, they pumped each other for information and hands-on help.

One of the biggest challenges Brailsford and his council colleagues faced was getting help from their IS departments. It wasn't that IS wasn't interested-the Internet was and is the hottest thing happening from a technical standpoint-it was the time and ROI problems. Stackelbeck was one of two Hallmark IS people devoted part-time to the Web, but he and his colleague (a young intern, actually) were supposed to be researching Web technologies, not helping with intranets.

Luckily for Brailsford, however, Stackelbeck is not a big believer in research for research's sake. "Unless it's applied somewhere, R&D isn't really beneficial," he says. "I always look for opportunities to do R&D using a real project."

Here is where Brailsford had an angle to play. Hallmark's IS group had recently decided to disperse its IS people out into the different functional areas and business units of the company to better align business and IS goals. Colocation, as it is known, is a great way for IS people to get to know the businesspeople they serve and to better understand the business problems those people face.

Stackelbeck sat right near Brailsford's research group and was responsible for the digital happiness of the portion of the research group devoted to studying Hallmark's retailers. The two men got to talking about their Internet passions and began hatching ways to help each other.

But they faced some problems. Brailsford, desperate to make sure that the VOM would be more of a shout than a whimper, fretted that he was developing something that wouldn't have enough impact within the research department, let alone the rest of Hallmark. Stackelbeck, meanwhile, worried that the intranet might become too successful, thereby creating a support nightmare for the overworked IS department. "The primary danger in research projects is that they will be wildly successful and IS won't have the manpower to support them," he says.

So the two men set about developing an intranet that was deep enough to provide value to the researchers but simple enough for the researchers to maintain without much help from IS. This kind of challenge excites good engineers like Stackelbeck, because it means creating an almost-living entity that people can use on their own. "It doesn't have to be the geekiest solution to be successful," says Stackelbeck. His contribution was twofold. First, he created a set of webpage templates that enabled people to publish content without IS's help. And then he trained 18 to 20 research people to use a simple Web publishing tool and maintain their webpages.

Stackelbeck gave researchers their own webpages corresponding to their research responsibilities and interests. To keep the site from appearing fragmented or overly complex, he linked the individual pages to a main homepage with clear navigation options. Stackelbeck's philosophy is that if businesspeople can be engaged and shoulder the responsibility for an application and its maintenance, he lets them.

Soon after the new site went live in 1998, however, Brailsford and Stackelbeck saw that it wasn't having a great deal of impact outside of consumer research. As the site grew fat with content, Brailsford's ROI dreams began to ebb and Stackelbeck's fears of support costs began to flow. Both men realized they needed to broaden the appeal of the intranet to make it a more exciting log-on. Perhaps the most universally compelling information within Hallmark is the point-of-sale (POS) data from retailers. These are the hard numbers that show whether Snoopy is kicking Woodstock's butt in the card aisles, or whether the boysenberry-scented candles are smelling sweeter to consumers this week than the French vanilla.

Stackelbeck's day job is all about that data. He maintains Hallmark's retail information systems, which include an old POS legacy system that holds all of Hallmark's historical data. Stackelbeck had been wrestling with how to make the old system get along with a new POS system Hallmark wanted to use to access new data. "You'd query each system for the same thing, and they'd come up with different answers," he says. Stackelbeck realized that he could get the systems working together by linking them with Internet technologies. From there, it was just a short leap of logic to suggest giving those Internet technologies a home on Brailsford's website. "It seemed like there was enough of a relationship between the research on the VOM site and the POS data that this would not be a tough sell to IS," recalls Stackelbeck. "I thought, why make a request for a new project when I could use one we already had running?"

Suddenly, Brailsford and Stackelbeck had their major problems solved. VOM had the information that all of Hallmark wanted and gave the research group a platform to add deeper context to it. And Stackelbeck had the POS link to justify requesting extra time to support the site. Brailsford plans to funnel other urgent sales data into the intranet, making it a gateway for business information about Hallmark's current, past and future market position. "The old retail sales databases that people used here had no context around them," says Brailsford. "All they could pull up is the fact that Valentine's Day sales were up 2 percent. We want to get to the point where people can use the site to understand why sales were up 2 percent. We want to make it into an information portal."

Brailsford knows he can go back to Stackelbeck for even more help toward that goal, because his site is on the map. With POS data running through it, IS has much more reason to siphon off valuable support to the site because it is in essence a part of the company's POS system now. Not to mention the fact that he can offer up the chance for IS people to experiment with Java, help develop a website and create an Internet infrastructure. "IS people are looking around the job market and saying, 'Boy, if you don't have Web development, you're only half a programmer,'" says Brailsford. "Plus they like being part of a skunk works, because if they can get it done in addition to all the other stuff they have on their plates, they can be a real hero."

In their spare time, of course.


The Genie has left the bottle

It's human nature to trust someone you know more than someone you don't. When it comes to computer issues, this is particularly true. Businesspeople start to understand how overworked IS people are (or aren't) when they can see them laboring in the next cubicle. When they run into IS people in the hallway, relationships are built and technology problems that might not have made their way to the help desk get aired more easily. The main downside of colocation is that everyone can become a little too chummy. Suddenly, the local IS guy who is supposed to be consulting on application development or launching a website is unfreezing frozen laptops all day long while the pimple-faced help desk jockey who's paid one-third as much reads cyberpunk novels all day.

"Putting IS people in the business units makes them more responsive, but the cost of IT tends to grow," says Michael Gerrard, research director in business management of IT for Gartner Group Inc., a Stamford, Conn.-based consultancy.

Should you be willing to trade productivity for responsiveness? Probably. Because it's easier to work on productivity than it is to mend the communication gap that develops between IS and the business when IS is hidden away in a central location. "Your net costs are lower when IS is centralized, but responsiveness goes down," says Gerrard. "IT is so pervasive and important today that it needs to be managed in a very customer-sensitive way. So the priority for IS should be operating at the business-sensitive level rather than the lowest-cost, most-efficient level."

"We were located really close to the researchers, which made it easy," says Kurt Stackelbeck, an IT specialist at Hallmark, who helped Hallmark's research people build a grassroots intranet. "Projects like the intranet will arise if you can have the hallway conversations and see the enthusiasm in the faces of the businesspeople."

But colocation in and of itself isn't enough. "Sitting next to each other isn't essential, but a good working relationship is absolutely key," says Stackelbeck. "Businesspeople have to be committed to a project if you want it to be successful." IS can try to build that commitment from afar, but it's easier to do it when you all visit the same watercooler.

Join the newsletter!

Error: Please check your email address.

More about GartnerGartnerGatewayLevi StraussQualcommSprintSprintSun Microsystems

Show Comments

Market Place