Planning Out a Portal

After nearly two years of infatuation with the corporate portal, the enterprise is waking up. Many companies that bought into portals as the killer app for knowledge management have found themselves with solutions that don't deliver the expected functions or performance. In fact, these tales of horror are as common as the highly publicized success stories that vendors spin.

"People have been, and are still being, sucked in by the hype and sexiness of portals," says Gene Pfifer, vice president and research director of Internet strategies at the Gartner Group in stamford, Conn. But "putting [a portal] in and making it work isn't anywhere near as easy as advertised."

When done right, portals can deliver great value to an organization by consolidating information from a vast array of sources, providing a centralized means of collaboration and offering a personalized workspace for both individuals and teams. But things can also go horribly wrong.

Fools Rush In

Planning ahead is essential to any project, but in today's frantic business climate it's easy to be stampeded into launching a portal. Before even calling your first vendor, understand the exact business value the portal will deliver. The company that forces itself to go through a quick but rigorous evaluation process to identify at least its minimum goals won't have to pay the price later.

The point is to get the project on firm footing before you start wrestling with the technical details of implementation and deployment. According to Kathleen Hall, a Giga analyst, it may even make sense to delay some of the technical challenges if possible. "While the initial launch of the portal needs to be compelling, there's no reason content or services can't be added incrementally based on demand," she says.

Thoughtful portal builders will first draw a roadmap and ask pertinent questions. Who will use the portal? Will it eventually be rolled out to additional groups within the company, perhaps even to customers and partners? How large a project should it be, and which interest group should be the initial target? Answers to these kinds of questions will suggest how much the builders may take for granted in terms of user expertise.

Don't Cut Corners

Andy Roehr, senior managing director of business development at Perficient, a professional services firm in Austin, Texas, advocates addressing design details in advance, even if it slightly delays the project. His project team, which specializes in implementing software for high-tech start-ups and consultants, learned this lesson from experience.

Some initial shortcuts on the look-and-feel side of Perficient's portal produced complaints from the user consultant, and managers who were the key source of content didn't post material to the portal on a regular basis, a reluctance that also could be traced back to the core issue of usability, he recalls.

"It would have been worth it to spend more time up front on getting it to look good and thinking through how to populate it more thoroughly, because more people would have used it from the beginning," says Roehr.

Identifying target users' interests and needs as precisely as possible also has obvious implications for content planning. As Giga's Hall notes, knowing what you need up front allows for a more coherent, efficient--and likely more cost-effective--content acquisition process.

Don't Be Shy

How, then, do you find out what users need from a portal before it exists? Ask them. Relying on the IS people who will do most of the portal building to know their users' needs is a common practice--and a dangerous one. A successful portal must offer workers compelling reasons to change their work habits. Those reasons are close to impossible to ascertain from the outside and much more subtle than an IT-based user profile.

Compaq Computer of Houston discovered this truth. The company's marketing and communications division is building a portal for its global sales force in hopes of reducing time spent answering e-mail, locating marketing materials and coordinating transatlantic conference calls. According to Sarah Thomas, the Internet strategist for Compaq's marketing and communications division and the lead manager on the project, a portal was the right solution because it could centralize and simplify both communications and the distribution of materials to members of the sales force.

In planning the deployment, the development team concluded it needed to know how the 2,000 individual users--most of whom work from a remote office or their homes--spend their days. To get at this critical information, Thomas and her associates spent five months doing extensive e-mail surveys, focus groups and one-on-one interviews. "You really have to walk in their shoes if there's any hope of the portal being a success," she says.

The team found that the users wanted the portal to supply basic information they needed to do their work, such as account news, product updates and information specific to their own teams. But they did not want to have to spend as much time at the portal as the e-mail system required.

That feedback shaped how the Compaq team put the portal together. But despite her team's efforts, Thomas expects that some usability issues probably won't surface until people are actively using the portal. Her advice is to do as much homework about individual work habits as possible before, during and after the portal implementation.

Don't Make It Hard

Once you understand your user base, the work can begin. But another common pitfall is building a portal that isn't tailored to user needs. By making people, especially those who aren't used to working in a Web environment, struggle to find what they need, you risk losing them altogether.

A user having to click more than three times to get to business-critical information can lose patience, according to Bob DiGiamo, senior vice president of IT at New York City-based investment bank Lehman Brothers. He acquired that perception while setting out to build a minimum-click portal for Lehman's globally distributed core of bankers, which had do appeal to technically savvy junior staff to senior financiers. He couldn't afford to waste the high rollers' time by forcing them to search the Web for relevant information.

DiGiamo decided that the portal should have a good deal of personalization built in--even before the portal was launched. That meant taking the time to figure out which companies each banker was responsible for, which country or region he or she called home, and which subset of the hundreds of available information sources the target users actually relied on.

"The idea I was running with is that we should be smart for our bankers, so they could work more efficiently," says DiGiamo. The portal currently is being tested with 50 people in New York, but the true test will come when 1,000 users go online this fall.

Don't Get Cocky

Another pitfall to avoid at all costs is the assumption that the users will flock to the portal, particularly if those users happen to be outside your organization. In early 1999, Pleasanton, Calif.-based scanner manufacturer Visioneer began to address customer complaints centered on the need for more real-time information about inventory levels and product delivery. The solution seemed clear: a Web-based customer portal where internal sales people and customers, including Costco, Office Depot and Best Buy, could check to see what was available, how many products were coming and when.

Jeanne Dorcich, Visioneer's vice president of sales and administration, was concerned about getting customers to use the portal. Her group's imperative was to make using it easy. So the developers gave customers access in only two clicks to reports about inventory levels, which are generated from a back-end ERP system, and projected delivery times, which are provided by Federal Express and other shipping sources.

Yet even with this direct information design, use of the portal has met resistance from some technology-averse customers. The challenge, as Dorcich sees it, is to get them to phase out the practice of phoning Visioneer. It will take time, she says. "Pushing back on customers a little at a time is what it has taken, continuously showing them how easy it is to use and monitoring whether they're using the system."

How can you avoid or overcome portal pitfalls? No project will be a success overnight, perhaps not over a few months, so persistence is necessary. Bypass the vendor hype, stay focused on your goals and flexible to change. Strive to understand the portal's business value and its target audience. Listen to feedback and learn from others.

And by all means, look before you leap.