Despite its corporate intranet, its vast network of servers, and plenty of business intelligence tools, Ericsson Research Canada knew full well that there were lots of duplicate efforts among its 103,000 employees. One big reason, says Anders Hemre, the company's chief knowledge officer, is that the employees, like workers everywhere, tend to rely on personal networks for information, rather than on a central data repository.
Despite advances in technology, often it's simply easier to ask the person in the next cube about a project, schedule, or what have you. Trouble is, what's efficient for one person at one moment is not necessarily what's best for the organization as a whole. No matter how bright the person next door may be, the explosion in information of all types guarantees that the shoulder-tapping method will often yield less than optimum results.
So last year, when Hemre was charged with improving the flow of internal information, he modeled his approach on a very simple structure: the lunch table. Believing that casual, give-and-take lunchtime conversations on current issues generate the most useful information, Hemre spent a year working with a consultant to organize extradepartmental "communities of practice" that would draw together people with similar business interests for loosely structured brainstorming and dialoguing sessions.
Then he added a technologic component. Organik, a searchable database organized around user-submitted questions and answers, provides a place for the groups to log insights that emerge from the meetings. "They go through face-to-face discussions about the issues first, then we take them to technology," says Hemre. Users can later query the database with natural-language questions and receive answers to similar questions. They can also find people Organik has identified as subject experts based on their previous responses.
Why Make More Work for Yourself?
The idea that electronically stored information should follow, rather than change, human interaction is one important development in the effort to help people cope with the oft-lamented "information overload" problem. "Too many people who do knowledge management create extra work as they add information" to the corporate storehouse, says John Seely Brown, chief scientist at Xerox Corp. and a leading expert on search technology. "Technology should be a byproduct of what you're already doing."
While Ericsson's program is only in the pilot stage, Hemre says that "we [already] see a big difference between this and asking people to post their documents on a shared server, because this leads to better dialogue." He says the next step will be to integrate sources like an intranet and business intelligence tools into Organik, so that employees don't miss valuable information that resides elsewhere.
Companies have been steadily increasing their efforts to leverage all the information that exists within their walls, whether in databases, on the Web, or in employees' minds. Collaborative applications accounted for nearly 40 percent of knowledge management software revenues in 1999, according to IDC, outpacing sales of content management and data warehousing products. Often the technology is nothing fancy, but is deployed in new ways.
During a recent series of Web-based meetings, for example, Agilent Corp. CFO Robert Walker made sure participants had the chance to "chat" online, just as they would have had they met in person. "That's where information gets exchanged," he says. Participants could type comments and questions to one another during the management presentation, the equivalent of "whispering in the back of the room," Walker says, and with management's blessing. Walker made it a point to observe the "fascinating side conversations" that took place, giving him insight into employees' reactions.
Making Data More Relevant
These approaches work, experts say, because they go beyond making data available — they make it relevant. That principle is guiding a number of technology developments. Search engine enhancer Outride, for instance, which spun out from Xerox's Palo Alto (Calif.) Research Center (PARC) last December (and has since been acquired by Google), acts like a helpful librarian, refining and expanding searches based on a user's search history. The software "reads" text documents (anything from Web pages to news feeds to internal E-mails), summarizes and categorizes them, then delivers the most pertinent ones to the user. Inxight, another PARC spinout, adds a visual component; its Hyberbolic Tree software allows word-weary users to turn text into maps, thus making Web sites more navigable and data easier to interpret.
Another product that emphasizes visual presentation is ClearForest. It reconfigures news feeds and other data into whirlpool-like arrangements in which color-coded arrows lead the reader from a main topic, such as a competitor, to a related topic, such as companies the competitor is acquiring, with one-sentence summaries for each topic. Dow Chemical Corp. and Eastman Kodak Co. turned to the software to help them sift through reams of patent documentation as they searched for technical experts for hire.
Last year, Credit Suisse began to offer the product to its high-net-worth investors in Italy, and will soon bundle it with in-house research and roll it out to customers across Europe. "We see it as a competitive advantage," says David Chinn, marketing director for Creditsuisse.net. "Clients will be able to quickly search through news" and make better investment decisions.
Reader Comments» Post a comment