News

War on information overload spawns growth

• By Colleen Frye
• October 23, 2002

The value proposition of information-retrieval technology has expanded dramatically from traditional search technology, and market growth over the next few years is predicted to reflect that, according to new research from Delphi Group, Boston.

While Delphi expects that, due to the economy, the information-retrieval market will show a 10% decline in 2002 (from $400 million in 2001), double-digit growth is expected through 2004, putting the market at a projected $540 million at that time.

Driving this growth is the continued information overload of unstructured data that Global 2000 companies face, as well as the effectiveness of advanced information-retrieval technologies. ''In this era of downsizing, managers are coming to the realization that wasting time trying to find things, either in internal document stores or on the Web itself, is a luxury they can't afford,'' said Hadley Reynolds, Delphi's information retrieval director of research. ''The newer information-retrieval technology suites are providing personalized alert systems, the ability to filter news, and a whole range of personalization and information-location aids that haven't been broadly available.''

In addition, noted Reynolds, employees across the enterprise can benefit from information-retrieval technology, which is ''the opposite of putting in an SAP or making a large investment in an enterprise application product that winds up being used by only a few people.''

Information-retrieval technology has its roots in computer science and artificial intelligence. While it may be heavily automated, it may still rely on humans -- librarians, in fact - in large part to get things done.

An information-retrieval system organizes, classifies, and retrieves unstructured and semistructured data that may reside in newsfeeds, Lotus Notes databases, Exchange public folders, enterprise applications like PeopleSoft and a range of other underlying sources. Delphi breaks down information-retrieval software into five categories: basic search, advanced search, taxonomy/classification, question answering/enterprise management and ontologies/topic maps. Delphi projects that advanced search and taxonomy/classification products will grow the fastest through 2004; the top market leader in both of those categories is Verity, followed by Autonomy.

Research indicates that advances in information-retrieval technology are having a particular impact within two important mainstream market segments: customer relationship management (CRM) and business intelligence (BI).

''The e-commerce application of information retrieval was not widely appreciated in the first generation of commercial Web sites, but it didn't take long before the proprietors of public sites learned by bitter experience that you have to have good search capability; the same is true for enterprise services,'' said Reynolds, particularly in the area of self-service CRM.

Meanwhile, information retrieval opens up business intelligence around content interaction.

Information-retrieval software, explained Reynolds, can identify where customers are finding the most confusion or asking the most questions, particularly in a service environment. Reynolds said there has been speculation about a blending of the BI/information-retrieval segment. But, he noted, ''it will probably be integrated through someone in the CRM world.''

Reynolds cautions that while information-retrieval technology does not add a lot of complexity to a technology infrastructure, an information-retrieval implementation does require a good deal of care and feeding. The setup itself typically necessitates subject-matter experts and an expertise in linguistics or library science, and ''the organizational changes and language changes are ongoing.''