While engines such as Google and Yahoo! have made Web searches for obscure companies easier over the past several months, Dale Kurschner, publisher-CEO of the news service, prefers Subjex.com, a "natural language" search that allows him to have a "conversation" with the search function.
"Jupiter surveys show that 80% of online users will abandon a site if the search functionality is poor. Businesses must zero in on search functionality on their properties in order to survive. Search should be the kernel, driving overall development and facilitating business processes that interface directly with the online audience. ..."
From "Search Technology: Resurrecting the Web's Workhorse," Jupiter Research, 2000
Developed by PageLab Network Inc., Minneapolis, Subjex is a software application for on-site Web searches, private databases and wireless device infrastructures. It narrows down the "frame" created for the search and specifies precisely what information is wanted and, perhaps more important, what is not. Users search for information using plain, everyday language.
"The way the system is set up, I can ask it something, and the response may be a little sassy, but in a good way," Kurschner said. "We try to make financial news fun, and Subjex.com brings in those added elements to subjects that can be very boring."
Andrew Hyder, CEO of Subjex, which launched in 1999, said the company helps b-to-b companies save time because its artificial intelligence system assists with questions and concerns buyers may have after the initial point of sale. The technology costs $25,000 to license, and clients include 3M Co. and ASP Corp.
"There is no learning curve," Hyder said. "People just talk it out. What Subjex.com will do is keep asking questions" until the user's question is answered.
Despite the necessity of search engines, most b-to-b Web sites fall short of offering efficient site search functions. A June 2000 Forrester Research Inc. study of 30 b-to-b Web sites found that 93% either didn't offer a keyword search at all or had search engines that failed basic tasks, including missing critical content and burying the most relevant results at the bottom of retrieval lists.