-->

Keep up with all of the essential KM news with a FREE subscription to KMWorld magazine. Find out more and subscribe today!

Web self-service: searching for answers

After an extensive evaluation of natural language search tools, Mentor Graphics chose InQuira. It did a better job of quickly finding the answers, and whisked the user directly to the relevant location inside large documents. It was deployed as part of Mentor’s self-service portal called SupportNet, as well as being available to its CSRs. As it turned out, the implementation was just the first step in Mentor Graphics’ launch into self-service.

Once the system was up and running, Mentor Graphics began analyzing the results. Five question areas accounted for 80 percent of the inquiries. The leading use of self-service was to find the answer to a technical question, followed by downloading new versions of software and accessing technical documentation.

"We found that more than half the time, SupportNet was answering the customer’s technical question," says Floodeen. "And although this was a considerable achievement as a first effort, we wanted to improve our success rate."

Mentor Graphics reassigned several of its senior analysts from call center duty to a new role of knowledge managers. Those experienced staff members examine topics for which SupportNet did not answer a customer’s inquiry and then bridge the gap by incorporating new information into the system. Now, more than 70 percent of customers’ questions are answered correctly. Content is tagged so that the user is directed toward the exact spot in a document that matches the inquiry. Targeting more specific information, along with continuous enhancement of content, has produced a dramatic change.

The launch of self-service has resulted in an increase in the number of questions by a factor of 10. "In the call center, we typically received about 5,000 calls per month," says Floodeen, "but we are getting over 50,000 questions per month on the Web site."

Customers have found it so easy to get answers from SupportNet that their need for assisted support has decreased greatly. In addition, the speed of response makes the system appealing, and beginners are able to ask very basic questions without embarrassment. The availability of self-service has lowered the number of questions directed at the call center, during a period of substantial growth in customer base and overall need for support. That has enabled the CSRs to focus on the more complex issues, as well as to invest more time in developing customer content for SupportNet.

Intelligent search is a critical element in the InQuira solution. "Actually understanding the meaning of the question is very important in producing results," says Jason Hekl, VP of corporate marketing at InQuira. "The same term can have different meanings in a different context."

InQuira’s ontology allows the search engine to focus on the intent of the searcher. Equally important, Hekl adds, is maintaining top-quality content. "Completing the loop to validate the accuracy of content and then seeing how it is used are both essential parts of effective self-service," he says.

Personalizing search

KNOVA 7, a customer relationship management (CRM) solution from Consona, is a suite of intelligent customer experience solutions that includes self-service, contact center and forums applications.

"Users need a highly personalized solution with sophisticated search technology that understands the concept and the context for their searches," says Martin Thompson, product marketing manager at KNOVA. "In addition, search capabilities should be linked to various other tools to help resolve issues as quickly as possible." For example, KNOVA includes an array of built-in processes such as wizards and guided search that can be tailored to provide the intended result.

An important part of KNOVA’s design, Thompson points out, is that the steps that help a user find an answer should also become a way of helping the system learn. "Methods of finding a resolution should be scored for effectiveness," he says, "so that they get priority when the user conducts a search."

New information captured on the fly during a self-service or agent-assisted session can be captured and incorporated into the solution. KNOVA also has forums that capture customer and agent knowledge. The information can be added to the knowledgebase, and can become a valuable component of the overall solution. The ability to use ratings and add customer knowledge are two elements that leverage Web 2.0 social networking features and work to enhance overall system performance.

Self-service Web sites can take pressure off CSRs and improve customer satisfaction, but success is not just a case of putting a new interface on a knowledgebase developed for internal use.

"The information should be structured and searchable in a way that is aligned with the consumer’s needs," says Jupiter Research's McGeary. "The knowledgebase is a constantly changing entity, and it needs human input by subject matter experts so the content can be validated."

To guide users to the information they need, McGeary recommends a blend of navigation and search so that the right content can be located in the most efficient way possible. But to a greater degree than other search applications, search on self-service Web sites depends on a deep understanding of meaning and intent. Therefore, a good natural language query ability and a strong underlying ontology are critical.

KMWorld Covers
Free
for qualified subscribers
Subscribe Now Current Issue Past Issues