Searching for results
The company sought an enterprise knowledge management platform compatible with Remedy Service Desk from BMC Software, which the company used for incident management, and one with a strong search capability. The company selected the Unified Knowledge Platform from RightAnswers to provide agents at its four technical support centers with high-impact, searchable knowledge solutions.
RightAnswers delivers cloud-based knowledge management and self-service solutions for customer service and IT support operations. From a content perspective, RightAnswers provides knowledge for technology support, particularly for desktop software such as Outlook, but including operating systems, Oracle, Citrix, smart phones and other technology used by the great majority of office workers. "We ship the product preloaded with relevant content tailored to each customer," says Simon Yelsky, VP of product development at RightAnswers, "but we also provide a structured method for customers to add their own articles."
The natural language open source search engine features RightAnswers' algorithms. "Our search tool bridges the gap between a single search box that might pull up too many articles, and a guided Q&A approach that might not be effective when the knowledgebase contains a lot of content," Yelsky explains. "To improve the accuracy and relevance of search results, the system offers guidance to users by asking them to provide additional search terms."
Presentation of content plays a large role in achieving search efficiency. "An agent who has to answer a question in three minutes cannot read a 10-page document," Yelsky explains. "If a document is used a lot, breaking it down into smaller chunks makes sense."
Agents can report situations in which they did not find an answer, and can make recommendations for a new article to be created. Users also have the opportunity to evaluate the usefulness of search sessions. That type of feedback is essential for improving search effectiveness.
A continual effort
The Unified Knowledge Platform allowed the office supply company to reduce its response time by 50 percent because answers could be found more quickly. Time spent on calls was reduced by 30 percent, and 35 to 45 percent of calls are directly impacted by the knowledgebase. More incidents are resolved during the first call, and Level 1 training time was reduced by 66 percent.
Effective search requires a thorough understanding of the user as well as content that is correctly organized, according to Seth Earley, CEO of Earley & Associates, a consulting company specializing in content management and search strategy. "Developers should consider the attributes that users will care about or want to search on," he says. Anticipating their needs is essential because it helps narrow the search universe. "Language is ambiguous, and users' search terms are not precise. They are an approximation of their needs," Earley says, "so having a mental model of the user is critical."
The most successful enterprises focus on regular user testing of search, according to Gingras, to be sure that the results are relevant to the users' needs. "In general, search technology is mature and effective," he says, "but it is not a one-size-fits-all, and it's important to view it as an ongoing effort rather than as a project that is over when the technology is deployed."