Four Tips to Optimize Search in Your Knowledge Management System
Search is a critical component of any knowledge management system. Yet, the search function of many systems returns poor results, not because of poor technology, but rather because the search function didn’t anticipate the correct user behavior. The following four tips will help you optimize the search function in your knowledge management system to improve the results for your customers.
Tip 1: Focus On Problems, Not Solutions.
Most often, users do not know which knowledge management article contains their solution—only the problem they are facing. This is particularly common in self-service knowledge management. Thus, writing answer-centric content and optimizing the search engine to find and retrieve answers is a recipe for failure.
Knowledge management content needs to clearly state the problem, issue or question that the content addresses. In this way, when users search the knowledge management system by describing the problem, they are much more likely to find a solution. Knowledge management authors should re-think their approach to solution authoring—titling content with problem statements and describing the solution information in the body of the knowledge management article.
Tip 2: Make Less Do More.
The average knowledge management search query is about two words, which is usually insufficient to locate a specific resolution. This leads to far too many results, forcing the user to weed through excess information. While it is impossible to force the user to enter a more complete query, several best practices can be used to produce a better user experience.
First, the most relevant results for the keywords must be returned. To do this, the knowledge management system should employ weighting of certain article fields. Fields, such as titles and metadata, should be considered more heavily than body text. In this way, results at the top of a two-word query will contain both of these keywords in the most important sections of the article. Additionally, high-value knowledge articles can be tagged as being more important and will always rise to the top of the results list when found, helping ensure that key content will be returned at the top of the result set.
Second, users can be encouraged to provide additional information during the search session by providing visual cues to filter the results based on additional tags or by prompting the user to answer clarifying questions. Another option is to employ search-as-you-type or automatic search completion. As users type one or two words into the search box, they will see examples of more-specific search queries from which they can choose. If users see a query that matches their needs, they will choose the more specific search string and thereby retrieve more-specific results.
Finally, it is critical to leverage contextual information to filter results dynamically. Integration strategies now allow the knowledge management system to integrate to other components of the IT infrastructure. When integrated to other systems, the knowledge management system is aware of various user-profile elements, from location to contact history to current product ownership. By leveraging this information to proactively contextualize search results, the knowledge management system can provide more-personalized results. For example, a query for “ATM fees” can be translated to “ATM fees for premiere checking users in New York.”
Tip 3: This is What You Are Looking For.
Even if the article a user needs is ranked at the very top of the search results, it might as well not exist if the user cannot recognize its relevance. Several strategies can help users recognize a content item as the needed solution.
Descriptive titles. Article titles should be problem-centric, describing the issue the user is likely experiencing. These titles should be as descriptive as possible, using terminology likely to be used by the customer. Additionally, visual cues, such as ratings and content type identifiers, can help the user identify the right content item.
Relevant excerpts. The best way to communicate relevancy is to show users why the result matched their query. Contextual excerpts present the portion of the knowledge base article that best matches the query. This text snippet allows users to read a small portion of the article without clicking through to determine if the article will answer their question. Bolding keyword matches within the contextual excerpt will provide further information to users as to why the result was returned.
Focused article topics. Finally, a knowledgebase article should focus on a single problem and solution. Long documents covering a range of topics are not only more time consuming to read, but also are difficult to recognize as a solution to a particular issue.
Tip 4: Test Your Search.
No matter how well you construct the search function in your knowledge management system, it is worthless if you don’t test it to see if it returns the anticipated results. Testing should be done with real-life queries based on user input, not what you think your users are looking for. And, it should be done with the language of real users, not your knowledge experts.
Start with a typical query, such as ATM fees, then see if the content found is relevant to that query. Check that highly ranked articles actually match the query and that lower-ranked articles do not.
Search is an important tool, but even installing the best search engine will likely fail if its implementation is based on something other than the way your users think and talk. Follow these four simple tips to help make your search function the best it can be and something that optimizes your customer’s engagement with you.
KANA, a Verint Company, is a leading provider of cloud and on-premises customer service solutions. KANA helps global organizations—including many of the Fortune 500, mid-market businesses and public sector agencies—optimize their engagements with consistent and contextual customer journeys across agent, Web, social and mobile experiences. Using KANA solutions, organizations can reduce operational costs, increase resolution rates and improve brand loyalty. Learn more at www.kana.com.