A New Knowledge Management Paradigm
For companies that sell into a competitive market, purchase decisions often hinge upon customer service quality. Differentiating on service levels in the face of rapid growth, however, especially when acquisitions are involved, can be challenging. Service call volumes increase, and the company must integrate the support operations of acquired companies, all while improving support delivery and customer satisfaction.
Other challenges also exist--for many industries, customers are very technically savvy and prefer having the resources available to solve problems themselves. From a customer satisfaction and an efficiency standpoint, these self-service options become even more critical.
An interesting example: Due to a series of strategic acquisitions, a company was maintaining multiple incongruous knowledge base systems. Different interfaces made for inconsistent user experiences, and maintenance was a challenge. The company was growing, but risked increased customer dissatisfaction if it did not consolidate its significant knowledge assets into one integrated system and user experience, and empower customers to be more successful and self-sufficient with their products.
The company took this as an opportunity to change its knowledge management (KM) paradigm. Below, are the steps the company took to consolidate its KM systems and improve Web self-service to deliver a better, more productive user experience.
Consolidate Knowledge Assets
At the project's start, the company had several knowledge bases due to recent acquisitions. Combining all knowledge assets, with their respective tags and classifications, into one system would have been painful. It selected a new system based on natural language processing, capable of both understanding customer service requests for their underlying intent, and of indexing knowledge assets for their actual meaning. This eliminated the need to tag each content item and pass structured field data into the search engine to retrieve relevant results. Retrieval is driven less by a manual, difficult to maintain classification structure, and more by a robust, flexible language model.
Building the integration into the call tracking system was relatively straightforward. Now, when call center representatives click the knowledge base button, they immediately retrieve a set of relevant search results, and can link specific knowledge base articles to the case. Or, if the system does not return the right information, the agent can author a new knowledge base article and submit it to an automated workflow for review and approval.
Most companies try to document every case that customers submit, but that really does not measure anything about what customers actually need; it just reflects the incoming calls rather than the changing needs of its customers. Taking the KM evolution a bit further, the company leveraged an intent-based analytics module to enable deeper analysis of customer searches and inquires--how they cluster, in terms of frequency, what percentage retrieves relevant results, et cetera. This insight allows the company to quickly focus on the issues in which customers are most interested, and then adjust the front or the back end accordingly to address customer needs.
Provide the Right Search Experience to Users
Consolidating information into one system represented a first step toward empowering customers; however, the challenge to retrieve relevant information from the system and present the right self-service experience still remained. Many customers rely on search to find the information they need to resolve their problems, but a challenge is the frequency of search terms and the density of those terms within documents and knowledge base articles. Conventional statistical-based search engines struggle with returning relevant results when much of the indexed data is similar or concentrated.
The new system provides feedback to the user when it detects that not enough search terms are provided or the search relevancy scores are too low. The system triggers a message explaining that he will retrieve better results if he is more explicit with what he needs. This often surprises many customers, who have been conditioned by today's Internet search engines to expect less relevant results as the length of their search query grows--they feel they have to boil it down to two to three words. This new support model actually understands customers' search requests for their true meaning.
Part of that effort includes providing a user interface that integrates search and dynamic navigation into one user experience to give users the flexibility to choose the method by which they interact with the self-service knowledge base. Users can be explicit with their queries, and trust the system to understand the request and respond appropriately. Or they can enter keyword searches, and rely on the dynamic navigation links to refine or expand what information is returned to the user.
Monitor User Behavior and Analyze Knowledge Effectiveness
Web users are becoming more savvy and this is particularly true for users in more technical fields. Companies must adjust accordingly--user feedback and behavior can give you the tools to create a responsive, intelligent support organization, but it requires a paradigm shift in how you manage knowledge throughout the enterprise. Applying these insights to identify and resolve knowledge gaps will consequently improve the overall customer experience, and move your customer service a step closer to the intelligence of your customer.
About the Author
Jason Hekl is senior director of marketing at InQuira. Please visit www.inquira.com