Adaptive Search and Resolution for Service and Support
If lowering costs, closing more issues at first contact, or increasing customer satisfaction are on your contact center's to-do list, chances are you have been considering a knowledge management suite, probably including business process support that can integrate with your CRM or incident management system.
Great search is key to achieving the benefits of such a solution—there's no point in managing knowledge if it's too hard to find—but most search solutions for service and support fall short for three reasons:
Search for service and support is different—In the standard search model, people know what they're looking for even if they're not sure how to ask for it. In service and support, customers have an issue, but they don't know how to describe the content that will resolve it. In a customer service environment, a natural language processing (NLP) search needs to be geared toward guiding people to a resolution even though they don't know what it will be.
Search must adapt to user queries—Search companies are often founded on a technology vision: once users have associative search, case-based reasoning, query intent, natural language, pattern matching or some other technology du jour, all will be well. In the real world, searchers search in a range of ways—they don't care about the technology. Effective search for customer service must be adaptive: it must recognize the nature of the query then process each query appropriately. This means it must have many search technologies available, and know when to use each one.
Search alone doesn't resolve all problems—In many cases, the right resolution is to do something for the user in a transactional application, or find the right expert, or guide them through a process. None of these situations is handled by a list of document results, or even any one specific document. They require integration, collaboration and business process support. In other words, the contact center requires more than adaptive search; it requires adaptive resolution where search is integrated into the business process.
Effective Service
Why are keyword search solutions that work so well for researchers unable to meet the challenges of service and support? Let's look at the key requirements that make search an effective component of service resolution.
Understand query intent. The concepts and wording a customer uses to describe a problem may be entirely disconnected from the focus of the resolution document, which tends to feature the root cause and fix. This is exacerbated by content authors' desire to focus on the positive (the solution) rather than the problem, even though the user only knows the problem. By focusing on the intent of the query, effective search will help the user bridge from the problem to the solution.
Support different search needs. At the start of a service resolution process, the user is focused on finding an existing answer. If that doesn't happen quickly, the focus shifts to finding analogous or related content that might suggest a debugging approach. Eventually, the user may move into a research process, where the required content is more about the system and diagnostic procedures and less about the specific problem. The system must support each of these types of search.
Match user and enterprise language. Technology, financial services, telecommunications and other specialized fields each use distinctive language. In part, the language barrier is just jargon—what a user calls the "blue screen of death," Microsoft calls a "stop error." But the differences can go deeper. Users and enterprises can describe problems using completely different concepts. In the customer service environment, search must bridge the gap between very different ways of saying things.
Support different search styles. A detailed examination of a contact center search log reveals there is no single way that people ask questions. Many enter vague one, two or three word queries; others use detailed keywords from the domain; a few use advanced Boolean searches; and a few others type in complete natural language questions. Some prefer to search, while others are more comfortable navigating through an information architecture. Search optimized for any single one of these query types will invariably fail to deliver good results for others.
Make context as important as content. If the searcher is an employee solving a customer's problem, or a customer that is registered with the support site, his or her context must be blended with the query to resolve the issue. This includes data in the user's profile, the CRM system, the billing system or other systems that contain relevant data.
Search and Beyond
Search IS a critical component of improving service resolution, but more is required. Search that drives effective resolution is powered by knowledge and supported with query intent, authoring, collaboration, business process and channel-specific applications. This is the basis of a Service Resolution Management (SRM) system and the result is seamless delivery of information at the right time, through any channel.
Knova Software is a leading provider of service resolution management applications that reduce service costs, increase revenues and improve customer satisfaction. Built on a next-generation search and knowledge management platform, Knova's suite of knowledge-empowered customer service applications automate the resolution process across multiple channels including contact centers, help desks, email, forums and self-service sites. For more information, visit
www.knova.com.