Myth Vs. Reality
Enterprise Search is Not Knowledge

The myth is that enterprise search and knowledge management are “equivalent” technologies. It’s true that both can help customers and agents sift through large amounts of information. By facilitating searches across a wide variety of sources, enterprise search and knowledge management (KM) enable users to access knowledge hidden in the myriad of corporate systems, websites, applications and databases that store information.

But that’s where the similarities end. Search is a tool that can be used as a building block for a wide variety of solutions from knowledge management to e-commerce to browsing videos on YouTube. Knowledge management is a solution designed to link customers and call center agents with answers quickly and efficiently, to reduce support costs and enhance the customer experience.

The reality is that when it comes to helping an online customer find an answer or enabling an agent to resolve an issue in your call center, enterprise search is just one small piece of a much larger solution, and frankly, isn’t up to the job

Myth #1: There’s no difference.
Enterprise search finds information through keyword searches. But keyword searching has severe limitations because it treats search phrases as independent terms. For example, when you type in “MyProduct software download,” enterprise search looks for documents that contain any of these words. Typically, you get back a huge results list with many documents unrelated to what you are looking for. Think of a Google-style search—the results produced can be in the thousands, and many bear little relationship to your search phrase, burying useful results underneath a mountain of waste.

Some enterprise search systems try to prioritize the results, such as giving priority to documents that contain all three words, or where the words are close together. This can improve the results set, but still produces a significant number of off-target results. For agents, the information flood increases research and training time, and often requires them to use applications outside their desktop to find an answer. For customers, it’s just a lot of frustration and wasted time.

The reality: Search is a tool, KM is a solution. The problem is that enterprise search cannot understand the intent of your inquiry. Enterprise search focuses on keywords, because keywords are all it has. The search engine typically doesn’t know anything about the person asking the question, why they are asking the question or where the answers are stored.

This is where KM excels. KM solutions (such as InQuira’s) apply contextual information to find a relevant answer that matches the intent of the question as quickly as possible. For example, when a self-service customer asks a question, often they have logged in and the system knows what products they own or their level of technical expertise. It knows that product manuals are a good source of technical information and price books have information to answer pricing or discount questions.

Using natural language processing, business rules and ontologies to discover the “true intent” of each inquiry, KM refines the likely results so that only the best answers related to the intent are presented. This eliminates irrelevant documents that just happen to contain the three words “MyProduct,” “software,” and “download” that will be served up by enterprise search.

KM simplifies searching by recognizing not all users search in the same way. A user may ask “Latest software for MyProduct” or search for “MyProduct downloads,” “MyProduct newest version,” or “MyProduct software.” While enterprise search treats each of these as a different query—which means the results will be different each time—KM categorizes all these terms as part of the same intent. Analytics and rules treat all these phrases as the “software download intent.” Instead of searching all repositories and producing a massive results list, KM directs the search to the most relevant knowledge source, such as the download page within the support site or a MyProduct micro-site. The success of searching is further improved through integrated analytics that provide insight into which questions are not yielding a positive experience and where an answer cannot be found. This information helps you continuously refine and improve the user experience.

Advanced KM combines the intent discovery with “directed knowledge,” which is entirely missing from enterprise search technology. When a search request is equated with a specific intent, KM can present the user with a variety of predefined responses for that intent. For example, KM can offer a wizard to guide a customer through the process of picking the right model and precisely defining the nature of the problem. It can then deliver the most relevant troubleshooting procedure and accompany the answer with related knowledge, such as links to the appropriate user manual, a discussion forum of customer tips, in-context FAQs and even a special offer related to the inquiry. Again, analytics help to improve the experience by identifying the most frequently asked questions which are prime candidates for directed knowledge, such as the perennial favorite, “How do I reset my password?”

Myth #2: DIY knowledge management is easy.
So, enterprise search alone won’t cut it for delivering a great search experience. Then what about combining enterprise search with enterprise content management? After all, content management is all about documents and document control.

Unfortunately, content management provides few capabilities to increase the ability of an agent or customer to find an answer. As Forrester notes, companies get knowledge management wrong by expecting content management to serve the knowledge management function. That’s because these systems are all about getting content into the system, not out. Like enterprise search, content management generally offers only basic hierarchical and keyword searching—which is appropriate for content authors with domain expertise—but not so good in the call center or online where the user’s skill is at varying levels.

The reality: Search plus content does not equal knowledge. When you combine enterprise search and enterprise content management for do-it-yourself knowledge management, you are still not addressing the fundamental needs of your users—a quick route to the most relevant and accurate answer. Are piles and piles of documents presented in endless results lists really what you want to offer to your users?