Using Data to Create A Knowledgebase
Create a knowledgebase and expand it with each new question customers ask and you answer.
Create a knowledgebase and expand it with each new question customers ask and you answer.
Knowledgebases are an excellent way to provide information to customers on a self-serve basis, particularly for complex services such as tech support. If all you ask your visitors to do is sign in, then a frequently asked questions (FAQ) list is probably sufficient. However, if your visitors come to you with questions about installing a new sound card, upgrading the BIOS, installing a modem, or cleaning up after a virus, you’ll most likely need more than an FAQ.
Good knowledgebases are built over time. You shouldn’t have to anticipate every question that might come in but rather have the knowledgebase grow with each new question that is asked and answered. Dell, among others, gives customers access to the same knowledgebase that its tech support staff sees. That’s powerful motivation to savvy customers to try to go it alone rather than wait on hold to be walked through a solution.
The main problem with knowledgebases is that while they may be neural from the perspective of the company, they’re static from the perspective of the customer. A customer may have tried one or more solutions in a knowledgebase and knows that Solution A doesn’t work until you’ve applied a particular operating system patch discussed in Solution B. However, there’s no way for the customer to communicate that information to tech support, let alone to other customers.
When I was trying to get that nasty Nimda virus off my computer, I scoured McAfee.com for detailed instructions. There was a brief set of instructions, which turned out to be not entirely accurate, and there was no way to contact tech support. I could chat with a live agent about my account, but not about the virus. It made me wonder what business the company is in — account management or virus protection. I would have liked to edit the instructions it provided. First, I would tell customers not to delete the files manually, since that causes the virus to replicate itself more aggressively. Next, I would mention that if system files are infected, you need to reformat your hard drive and start from scratch. McAfee’s paltry directions made it sound so easy. It was really a chore.
Unfortunately, McAfee’s site doesn’t offer a knowledgebase. Even when sites do offer knowledgebases, they are usually missing a valuable opportunity to harness the knowledge of customers. Does a knowledgebase necessarily have to be limited to the knowledge in the company? Michael Heumann, president of KnowledgeFilter doesn’t think so. His product, KnowledgeCenter, permits sites to leverage the knowledge of visitors, “turning traffic into content.”
KnowledgeCenter typically begins with company-created content. Visitors are then permitted (and welcome) to comment on content, including rating it as good, bad, or neutral. Visitors can vote on other visitors’ comments. Good content and comments rise to the top. This helps filter worthless comments. The users decide what is useful and what’s not. The company can moderate content so anything that’s spam, offensive, or otherwise inappropriate won’t make it onto the site.
This neural approach to a knowledgebase has several advantages over a traditional knowledgebase. First, content is more valuable. Second, content is essentially self-creating, thus free. Third, content is constantly updated. For example, an answer to a question that assumes a particular version of a browser is being used will almost certainly be commented upon and corrected by another visitor. Fourth, the knowledge is always organized in such a way that all content about a particular topic is kept together; contrast this with your typical forum or threaded discussion group where the same topic comes repeatedly in a disjointed way. Finally, interactivity increases the stickiness of the site.
Whether you call this an organic knowledgebase, a neural network, or user-generated content, the result is the same: improved, less expensive, stickier content.
A living knowledgebase is ideal for either technical content or deep content — a wine rating site or homeopathy. KnowledgeCenter might be ahead of its time, but I’m ready for its functionally now.