AppDirect Blog

News & Updates

The Stale Knowledge Base Syndrome

By Patrick / Oct 17, 2014

More than half of knowledge management programs fail.

A decade ago, failure rates were as high as 80 percent . This idea of acquiring and retaining expertise was (and still is) a strategic tool for maximizing the intangible assets of an organization. However, despite advances in KM technology, making individual tacit knowledge (the knowledge inside people’s heads), explicit and accessible across all members of an organization—is complex.

WHY KM FAILS

So much research has been poured into understanding why knowledge management projects never seem to pan out the way they were intended to.  For one, knowledge bases are really hard to keep up to date. They are riddled with complex workflows and processes and take way too much additional effort for agents to contribute their learnings. That means that really important ideas, modifications and updates get pushed into the “I’ll write it later” pile, a dungeon of tacit ideas that will never see the light of day.

On top of the complexity of workflows – organizations exercise tight control over the creation and distribution of knowledge. Granted, we’re not hiring tech support agents to be Hemingways, but they should have a simpler way to share new solutions they discover with their team. The process of creating or even proposing a new knowledge base article is inefficient. If and when a KB article makes it into production, it will undergo edits, validations and approvals. Meanwhile, other agents struggle to solve a problem with a known answer!

KM expert, David Kay, makes a pretty compelling point in his article Technology Should Enable, Not Enforce :

"We give agents live phones, but we control them so much when it comes to written content."

We are controlling permissions to avoid the risk of making mistakes (I get that!) but collective knowledge and wisdom are suffering because of it. Knowledge bases get stale and outdated because they are not designed to empower real-time sharing between agents and that is a must if we want to resolve the expanding scope of customer problems.

Tech products are literally evolving in real time, new features are being added every second… Especially if you’re in the multi-vendor landscape, your knowledge base often can’t keep pace with all the possible combinations and permutations.

FROM KNOWLEDGE BASE TO A KNOWLEDGE NETWORK

A Knowledge Network is about tapping into many sources of knowledge rather than a single one

Our knowledge bases fail to capture the ever-changing corpus of knowledge that we need to support our customers. Knowledge can be locked anywhere from your ticketing system, user communities, people’s heads, emails, 1-1 conversations,  the web… The point is – we need to capture it and make it accessible to everyone on our team wherever it happens.

At AppHelp, we’re taking one small step for support teams, towards a giant leap for the customer support industry,  with our new product Reveal. Reveal allows agents to find, capture, share and collaborate on any content they discover on the open web.

Support teams that handle many multi-vendor requests  turn to the web over half of the time to find solutions to their customers’ problems. Yet despite constant algorithm updates aimed at improving the quality and relevancy of search results, search engines serve up far more results than one could possibly need.

The more time agents spend digging for solutions, the higher the risk of losing customers. We need a way to filter the noise on the web, and most importantly, an instant and simple way to share what we learn.

The point I’m trying to make here is not that we need to kill the knowledge base and rely solely on the web. It’s that we need to tap into many different sources of knowledge and we need to be able to capture it wherever it happens, with no additional effort. Because many tech support agents are already using the web to find answers, it’s a great place to start.

Interested in being an early adopter of Reveal ? Sign up here to request access.