Building a People-First Knowledge Base
It’s happened to me more than once: someone asks me to get a new knowledge base going, and during the initial assessment we find an old knowledge base just sitting around gathering dust.
At one company where I worked, I was hired specifically because the VP for Customer Care realized how important it was to have a solid, well-managed knowledge base.
When I got into the job and started digging, the first thing I found was a “sort-of” KB that some SMEs had put together in Microsoft One-Note. Yeah, there was a lot of info in there, but it was out of date, inconsistent and impossible to find. You had to know exactly what you were looking for to find anything – not a great experience, especially for new employees.
I dug a little deeper and discovered that some team managers had made a half-hearted attempt five years before to use Salesforce Knowledge. That was a whole lot better than One Note. But nobody was using it, and the people who had started it up had long since left for other jobs.
I’ve seen this pattern repeated in my consulting work time and again. Someone in the past thought it would be a good idea to have a KB, they tried to build something and then it just … went away.
In all of the cases, the failure boiled down to a simple reason: no one was thinking about the people it served. They were just putting together a system to hold some knowledge artifacts.
But the secret to a successful Knowledge Base (KB) isn't just about the system or even the knowledge it holds; it has put people first. When you empower your team to own the knowledge they use and create, you unlock the true potential of your knowledge assets.
Here’s how to foster a human-centered KB:
Empower Knowledge Users: Your frontline agents are the key to success. They should be using the KB daily to solve customer issues and are often the first to spot knowledge gaps or inaccuracies. When you give them the tools to highlight gaps or areas for improvement, that creates a sense of ownership – meaning they’re more likely to use it.
Leverage Knowledge Contributors (SMEs): In an ideal world, every single agent can share their know-how; that’s the promise of frameworks like Knowledge Centered Service (KCS). In my experience implementing KM programs, I’ve found that Subject Matter Experts (SMEs) are your go-to people for "know-how" and tribal knowledge. They are the ones that should create and update articles directly in the KB as they work. This captures their valuable insights and frees them from repeatedly answering the same questions.
Support Knowledge Curators: Even when you have a healthy number of expert contributors, you still need folks who can curate all the submissions, to make sure that they meet specifications, aren’t duplicates and so on. Knowledge Curators work right beside Contributors, reviewing submissions, ensuring consistency with style guides, and maintaining the KB's overall health and findability. In fact, it’s usually best if your Curators are drawn from the ranks of the Knowledge Contributors.
Define the Knowledge Coordinator/Manager Role: Many companies who don’t have a knowledge management function are adverse to creating management-heavy teams to do the work. Still, it’s crucial to have at least one person who "owns" the KB, overseeing daily operations, monitoring its health, planning improvements, and serving as the primary point of contact for leadership and cross-functional teams. They are a change agent, not a gatekeeper, facilitating knowledge sharing across the organization.
Cultivate a Culture of Sharing: Encourage open communication, collaboration, and reward employees for their valuable contributions. Implement forums like communities of practice (or "Knowledge Cafés" or use channels like Slack/Teams for informal sharing and new knowledge flagging.
At the end of the day, it’s important to remember: knowledge management is something that you do for (and with) people, not to them. Following a collaborative approach ensures your KB will be something people actually like to use because it’s continuously fresh, relevant, and truly valuable.
KM News Notes
Cresta Unveils New "Generative AI for CX" Features to Proactively Assist Agents
When most people think about KM for Customer Service Contact Centers, they think of customer self-help chat bots. But AI-based agent assist is a key tool for contact center leaders. Recently, Cresta announced a significant expansion of its generative AI capabilities that moves beyond reactive assistance to proactive knowledge delivery for customer support agents. The updated Cresta Agent Assist will now automatically surface knowledge base articles, step-by-step guides, and real-time coaching suggestions based on the live context of a customer conversation, without the agent needing to manually search. This addresses a key challenge in knowledge management: ensuring agents can find the right information precisely when they need it, reducing hold times and improving first-contact resolution.
Salesforce Study Reveals AI's Impact on Customer Service Knowledge Gaps
A recent report from Salesforce provides data-backed evidence that AI is crucial for knowledge management success in customer support. One key finding: 81% of service agents believe AI is most valuable for its ability to automatically surface knowledge from various sources. The study shows that companies using AI-powered knowledge management tools saw a significant decrease in agent onboarding time and more consistency in case resolution.
Glean Launches "Workplace Search" Assistant to Unify Enterprise Knowledge for Support Teams
A big challenge for customer support teams in virtually every organization is having to find knowledge that’s trapped in silos across the company. Tech startup Glean has officially launched its new workplace assistant designed to solve this problem. The Glean assistant connects to over 100 different applications – like Zendesk, Slack, Jira, and Confluence – to create a single, unified search experience. When a support agent has a complex customer query, they can ask the Glean assistant one question and get an answer from resolved tickets, internal wikis, team chats and more. This approach drastically cuts down on search time and improves the quality of agent responses.
