The support team saw a spike in calls about a new issue.
It was a tricky one. Most agents needed 30 minutes to solve it. Stephanie figured out how to fix the issue in just five minutes. The other agents would benefit from Stephanie's solution, but there wasn't a clear way to share it.
Knowledge gaps like this are a big challenge for contact center leaders. Agents spend too much time searching and struggle to give the right answer. Fixing these gaps could improve agent morale, improve contact center efficiency, and keep customers happier.
I partnered with USU's Hugo Ramadier to host a virtual "Ask Me Anything" to answer leaders' toughest questions about knowledge management in the contact center. The questions spanned everything from knowledge management best practices to how AI can help.
One thing was clear: contact center leaders want an AI-ready knowledge strategy. Most contact centers will be using AI for knowledge management if they aren't already.
This article will help you build that strategy.
Your goal should be: get the right answer, quickly.
This goal enables an AI-ready knowledge strategy. Now you can work backwards and decide, "How do we get there?"
One client used this reframe to cut new hire training time by 50%. The old program focused on getting agents to memorize a slew of complex details about products and procedures. Agents struggled to remember the correct answer and froze when it came time to answer live phone calls.
Reframing to "get the right answer, quickly" led my client to implement a knowledge tool that eliminated memorization and allowed agents to quickly pull information from the correct source. Their confidence soared and so did their right answers.
Once you reframe your goal, it's time to build your plan.
We asked our webinar participants to rank their knowledge management priorities. The group consensus reflected a healthy approach:
• Accuracy: keeping knowledge accurate
• Governance: managing knowledge
• Training: using knowledge to improve performance
• Technology: finding knowledge management solutions
This is the exact order I recommend you address your knowledge management plan. Let's start with accuracy.
You need accurate knowledge before your strategy is AI-ready. AI will give wrong answers if it's pulling from a source that's incorrect. (Human agents will, too.)
Our webinar participants were asked to weigh in on who was more likely to give the wrong answer to a customer: a human agent or AI. The group provided a sage response:
• AI: 14%
• Human: 14%
• Both equally likely/unlikely: 71%
There are two knowledge base best practices that can help improve your accuracy.
The first is to use one source of information whenever possible. Knowledge gaps get amplified when different systems contain conflicting or out-of-date information.
I once managed a training team where it took three months to get the training knowledge base updated. We fixed the problem by revising our training to use the same knowledge base that agents used when handling live contacts.
The next best practice is to create a governance plan.
You need a solid governance plan if you want to keep your data accurate and up-to-date. Knowledge governance consists of a few critical steps:
• Keeping information accurate and up-to-date
• Adding new knowledge as it becomes available
• Making sure knowledge is easy to access when needed
The biggest question is, "Who owns governance?"
We asked our webinar participants who owned knowledge in their contact center. The most popular answer was "everyone," which is a potential red flag. It's hard to instill accountability without a clear owner.
The best approach depends on your situation.
A centralized approach makes a specific person or team responsible for managing knowledge. This works well for large amounts of ever-changing knowledge that requires a lot of work to manage.
A role-based approach assigns ownership of different types of knowledge to different teams. For example, product keeps product information updated, training updates procedures, managers keep policies up-to-date. This works well when teams have well-established roles.
A hybrid approach has a central owner managing the knowledge system combined with role-based experts who keep their part of the knowledge base updated.
For smaller teams, governance can be an opportunity to engage agents. Assign each agent a knowledge domain and make them responsible for keeping it updated. I did this when I managed a small contact center and agents loved it.
Now it's time to use knowledge to unlock agent performance. This is where the "quickly" in "get the right answer, quickly" comes into play.
Train agents to use their knowledge management tools rather than rely on memory alone. This empowers them to answer nearly any question quickly so long as the answer is available via the tool.
Start in new hire training. Design your training around scenarios your agents will actually encounter. Order them from simplest to most complex. In each scenario, challenge agents to use their knowledge management tools to find the correct procedure, policy, or answer to serve their customer.
It starts slow at first, but agents rapidly gain speed and confidence as they become adept at using their knowledge tools.
I've helped several teams reduce new hire training by 50% using this approach. The best part was agents had more confidence and performed better when they learned to rely on knowledge tools, and not their memories.
The focus on performance should continue after initial training. Embed reminders to use knowledge tools into your existing agent feedback systems:
• Quality assurance monitoring: was the correct answer given, quickly?
• Quick questions from agents: share the source, not the answer
• One-on-ones: include knowledge management as part of regular discussions
Your knowledge strategy is AI-ready if you've made it this far. Your knowledge is accurate, there's a clear governance plan, and your agents rely on their tools rather than their memories.
Our webinar participants asked so many great questions that we couldn't answer them all during the live event. Here are some additional questions and answers.
The companies that are getting knowledge right follow these steps, regardless of tools or technology:
Training gets more effective the closer it resembles the actual work. The best contact center training programs lean into that.
In practice, that means designing contact center training around remote, one-on-one customer interactions as much as possible. Remote and hybrid teams already do this. Onsite teams can do it, too.
One contact center training team took this approach to the next level for onsite agent training. They eliminated virtually all classroom training and had agents attend training from their workstations. Some training was still done in a group setting via video conferencing, but agents quickly learned to use the tools at their desk to solve problems and help customers.
Agents typically over-rely on escalation and support team members when they worry about making the wrong decision. Boost their confidence by providing clear, well-documented empowerment procedures. For example, one company allowed agents to give customers account credits of up to $1,000 without getting approval from a supervisor. Agents were given clear guidelines to follow that helped them determine when to give a credit and how much was warranted.
Every credit above $250 was audited. Agents couldn't get in trouble for issuing a credit within the guidelines, but the audit process was designed to help identify any inconsistencies or opportunities for calibration.
AI is a powerful tool that can work wonders for knowledge management. But it isn't a strategy. Get your knowledge house in order before implementing AI, and you'll be positioned to take full advantage of what it can do for your contact center.
USU Customer Service Knowledge Management (CSKM) is a knowledge empowerment platform purpose-built for customer service teams with 50 or more agents. It turns scattered content into one reliable source of truth for agents, customers, and AI tools.
USU CSKM includes built-in editorial workflows and approval processes to support the governance approach described in this article, along with intelligent search, decision trees for complex procedures, and native integrations with Salesforce, Zendesk, and Intercom. Knowledge analytics let you measure what's working and spot gaps before they become customer-facing problems.
The platform's AI capabilities are designed around a controlled architecture, meaning AI-generated answers draw only from your approved, governed knowledge, keeping responses accurate and compliant.
Teams using USU CSKM have reported 50% faster search times, 80% shorter training periods, and 70% higher first-contact resolution rates. The platform deploys in as few as 8 weeks without heavy IT involvement, and teams typically see 90% adoption within the first week.
Learn more at usu.com/en/knowledge-management
Not sure where your knowledge gaps are? USU offers a free knowledge audit to help contact center leaders assess the current state of their knowledge, identify opportunities for improvement, and build a roadmap toward an AI-ready strategy.