7 Steps to Level Up Your Contact Center’s Internal Knowledge Base (IKB) with AI

A key challenge in contact centers today is the fact that information and processes are changing faster than ever. Company updates are more frequent. Price points and promotions change daily. Critical updates must be readily and easily accessible by agents who struggle to keep up with the change needed to deliver outstanding customer experiences.  

Even a well-maintained IKB only helps if an agent can find the right answer while a customer is still on the line. That’s the gap Companion Agent closes: as the evolution of real-time agent assist, this AI agent assist software listens to the conversation as it happens and automatically surfaces the exact knowledge-base article, policy, or answer an agent needs, live, mid-call, so the IKB becomes something agents are guided to rather than something they have to remember to search.

That’s where a dynamic internal knowledge base (IKB) comes into play, because the fact is, outdated knowledge bases are common across contact centers of any size. And this can cause a few issues:

• Many articles, news flashes, and company updates are rarely checked for accuracy or relevancy within the Knowledge Base
• Agents aren’t utilizing the IKB properly. For example, searches aren’t done on the IKB. Instead, agents look to their supervisors first rather than checking for updates in the tool 
• Customers receive inaccurate information, which can cause the customer to call back and possibly churn
  

As a result, it’s mission-critical to maintain and keep a company-wide IKB up-to-date. The IKB should be the agent’s go-to whenever they need help. By providing prompt and accurate information, the agent can do their job efficiently, be mindful of the customer’s time, and improve company credibility.  

In this blog, we’ll share seven tried and true ways to ensure your internal knowledge base is used to its full potential to enhance agent performance and deliver a world-class customer service experience.

We just released a new Guide on reinventing quality programs with Contact Center AI. It includes best practices from industry leaders, actionable insights and transformative use cases and KPIs.

Determine Who Is Using the IKB

Agents, supervisors, trainers, and operations managers are all trained to use the IKB, but how many of them are taking advantage of it on a day-to-day basis? The intention is to deliver valid information about a product or company easily while on customer calls, but are they continuing to use it for that reason? By understanding who is utilizing your IKB and on what topics, you can identify topics to create coaching content around.

Some tips to consider:

• Start with monthly audits of who is accessing the IKB. This creates a population of focus groups to better drive broader adoption. 
• Identify when agent adoption drops off, as in most cases, new agents are the heaviest users of the IKB. 
• Think of ways to re-engage tenured agents. As they grow more experienced, their product knowledge grows as well, and they may not see the need for an IKB. However, many updates and special articles are shared in the IKB, and due to lack of usage, they might provide outdated or inaccurate information.
• Determine if supervisors and trainers are also utilizing the IKB, as they are responsible for both promoting use for agents and ensuring the relevancy of the content.

Find Out How Often It’s Used, and How It’s Used

Tracking and monitoring what articles are visited and what keywords are being searched bring to light how the IKB is being used, and uncovers gaps that can be mitigated. This helps you clean up old material; surface relevant, but overlooked information; and better tailor the IKB to your organization's needs. 

Try asking these questions:

• Are keywords and inquiries being tracked that return no information to the user?
• Are users flagging assets that aren’t relevant to their search?
• What articles are the most heavily trafficked, and are they up to date? 
• Is there an opportunity to train agents to mitigate heavily trafficked topics?
  

Utilize Contact Center AI to Better Curate Your IKB

Contact center AI allows contact centers to analyze 100% of voice calls, which provides a massive breadth of insights into both the contents of the call and the agent’s performance. These insights are valuable in making data-driven decisions on what resources should be included throughout the IKB, what content should be displayed most prominently, and what areas require more targeted coaching outside of the knowledge base. 

Examples of insights you can gather from contact center AI to fuel a better IKB include:

• Are agents able to answer FAQs from customers?
• How often are agents putting customers on hold to check on how to handle requests?
• Are agents going through the proper customer verification processes?
• Are agents following mandatory compliance guidelines?  
  

By the way, IKBs are just one of the many AI-driven transformations you can make on quality programs in the modern contact center. Check out our new Guide, The Future of Quality Assurance in the Contact Center for more.

Test If the IKB is Intuitive

Layout and usability are key for driving adoption and continued use of an IKB. So tracking UX and UI metrics will provide the insights you need to improve overall intuition of the IKB. Questions to ask include:

• How accurate is search?
• What is the hyperlink structure?
• How is information categorized?
• Do articles give a clear explanation of the information?
• Do articles provide related information?
• How many articles does an agent need to visit to get the information they need?

Re-think How You Onboard and Train New IKB Users

Providing best practices for utilizing the IKB to its full potential will help drive adoption and retention. If the information being presented doesn’t match with how it’s being used in the live contact center environment, it’ll more likely turn agents away from using it and understanding the importance of having the IKB. Some ways to train agents on the IKB include:

• Asking agents to do a “treasure hunt” to find a certain piece of content.
• Walking agents through the major and subcategories.
• Providing examples and real-world use cases for using the IKB
• Creating a 360º feedback loop to better understand agent requirements and how they interact with the IKB.

Create a Process for How the IKB is Updated

The IKB should be treated just like an organization’s website, collateral, marketing, and social media. It needs to be consistent and on-message. Some things to consider:

• Does your organization look at a “shelf date” on the material published to the IKB?
• Is there a time frame for when material should be revisited for updates or removal from the IKB?
• Is information verified for accuracy on a regular basis?
• Is there a process for adding new content to the IKB?

Utilize Refresher Courses

Beyond initial training, refresher courses (sometimes called continuous improvement) help maintain agent skills and keep everyone on the same page. With contact center AI, refresher courses can be better tailored to individual agents, identifying areas of improvement and knowledge gaps. An IKB is a great place to promote refreshers, allowing organizations to present the material in digestible “micro” formats.

Put Your Knowledge Base to Work in Real Time with Companion Agent

All seven steps above make your IKB more accurate and better organized, but accuracy only pays off if agents actually use the knowledge base in the moment they need it. That’s where Companion Agent, Observe.AI’s real-time agent assist software, changes the equation. Instead of asking agents to pause a call, switch screens, and search for an article, Companion Agent listens to the live conversation, understands what the customer is asking, and pushes the matching knowledge-base content straight to the agent’s screen as the call is happening.

As part of Observe.AI’s purpose-built AI agents on one CX platform, Companion Agent turns a static IKB into a live coach for every agent, then summarizes the call afterward so supervisors can see exactly which knowledge was surfaced and used. That means less time hunting through folders, fewer callbacks caused by outdated answers, and a knowledge base that finally gets used the way it was designed to.

• Surfaces relevant IKB articles, scripts, and policies automatically, based on what the customer is saying in real time
• Eliminates manual searching, so agents get real-time agent assist instead of toggling between systems mid-call
• Summarizes the call afterward, highlighting which knowledge-base content was used so teams can keep the IKB accurate and relevant

Ready to learn more about contact center AI?

Check out our guide on The Future of Quality Assurance in the Contact Center, which includes transformational use cases, best practices from leaders in the space, and more to reinvent your quality program.

Or watch our most recent webinar on 5 Secrets for Building an Ultra-efficient Contact Center with Root Insurance and Talkdesk.

About the Author

Adrian Valenzuela leads onboarding at Observe.AI. Adrian has a track record of moving contact centers to the cloud in high-growth environments. Adrian has worked in implementation management, business process optimization, analytics, contact center operations, and customer success. Connect with him on LinkedIn or reach out to adrian@observe.ai.