Top Priorities for Investing in AI: Knowledge Management

When it comes to AI solutions for the contact center, we’re spoiled for choice. For organizations dabbling in AI for the first time, a common qualm is, Where on earth do I start?

Naturally, most firms gravitate towards the technology with the most easy-to-understand, quick-RoI value proposition: chatbots. They allegedly reduce call volume while providing 24/7 customer support without needing to keep the contact center open and staffed 365 days a year.

But even a cutting-edge chatbot can’t function without a sufficient knowledge base, which feeds the responses the bot uses to interact with customers.

“Knowledge management is a nexus point between customer questions and your brand’s answers,” Sedarius Tekara Perrotta, chief innovation officer at Shelf.io, said in a recent webinar with CCW Digital. “It’s that one technology that impacts different metrics across the organization because it’s a tool for your agents to do their job.”

In the CCW Market Study: Future of the Contact Center in 2019, improving the knowledge base ranked #2 out of 21 surveyed priorities for the year. 

More than just a knowledge base

Knowledge management is often misperceived as the back-end knowledge base that agents refer to when they’re searching for an answer to a customer question, or the customer-facing FAQ page and how-to library on your company’s website, when it’s really the heart of all agent- and customer-facing applications.

“You install an AI-based knowledge management system as the foundation, it can then become your self-service portal - you don’t need a separate technology,” explained Perrotta. “It connects to your chatbot and integrates with your IVR.”

The first step towards optimizing the customer and agent experience

This integration means something else for your organization: 360-degree analytics that span all of the different contact center applications, providing a dashboard of insights into what  information your agents are searching for, what your customers are doing on the self-service portal, IVR, chatbot and so on. On top of that, the AI interprets patterns of activity and makes recommendations for how to optimize each individual channel.

“When somebody clicks on a piece of information and doesn’t find what they are looking for, now you have the ability using the AI to recognize that pattern,” said Perrotta. You can see what queries customer enter most frequently into the search bar, what information took three clicks for them to locate (thereby raising their effort level), or if some missing infrastructure led them to call into your contact center because they couldn’t find what they were looking for.

The analytics also provide a readout on what language customers and agents use when searching the knowledge base, so that you can optimize your keywords and tags to make items easier to find. For example, say you have several items filed under “rebate policy” while customers search for the term “refund policy,” they might not find what they’re looking for.

A knowledge management system would flag that unreturned search and alert you to the misclassification so you are continually updating your knowledge base according to real-time user behavior.  

With AI and analytical insights, you can also keep your internal-facing knowledge bases current, rather than letting them become a source of frustration for agents when they’re on the phone with a customer and can’t find the latest information regarding a recent policy change, specs on a new product, or details of an ongoing promotion.

Analyzing behavior and optimizing information

A good AI-powered knowledge management system makes recommendations based on the analytics it collects on customer and agent behavior. “It will tell you exactly what you need to do to keep your organization’s entire knowledge base up-to-date, accurate and trusted,” said Perrotta.

This empowers agents to excel at their jobs, because if disgruntled customers aren’t calling or initiating web chats because they can’t find information themselves or the IVR misdirected them to the wrong department, agents no longer need to deal with exasperated customers and can post higher CSAT scores for the customer interactions they do have.

“It also reduces call volume because if you set up your chatbot and self-service correctly, a good percentage of customers can find what they’re looking for without any human interaction.”

The most important thing to understand when considering different AI technologies like chatbots, speech analytics, self-service portals is how to use them in tandem so that you communicate a consistent brand message across all channels, and how they integrate so you can truly create a 360-degree view of the customer.

To learn more, download your copy of our Special Report on Knowledge Management.