The latest insights from your peers on the latest in Enterprise IT, straight to your inbox.
Advances in artificial intelligence are making call center conversations more human.
By Chris McGugan, Senior Vice President, Solutions & Technology, Avaya
These days, the call center is one of the most important elements of a company’s brand. In fact, it’s often the last line of defense in protecting the brand. Most of the time that your customers are interacting with your products things are working fine. But if something goes sideways or they have a question on your product or service, they end up making that phone call.
Now, when you call in, even if you punch zero, you may still be talking to a computer. Voice traffic is leveraging artificial intelligence and machine learning in a lot of ways, and now we’re using AI not to automate the call, but to make that call a better experience for the customer and a more profitable one for the company.
Another set of ears
Five years ago, the most a contact center system could do was look at a caller’s phone number, which indicated where the person was calling from and gave you some insight into their language preference. More recently, we’ve been able to match the number against our CRM system. This means we not only know that the caller is, say, an English speaker, but we may also know that, because of their account with us, they’re a high net worth individual who should be routed to a particularly skilled agent.
And today, we can do even more. With the advances in artificial intelligence and the speed of computation, we’re starting to use real-time AI in the call center. An AI agent “listening” to a call can provide real-time on-screen coaching to the human agent. For example, it can consider the sentiment in a customer’s word choices and alert the agent that the temperature of the conversation is rising. Or if the client looks like a good prospect, it may suggest upselling options.
The system also helps with fraud detection, because you can learn a lot from the language of an individual who may know just enough about a person to impersonate them, but not enough to really be them. All those sorts of things are absolutely top of mind for our contact center operators today; we’re going way above and beyond basic experience management.
The call center as strategic tool
Our clients are starting to use this new technology to connect all the dots. An insurance company I work with, for instance, is mining all of its existing policy information and call recordings, and then building a body of policyholder information around what factors look good and what looks scary.
In fact, many companies are paying much more attention to their contact centers and exerting more control over them. That’s because they’re no longer simply a “call center,” they’re an interaction center. And they’re so important to brand management and sales that the business side of the equation has started to take more control of their budget. It’s similar to the evolution of IT over the last decade, where we moved from IT having its own budget to a client-funded model.
Companies are right to make that change. The fact is that the amount of speech analytics information captured by most contact centers today is breathtaking. Our ability to humanize AI assistance is getting better and better every day. You don’t have the weird disconnects, you don’t have the slow handoffs, the interaction is becoming much more natural.
Getting comfortable with AI
People are also getting more used to these interactions. I’m amazed at how quickly digital assistants like Alexa are becoming part of our lives. We have seven of the devices throughout our house, and for our children, it’s very natural for them to wake up first thing in the morning and say, “Good morning, Alexa,” and have her tell them the day’s news and weather. (My wife, though, isn’t so chummy with Alexa yet. In fact, she finds her creepy and calls her “Miss A,” so she won’t wake up when we’re speaking about her.)
A similar process has happened with call centers. Take a large financial company we work with. They can take three billion calls a year, and they capture over well over 80% of their calls through interactive voice response. It’s the highest rate in the industry. They still take a lot of human calls, but they’ve built an incredibly easy-to-live-with automation platform.
I think when automation feels natural, consumers feel comfortable. It’s kind of like a video conference: If I have never met you, a video conference feels weird. But once I have a relationship with you, we could be on a video conference all day long.
On the operations side too, deploying the technology is also getting easier. The fact that it’s cloud-based generally means your time-to-deploy is shortened. If I can bring it in and test it with ten or 20 agents before I deploy for the whole center, that’s win-win-win. I can avoid mass disruption, I can avoid mass training, and I’m able to introduce the new technology in a very simple manner.
Now that we have such powerful tools available, the main imperative for companies is to capture as much data as they can. The more information you have that you can make decisions on, the better outcomes you’re going to get, and the more effective you can be.