Contact Center Solutions Featured Article

Coming Soon to Call Centers: Voice Analysis to Determine Customers' Moods

October 14, 2013

Call center agents have to be a lot of things to a lot of people: brand advocates, the targets of anger and frustration, the soothers of tempers, the source of education about a product or service. Sometimes, however, they find themselves acting as amateur psychologists. To be successful, they need to read voices and search out hidden language to determine how a customer is really feeling. Is he/she thanking the company and cutting the conversation short but really seething inside because the transaction is incomplete? Does the customer simply not understand the answer but is too embarrassed to admit it or risk looking stupid?

Call center work is hard work, and agents often have too many balls in the air as it is. This is why a technology called “computational voice analysis” could help call center agents in the near future determine what customers are saying – not from the words they speak, but from the way they speak.

The New York Times recently profiled the technology, highlighting developers such as Israeli company Beyond Verbal, which are using computational voice analysis to develop solutions for call centers and other customer services that seek to read and respond to consumers’ emotions in real time. Beyond Verbal reports that its software can detect 400 variations of different moods.

Image via Shutterstock

“It’s not what you say. It’s how you say it,” Dan Emodi, vice president of marketing at Beyond Verbal, told the Times. “Listening to these patterns, we can allow machines for the first time to understand the emotional side of our communications,” he added.

People say “thank you” a lot of ways. They may say it in a genuine, grateful way… and really mean it. They may say it as a rote phrase to end a conversation they’re not really interested in. They may say it sarcastically, to underscore to the other party that they’ve been anything but helpful. They may say it resentfully, assigning blame to the other. The human ear isn’t always attuned to determining the difference in vocal patterns to indicate the true meaning behind the words.

Some experts, however, worry about privacy with the technology. How many people would be comfortable knowing, for instance, that when they called their mortgage company, a computer is diving into their voice, looking for deep clues to their state of mind? Some analysts say that call centers choosing to use the technology will need to warn consumers, much the way they already do when recording a call. But after all, advanced speech analytics already mine calls for loud voices, voices speaking over one another and key phrases such as “close my account.” Is this much different?

We’ll find out soon, according to the Times article. Some companies report being ready to run with the technology by the end of this year. Want to try it out? Beyond Verbal allows you to test your mood via the spoken word on its app website.

Edited by Alisen Downey