How might machine learning tools evaluating emotion help with call center problems?

In many ways, the virtual assistance and interactive voice response (IVR) systems in today's call centers are still pretty primitive.

However, brand-new technologies are coming along that promise to really revolutionize the phone call as an emotional means of communication. That can have big effects for all sorts of customers and other callers who are trying to interact with a business over the phone.

It's hard to square some of these new technologies with the current system that predominates, where the robot voice on your phone really doesn't do any analysis of your end of the call, beyond handling the basic natural language processing that distinguishes one target phrase from another.

When you look at what's coming down the pike, it appears that we’re due for a quick explosion of modern artificial intelligence assistance that will make IVR systems much more responsive.

A Wired article called “This Call May Be Monitored for Tone and Emotion” chronicles the emergence of a program called Cogito that is a prime example of how voice analysis software is about to get much better.

One of the biggest takeaways for those who are interested in customer service is that Cogito can help to discern a caller’s state of mind and mentality. That can help companies to better serve callers who are already frustrated or irate, and who are likely to get more frustrated and irate as they deal with voice menus. For example, writer Tom Simonite addresses the use of evaluation of phrases like “this is ridiculous” that can clearly show technologies whether a caller is becoming aggravated by the interface that he or she is using.

In a somewhat more sinister use, the software can detect the emotional patterns of the employees as well. This can also improve customer service, although it tends to do it at the expense of the human worker – in other words, software that compels you to sit up straight and speak more pleasantly is probably not going to be a great union sell.

However, the potential here is evident – by looking at the emotional makeup of the caller’s voice patterns and not just the natural language, call centers can become worlds more responsive and really serve customers who want to figure out their bill, better understand product development, or ask safety questions.

In the end, this is all about taking an interface mired in hidebound architectures and making it something that's easy for people to use. Interface disorientation is a huge challenge in today's tech world – your customer experience is only as good as your interface. When these new tools really come to market, companies will race to adopt them, because the better user experience will give them a huge competitive edge. We are likely to see a sea change toward these new technologies that’s much more rapid and decisive than, say, cloud adoption, because in the end, new voice response technologies make a whole lot of sense.