May 20, 2021
Stay in the Know
The Evolution of Speech Technologies and the Influence on CX
A Conversation with Chris Hodges & Ellwood Neuer
Part 1: The Evolution and Acceptance of Speech Technologies
Speech Technologies are entering the mainstream, from IVR to Analytics, with an explosion in the adoption rate over the last year. The global IVR market alone is expected to grow by USD 2.5+ billion in the next four years.
Noble Systems’ Chris Hodges (SVP sales and marketing) and Ellwood Neuer (SVP solution engineering) recently sat down to discuss the growing use of speech and artificial intelligence, both at the consumer level and in contact centers.
When people think of speech as it relates to customer engagement, IVR is the first thing that comes to mind. But IVR has gotten a bad reputation over the years. What’s created that negative perception?
Ellwood Neuer: I think what created the negative perception is poorly designed IVR. We’ve all been in the “IVR jail” where you’re being asked the same questions repeatedly. They’re just poorly designed IVRs, with no way out.
Chris Hodges: Add to that a lack CTI integration between the IVR and the system of record. So that if you do get to an agent, they have to repeat questions the consumer has already answered, because they can’t access the information.
Another challenge with IVR is that companies tend to morph them over the years. Oftentimes customers don’t even know what is in their IVRs until it’s time to redo all of them. There’s poor change management and poor control over where the routing goes, and whether or not it works.
That’s right. As companies add all these things in their IVR, they very seldom go back and do full end-to-end regression testing. So, you end up with menus that don’t go anywhere, or transfers fail or go to dead end points, and the consumer just gets hung up on. Bad design creates a bad IVR experience, which gives it a bad reputation.
How has speech technology evolved and how has that impacted perceptions around IVR?
Speech technologies have evolved dramatically. If you just look at the Android or iPhone device that most people have in their hand. Those devices have changed the perception of speech, as have tools like Amazon Alexa for your house. They’ve also changed people’s behavior. Consumers are more used to speaking into devices, and when they do, they are inherently speaking a little slower and a little clearer.
Companies can use that same ‘natural language’ technology instead of just using basic IVR. Now, they can understand and allow you to do things like, say “I would like to change my air flight from London, as well as book a new flight for Atlanta.” And it takes that conversation and natural language understanding and realizes that you need to do two or three things on that call. And then it routes you to self-serve or sends you to the right folks that can handle the transaction that needs to take place.
But it’s the technology that has changed and evolved into that natural language understanding that has really dramatically impacted everyone’s perception of speech recognition. It allows your IVR to be more conversational, so you could just answer the phone, “Thank you for calling Noble Systems. How can I help you?” And then the person calling in just responds in their own words, rather than pushing buttons. Based off of what they actually say, the IVR natural language engine understands and can determine what the next steps are, as opposed to being navigated through a bunch of trees to figure out what you want.
What impact have Siri, Alexa, Cortana and other speech recognition engines had on how businesses approach voice?
I would say that, adding on to the previous answer, there’s no question that those home voicebot type platforms – whether it’s your phone or your Amazon ECHO device – have given consumers the understanding that speech recognition generally works and can be used for simple questions and answers from a device. Using artificial intelligence has more recognition.
Those devices have definitely done everything you just said. But they’ve also changed people’s behavior and how they speak, whether they inherently realize it or not. That fine balance of talking to that AI, so it can understand everything you’re saying, so that it can determine how to help you.
I think that’s the real impact, right? And you know, decades ago understanding speech was a huge cost component for the amount of servers and stuff to do it. And nowadays, it’s in your phone and in your remote as a standard thing.
And from a consumer-level, the expectation is that I should be able to talk to technology and have it understand me and gather the information for me. So, when I call into an IVR that I can’t talk to, that’s a problem where businesses need to catch up, right? Consumers want to self-serve, they want to use the app on their device. They want to be able to ask the question and just get the answer automatically, without having to explain themselves to an agent. Because ultimately all they’re trying to ever get to is a resolution. Whether that resolution is service or sales, ultimately, they just want to be done.
How has the use of speech technology changed during the extraordinary events triggered by the Covid pandemic?
With the Covid effect, we all know that there was a huge impact the first 90 or 120 days as businesses scrambled to get work from home agents up and running. And they started to rely heavily on their self-serve technology to take the calls. And then they realized how important that element could be and that the more containment they got within their IVR, then of course it’s easier to serve that caller without having live bodies. And when there are issues with connectivity or availability of their workforce, the self-serve IVR becomes even more valuable.
So, I think that in the past 12 months, the number one thing is the digital transformation of businesses. You know, they got five or six years’ worth of digital transformation, and some say a decade’s worth, in that much shorter period. Due to the digital transformation of content, speech technology can help overcome the challenges of getting information to customers.
Because it’s one thing to have speech technology that someone can talk to. But the key is whether the technology has the ability to get to the information to resolve that person’s call. And perhaps before the pandemic, a company may not have all of their information and content accessible in a digital format. But certainly post-pandemic, many businesses have evolved to the point where now speech technology can get access to that content. So if we can understand the request, the content is usually available to get the information to the caller.
And obviously a big driver, especially early days in the pandemic, was the sheer volume of inbound calls into customer care centers, whether it’s for travel or hotel reservation, credit card payments, car payments, or what have you. That caused a gigantic spike in terms of folks calling into those lines, which created problematic speed of answer issues and customer satisfaction issues. And it became even more urgent to fix the IVRs, deploy new IVRs, and deploy speech technologies as quickly as possible, to help compensate for that.