Virtual Assistants Make Speech Technology Relevant Worldwide
For decades, those of us in the speech technology world have tried to demonstrate the relevancy of our industry. It’s been a long progression from the early days of replacing touch tone input on interactive voice response (IVR) units to the speech-driven applications we see today. But with the advancements in deep neural networks (DNNs), artificial intelligence (AI), Big Data, and related technologies, intelligent virtual assistants (IVAs), chatbots, and in-home VA appliances are now ensuring speech technology’s place in history.
As former editor in chief David Myron pointed out in the Editor’s Letter for Speech Technology’s winter 2016 issue, there has been no shortage of innovation, and thanks to the technological leaps noted above, people are more aware and appreciative of speech technology. You are only as good as your worst application, and for years, speech-enabled IVR had a terrible reputation. But this has changed; across speech-enabled, self-help customer interaction channels, there are solid examples of speech apps doing a great job, enhancing the customer experience and driving real business results for enterprises.
Furthermore, though speech-enabled IVRs quickly gained traction in the United States, the more sophisticated and conversational IVAs and bots are springing up just as fast globally. Take Greece, for example, where all four of the major banks and all three mobile operators have deployed conversational self-service in the contact center, and some are piloting conversational social messengers and IVAs. For example, Eurobank has served more than 4 million customers since the debut of Omilia's conversational self-service solution. The solution exceeded numerous technical metrics, notably achieving greater than 94 percent semantic accuracy. Most importantly, the bank experienced human agent cost savings of approximately three euros for every one spent and recorded 85 percent customer satisfaction with its conversational self-service function. What’s more, 82 percent of customers said that they’d recommend using this platform.
Another great example can be found at the Radisson Blu Edwardian Hotels in London. Using Aspect Software technology, the hotel chain launched a “virtual host” service for hotel desks. Available at 12 Edwardian Hotels, “Edward” is an interactive text response (ITR) mobile SMS concierge service that was launched as a fun and personalized way for hotel guests to interact with the brand and have their service requests handled. Edward greets guests, remembers previous interactions if they’ve contacted him before, and allows them to use natural, conversational language to do numerous things, including checking and requesting hotel amenities—such as towels or room service, getting information about local bars and restaurants, and even expressing complaints—simply by sending a text message. Edward also is capable of having a dialogue with a customer and incorporates previous context into the dialogue. He even occasionally shows a sense of humor.
Two months after Edward’s launch, Aspect combed through the recorded interactions, categorized them, and found some pretty interesting results. Seventy-five percent of the sentences texted to Edward could be deflected from the front desk staff. Most of these (59 percent of the total) were divided between pleasantries that Edward could respond to automatically, general FAQs that could evoke a canned response, and personalized account questions that required a dip into CRM data associated with that phone number. Sixteen percent were service requests that could be forwarded to appropriate staff members, such as housekeeping or maintenance, without involving the front desk staff. Thirteen percent were not autonomous and referred to other texts, requiring context.
Then there is the work done globally by Nuance Communications with its Nina IVA. As just one example, within three months of deploying Nina at Swedbank, the IVA was handling 30,000 interactions a month, with a first contact resolution rate of 78 percent. That number has swelled to more than 50,000 per month today, the equivalent of 30 live agents.
This is just a smattering of examples. IVAs can provide a rich channel of customer interaction that in many cases can handle complete and complex customer inquiries, and they are getting better over time. With customer acceptance and usage of speech interfaces at an all-time high, expect to see increasingly capable and interesting assistants.
Nancy Jamison is a principal analyst in customer contact at Frost & Sullivan. She can be reached at firstname.lastname@example.org; follow her on Twitter @NancyJami.
Chatbots and IVAs get all the ink, but RPA applications are vital to a smooth contact center operation