Speech Continues to Advance in the Contact Center
They say businesses are only as good as the service they provide their customers. Hence, in an increasingly digital world where consumers are more accepting of customer self-service and non-human agents online and on the phone, it makes sense for companies to automate more of their customer service interactions, particularly with the aid of speech technology. To do otherwise risks frustration and abandonment from clientele seeking speedy help, experts agree.
Consider that the interactive voice response (IVR) systems that companies have been using for decades now are becoming more conversational with the help of artificial intelligence (AI) and speech technology advancements like voice-enabled chatbots. Intelligent virtual assistants (IVAs) can understand human speech in many languages and respond using text-to-speech, allowing customer service to authenticate callers via voice, look up orders, make appointments, accept payments, answer a variety of queries, and more. Speech-to-text voice transcription technology, long used to garner insights into call data, is now being increasingly employed for analytics to better understand customer needs and trends. And more businesses are relying on natural language understanding (NLU) to review and auto-respond to customer tickets.
These are just some of the speech technology innovations that have companies large and small abuzz with excitement as they look to streamline their customer service departments and call centers. But adopting these capabilities might not be as easy or inexpensive as they would like. Understanding the capabilities, limitations, challenges, and adoption steps required can help organizations choose the right solutions.
Umesh Sachdev, cofounder and CEO of conversational artificial intelligence provider Uniphore, says innovation in this space is evident in markets where call volumes continue to rise, customer expectations are high, and companies focus intently on solving the systemic problems of customer service.
“Speech tech started out as a niche customer service opportunity but has been steadily gaining momentum as consumers experience the power and efficiency of voice as an interface firsthand,” he says.
Danielle Moffat, managing director of intelligent sales and customer operations at Accenture Operations, notes that there has been tremendous progress in recent years in automating customer service interactions through various speech technologies that leverage AI and natural language processing (NLP) capabilities. Automatic speech recognition (ASR) is one such technology.
“Within customer service, we’ve seen the adoption of automated agents, such as speech technology agents, increase by as much as 30 percent over the past five years,” Moffat says. “There are more self-service options and customer service channels available today than even two or three years ago, and speech technology is often a component of a company’s overall channel strategy.”
Tracy Malingo, senior vice president of product strategy at customer engagement systems provider Verint Systems, can testify to this trend.
“At Verint, we’re seeing a lot of interest from customers who want to integrate voice-based customer service solutions to help them build a more robust customer service stack. A lot of the advances we’re seeing in the industry today center on innovations in NLP that allow customers to access a larger library of language with their customer service solutions. This allows them to provide more on-brand and on-goal customer support, even when it’s through an intelligent virtual assistant or other automated system,” Malingo explains.
Alexey Aylarov, cofounder and CEO of Voximplant, a provider of cloud platforms for developing communications applications, says the two primary domains in which speech technology continues to be integrated to improve customer interactions remain supporting communication during the call and analyzing communications after the call.
“During the call, basic IVR technology helps navigate the caller to the correct human agent and collect useful information before connecting the customer with the agent. After a call ends, speech analytics can help managers monitor how well their agents are adhering to call scripts and following expected protocols,” he says. “But today, speech analytics can also look for ongoing trends, topics, and terms that come up in conversations. Even more advanced speech analytics capabilities can also analyze customer sentiment and emotions expressed during a call. For example, if speech analytics identify an unpleasant shift in tone from the customer, it can provide feedback to the agent on how to save the call and keep the conversation flowing.”