isolated yellow rotary phone against a matching yellow background
PHOTO: Mike Meyers

As companies look to provide customers with better customer experience (CX), an increasing number are turning to voice solutions to provide quicker and better service in their contact centers.

The popularity of voice response has grown thanks in large part to the commercialization of Apple’s Siri, Amazon’s Alexa and similar technologies as well as customers' desire to contact companies while driving or at other times when using touchtone interfaces aren’t possible, said speech technology consultant and DMG Consulting CEO, Donna Fluss.

Fluss added that voice is the most ubiquitous and easiest form of interaction for most people, and the cost of the underlying technology for voice recognition has dropped.

Voice in Customer Service Can Reduce Friction

“When a customer is calling into the enterprise it’s because they tried web and mobile and still have questions, couldn’t complete their transaction, or because they want confirmation that they completed their transaction,” said Ken Arakelian, director of the technology advancement group at Nuance Communications. “Rarely is a customer calling for a friendly conversation. That said, the key to good CX on the phone channel is to enable the customer to get where they need to go to resolve their query quickly and without friction.”

For simple queries, customers can use touchtone entry (“press 1 to hear our hours of operation”), but if the business is more complex, voice and natural language understanding can remove the friction from phone customer service, Arakelian explained.

Natural language understanding allows customers to ask a question in their own terms (e.g., “I have an insurance claim,” “how do I report an insurance claim?" or “insurance claim question”) which prompts a similar follow-up question (e.g., "auto or homeowners?") or results in being transferred directly to an agent, depending on the structure of the company’s technology. Before natural language understanding, callers would have to use very specific wording or the system wouldn’t understand the query.

Related Article: A Good Chatbot Is Hard to Find

Speed Self-Service Processes, But Beware the Speed Bumps

“When it comes to self-service, research shows that customers want highly personalized, low-effort experiences,” said Chris Bauserman, vice president, segment and product marketing for NICE inContact. “That said, building an empathetic, human connection with the customer is equally as important; a successful self-service strategy should replicate what a great live agent does via traditional phone.”

Customers want to feel like you’re listening to them, regardless of industry, and with the strides made in conversational AI, there’s no longer a need for complicated phone menus or bots that feel unnatural, Bauserman explained. Today’s conversational AI combines artificial and human intelligence to provide smart, personalized experiences.

So the experience of replacing a lost ATM card goes from a long, complicated process of providing passwords, account numbers and other time-consuming information to a quick identification, then handoff from the interactive voice response (IVR) to a speedy hand-off from IVR to agent, who can deactivate the old card and ship a new one in moments.

“In this scenario, the customer experience has been augmented not only by increased operational efficiency and personalization, but also by heightened security; the bank was able to authenticate the customer’s identity more securely by using the customer’s own voice, rather than an easily-guessed PIN or password,” Bauserman said.

However, some natural language understanding systems are very basic, with limited vocabularies, so won’t understand anything but specific, common terminology, so can frustrate the caller rather than enhancing CX, Fluss said. “There are a lot of old, tired IVRs out there that need to be updated. It’s time for many companies to upgrade to what works better. It’s an investment that needs to be made. Making it easier for customers does make a difference.”

Related Article: Why Your Approach to Chatbot and IVR Projects Is All Wrong

Voice Alone Is Not Enough

However, voice solutions by themselves aren’t enough, cautioned Jen Snell, Verint Intelligent Self-Service vice president of product marketing. “While voice-based user interfaces are currently in-style, it doesn’t mean they’re always the best modality for all users. In fact, the typical enterprise is dealing with tens of thousands of customers all with varying traits, cultures, primary languages and intellectual abilities. We would argue that voice improves CX when it is delivered as one modality in a multi-modal interface.”

For the best CX, companies need to offer different options for end users, who have different preferences and abilities, said Snell, pointing to company research showing that voice outperforms touchtone with low-literate users, but there’s a sharp divide between age groups and their use of voice assistants, where 55% of people aged 18 to 49 use them, while only 37% over 50 use them.

“We recommend analyzing existing system data from customer interactions and developing an evidence-based view of your customer base and the issues around which they engage you,” said Snell. “From there, it’s pretty easy to decide if maybe a hands-free nature of a voice interaction adds unique value, or maybe the unique traits among your customers make touchtone more satisfying and accurate.”