Companies use natural language processing to create customized voice assistant experience

Companies use natural language processing to create customized voice assistant experience
Growing number of businesses are adopting natural language processing (NLP), the technology that underlies virtual assistants, to create a more engaging interface with their customers and employees, according to a study by IT management consulting company CapTech.

Businesses are increasingly considering custom NLP solutions for the workplace, seeking ways to reduce time needed for analysis and decision making, and providing employees the ease and enjoyment of interacting with virtual assistants, CapTech says.

“It’s about the intersection of NLP with data science, machine learning, and customer engagement heuristics working together to meet the user on their terms, in their environment, on their device, with the information they need at the exact time they need it,” said Vinnie Schoenfelder, CapTech’s Chief Technology Officer.

NLP also enables virtual assistants to move away from simply responding to commands or questions to carrying on a conversation with an understanding of intent, emotion and context.

Voice technology solutions may be applied across industries, including retail, hospitality, energy and utilities and financial services, CapTech said.

“The potential for NLP to improve customer experience and business processes is vast, but organizations must go beyond basic, scripted interactions to create seamless and enjoyable experiences,” said Schoenfelder. The barrier is usefulness, not security or cost. Customers and employees will not adopt a new channel if there are usability frustrations or if it does not demonstrate immediate value.

CapTech recently conducted a separate study that examines the profile of smart speaker users as well as adoption habits. Understanding users allows companies to determine whether NLP fits within their brand engagement strategy.

The study found that just over half of smart speaker users are millennials or younger, 32% are Gen X, and only 12% are Baby Boomers. In addition to age, the main variables that drive smart speaker usage are income and gender. More than half earn over US$75,000 a year and 60% are men. The most popular uses of smart speakers are playing music (82%), inquiries/information gathering (42%) and shopping (39%).

Suggestions for voice assistant deployment

CapTech suggests three key considerations for businesses that are thinking about designing NLP applications that will deliver a satisfying experience for customers or employees.

The first thing to consider when building an NLP application is the user experience, or the interaction model. The screen is getting smaller, from website to phone, and phone to watch, and finally to voice, where there is no screen. This has created a perception that the need for customer experience/user experience design is also shrinking. “On the contrary, as you lose visual cues and begin to use newer technologies that people have little experience with, you need more thoughtful interaction design,” Schoenfelder pointed out.

The second consideration is service. Most businesses have already developed many services for their web and mobile applications. Some assume that this means they already have what they need to leverage the voice channel. That may be true, but a good deal of service mapping is required at the middle tier. Businesses must be able to tie together the intent of the customer’s voice with the appropriate services at the back end, so that each request is satisfied appropriately.

The third is artificial intelligence and intent detection. “You need to understand what your customers want and how they feel in order to establish a trusted conversation. That’s key to keeping customers from becoming frustrated and abandoning the opportunity to interact with your brand,” Schoenfelder said. Companies will need to establish a way to determine whether voice-enabled devices truly understand and deliver what consumers are asking for.


Share to:
Comments ( 0 )