Conversational AI in Retail: The Missing Ingredient

The rise of COVID-19 changed retail forever, shifting existing trends toward omnichannel strategies into overdrive. Online-first strategies and contactless interaction became essential virtually overnight—and two years into the pandemic, they’ve stayed that way.

The good news is that technology developed alongside these challenges. Artificial intelligence (AI) has led to new ways to connect with customers, deepening brand relationships outside the walls of a traditional brick-and-mortar store. Conversational AI (see sidebar below) is one of the most promising ways to keep these customer relationships going strong from a distance.

Today, conversational AI shows up in every aspect of the retail industry, from marketing to customer service to shopping itself. It’s part of the entire relationship with consumers, influencing the quality of the long-term brand relationship—which in turn influences whether customers come back for another transaction. But conversational AI also introduces a new challenge: At every stage of this journey, consumers still prefer human interactions—or at least interactions that feel human.

Branding specialists have known this for a long time, with concepts like “brand voice” and “brand personality” explicitly attributing human characteristics to the non-human entity of the brand itself. Just look at the Common Language Marketing Dictionary’s definition of brand personality.

See the explicit comparison between the non-human brand and the human characteristics? Conversational AI systems must do something similar. After all, brands succeed when people form personal attachments with them; every contact point must maintain that personality. In a pandemic-altered retail environment, the question is how to bring more humanity to conversational AI systems.

Machine speech technologies are stepping in to bridge the gap by making AI more human. In fact, the concept of emotional intelligence (EI) is key to successful automated customer experience (ACX) deployment in voice channels: not just conversational AI, but conversational EI.

Introducing Conversational EI: Emotionally Intelligent Voicebots

People naturally adapt their tone of voice to fit the interaction. That’s an example of emotional intelligence, which Harvard Health Publishing describes as “the ability to identify and regulate our emotions, to recognize the emotions of other people and feel empathy toward them, and to use these abilities to communicate effectively…”

So how do you add an emotional element to a voicebot? You need two elements. The first component is speech emotion recognition (SER), software that classifies the content of the human user’s speech. That could be a system that sorts words and phrases into emotional categories. It could even be an AI model that learns to classify emotion based on acoustic features embedded in the user’s speech. Either way, SER is an established science, with patents dating back to the 1970s.

But emotion recognition is just the first piece of the puzzle. The second is the emotional expression to match.

That’s where ReadSpeaker’s emotionally augmented TTS voices can help. The term sounds highly technical—and the science of creating emotional speaking styles certainly is—but the benefits to retailers aren’t hard to grasp.

Wondering what emotionally intelligent TTS voices actually sound like? Experience them today with our Voice Application Demo.

Emotional TTS In Customer Service Applications: An Illustration

Say you have a bad online shopping experience and take your complaint to an automated contact center or online voicebot. If the standard TTS voice sticks to an upbeat tone, your experience goes from bad to worse. But with today’s TTS technology, you can hear the same familiar brand voice speaking in apologetic tones.

It’s the same thing people do in every conversation. Human customer service agents unthinkingly moderate tone to match the caller’s mood; emotional EI brings the same ability to ACX systems.

That creates better customer experiences by using an appropriate tone of voice for every conversation. It makes every shopping interaction unique. Ultimately, conversational EI builds stronger relationships between brands and consumers.

Let’s review a few more use cases for emotionally intelligent conversational AI.

Conversational EI Across the Shopper’s Journey

Every brand interaction is an opportunity to build a stronger relationship with consumers, and that includes automated customer service solutions. Conversational EI opens the whole universe of ACX applications to retailers, while providing more natural and successful interactions wherever brands and consumers meet.

Here are just a few places conversational EI can create better customer experiences:

By matching tone of voice to the unique needs of the individual conversation, emotional TTS creates a better brand experience across retail interactions. But to truly drive brand recognition, and the loyalty that comes with it, conversational EI must also be consistent and unique.

Combining Conversational EI with Custom Branded TTS

Just as brands create strong, unique logos in the visual domain, they need a clear marker of identity for voice channels: a literal brand voice.

A custom TTS voice from ReadSpeaker fills this need, building brand legacy while reinforcing customer loyalty. As conversational AI continues to show up in more and more applications, the sound of a brand’s voice—its quality and character—becomes a central asset. It would be unthinkable to share a logo with Amazon. Soon, using Alexa’s voice for your brand’s smart speaker app will be the same.

Most TTS voices are already used by numerous brands, or branded for the platform itself (here we return to Amazon Alexa). That dilutes brand recognition with every utterance.

Human voice actors may also work for multiple brands—and voice talent doesn’t allow users to adapt on the fly. To produce a new sentence, you have to book a new recording session. With TTS, you can produce new utterances in an instant—essential for cutting-edge conversational AI that uses natural language generation (NLG) to determine new replies to every question.

A custom TTS voice allows brands to stand out wherever they interact with consumers—and maintain consistency across all those channels. Combined with conversational EI, custom TTS voices create great human-bot experiences that unmistakably belong to your brand, every time. That gives brands the edge they need to compete in a contactless, omnichannel retail environment. Conversational AI allows brands to connect with customers on a limitless scale—but only conversational EI scales the human experience at the same time.