3 Ways to Tie Artificial Emotional Intelligence into the Customer Journey
In a time when so many of the interactions we used to have face-to-face are moving online and workforces are distributed, the type of organic emotional capture we used to naturally do—responding quickly to an upset in-store patron, for example, simply can’t be done. Gauging emotion from afar, in digital environments, is becoming even more important.
The relationship between technology and human emotion is one that’s increasingly characterized by blurred lines. Robots, more so now than ever before, are capable of mimicking, replicating, and, most importantly, understanding human behavior.
The emotional connection between humans and technology is moving from the figurative to the literal. On the one hand, you have technology that can more and more realistically mimic human emotion. At the Consumer Electronics Show this year, for example, Samsung subsidiary STAR Labs debuted a project called Neon, featuring realistic human avatars that can interact in real time and rely on individualized AI models to replicate human behavior.
On the other end of the spectrum, there’s the still-blossoming field of artificial emotional intelligence, or emotional AI. Instead of replicating human behavior and emotions, like Neon, emotional AI seeks to perceive and interpret human emotions. It has been around for a long time, primarily in the form of sentiment analysis—companies of all sizes have been investing in ways to understand human emotions through video or audio for more than a decade. We’re beginning to reach something approaching critical mass, however, with Gartner estimating that emotional AI will influence more than half of all the ads you’ll see online by 2024.
There are a number of companies already working to give computers the ability to read and translate our emotions. Boston-based Affectiva, for example, uses webcams to track facial expressions and gauge sentiment, even measuring heart rate by tracking color changes in a user’s face. Those reactions can be synthesized and tied directly to metrics like brand recall, purchase intent, and likelihood to share.
What’s more, the devices we as marketers use to reach our customers are becoming more friendly to emotional AI. Gartner predicts that by 2022, 80 percent of smartphones are going to come out of the factory with on-device AI capabilities, with emotional recognition as a key selling point.
These developments are certainly intriguing and they pose a number of questions for marketers, including “How might I leverage emotional AI to better connect with customers?” and “How would I even begin to take advantage of a capability like this?”
Here are a few suggestions to help you formulate answers to these questions:
Use It Where It Makes Sense
There are some parts of your customer journeys where emotion doesn’t play a factor—asking a customer what shipping address they might want to use, for example. There are others that tend to have more emotional charge to them, like putting in credit card info, which could have some anxiety attached to it, or trying to sift through a complicated website, which can be frustrating.
Looking at customer journeys, marketers should be trying to figure out what emotional factors are playing a role. As the field of emotional AI advances, we envision being able to draw up kinds of emotional journey maps. Until then, there are only a few touchpoints currently where users organically surface their emotions in observable ways and emotional AI can be leveraged in a high-impact fashion.
An example here could be customer service. Using sentiment analysis capabilities to analyze callers’ emotions through their voices before they are patched through to live agents can help guide agents’ approach. Same goes for video analysis—if a customer is engaging with a video virtual assistant, or even a live agent in a video chat, an emotional AI tool can capture imperceptible changes in mood that might not be apparent to the human eye and agents can use that information to adjust the conversation on the fly.
Don’t Be Creepy
Good customer experience is built on a foundation of trust, and it can be a fine line between analyzing emotion to provide meaningful customer experience and coming off as creepy to your customers.
Having a window into your customers’ emotions opens up the potential for manipulating them—leaning on fear-based tactics for someone identified as susceptible to it, for example. The goal should always be to enhance customer experience; to use emotional insights to serve up experiences that guide customers to the best resolution easily and without friction.
Use the Data Effectively
Artificial intelligence both consumes and creates an enormous amount of data. You already have a treasure trove of customer data throughout your business, but you need to have the right frameworks in place to feed data to emotional AI programs and contextualize the data it puts out.
Quantifying and acting on emotional data isn’t something many brands are ready to do. Using emotional AI to identify user emotion is one thing—knowing how to use that information to enhance CX is another. What specific actions do you take when a user is upset or frustrated? What about when they’re happy? Brands that have undertaken customer journey analysis and journey discovery have a head start.
Having a data-driven customer journey map puts emotional insights in context. An understanding of optimal journey paths provides a road map for the best resolutions no matter the customer sentiment, and real-time decisioning allows those next best actions to be carried out automatically. Finally, orchestration data can be used to front-load AI capabilities for faster, more accurate insights.
Blurring the lines between emotion and technology doesn’t always have to result in an uncanny valley. When it’s implemented in the right spots, done with the right intentions, and paired with the right framework, artificial emotional intelligence can open the door to empathy on a wider scale, making the once-intangible measurable and actionable.
Mark Smith is president of Kitewheel, a Boston-based provider of a cloud-based customer engagement hub for orchestrating real-time consumer journeys across any channel. Prior to joining Kitewheel he held leadership positions at Pitney Bowes Software, Portrait Software, and was the founder and president of the predictive analytics software firm, Quadstone. He has a Ph.D. in mathematics and statistics from the University of Edinburgh. Contact him at firstname.lastname@example.org.