Behavioral Signals Emotional Intelligence is the Key to Improving AI-Driven Customer Service

Emotional Intelligence is the Key to Improving AI-Driven Customer Service

We are in the midst of a customer service revolution. New technologies are being deployed rapidly across all industries, providing support in the form of AI-powered chatbots, agent-assistants, and sales intelligence. These tools are designed not just to interface with customers but to also provide key insights about what customers need, when they need them. And while we are already seeing huge leaps forward in the sophistication of these clients – chatbots that can handle conversations online and agent-assistants that can cue customer service reps with valuable information during a heated discussion – a gap remains. Humans, in general, are highly emotional. When someone has a question, they are in a certain emotional state – whether it is anger, fear, confusion, or something else entirely. AI in customer service will never truly be able to fully augment human agents until it can recognize these emotions and respond in kind. That’s where emotional intelligence comes in. New technologies are being developed that leverage what we know about human emotion and the cues they give based on key responses.

How Emotional Intelligence Informs Customer Service

Emotional intelligence refers to the five core components popularized by Daniel Goleman – self-awareness, self-regulation, motivation, empathy, and social skills. How someone scores on a spectrum for each of these indicates their EQ. Humans remain far more sophisticated in these areas, able to detect and respond to subtle cues in the voice of the customers they are talking to, but machines are getting smarter, and in a hybrid approach, they can be a powerful resource.

One way this is being done is by implementing “whisper agents” – a supplemental tool that customer service agents can rely on to provide key insights during a call with a customer. One of the greatest challenges in customer service is dealing with someone in a heightened emotional state. That negative energy can make it difficult to recognize when the agent is starting to react in inappropriate or overly aggressive ways. AI can be used to analyze how the agent responds and provide feedback if it detects vocal cues that it associates with anger, frustration, or other negative emotions. The same can be analyzed on the other end of the spectrum. If the customer is becoming frustrated or upset with the agent or the situation. Currently, AI alone is capable of handling these types of conversations, responding to customer concerns in real time based on their emotional responses.

Another situation in which an AI agent must work in tandem with its human counterparts is in recognizing when a human agent is needed. The benefit of AI being able to evaluate vocal cues and recognize when a customer is becoming angrier, frustrated, or upset, is that it can transition the call to a human agent who is better able to convey empathy on the call. This relegates human agents to a smaller, but more sensitive subset of calls.

Another major benefit of emotional intelligence in AI is that it can scale to the needs of individual calls in ways that human counterparts cannot. While humans are increasingly expected to provide a more personable experience on customer service calls but avoid those calls going overly long, AI can scale indefinitely to match the needs of these calls. If it means an hour to process a cancellation because the cues picked up on from the customer indicate that this would be beneficial, then the system is able to accommodate those needs.

The Future of Emotional Intelligence in AI

Many organizations are invested in increasingly robust AI to support their customer service teams. Allstate, for example, uses an AI agent called Amelia that helps service agents on their calls. IBM’s interactive voice response system is powered by Watson to do the same in direct contact with customers. This Customer Care Voice Agent integrates with call centers and provides an intermediary between automated response and human agent. MetLife is using AI in creative ways to support its claims experience. The AI proactively analyzes calls in real-time and provides messages to the customer service associates on how to improve the call, as it’s happening.

The arms race in AI is heating up as well. Mark Cuban muses that the first trillionaire will likely be someone in Artificial Intelligence and this segment is the future of an AI that can truly integrate with business activities. For this reason, we have many startups investing in emotional intelligence. Affectiva is increasingly well known for its use of facial recognition in AI while Behavioral Signals is building algorithms to identify emotions in vocal cues for use in a range of different applications, like their meilo app for contact centers.

The current trajectory for AI integration into customer service is supplemental. Humans are increasingly used to dealing with machines in the form of chatbots, automated phone systems, and personal voice assistants. As technology evolves and customers become better acclimated to interacting directly with machines, we’ll see a growing number of companies using AI as the front line in customer service, with human agents as a backup for particularly tough cases.