Using AI-Based Tools to Build Empathy in the Contact Center


Think about the best experience you’ve had as a customer. Chances are it was superior to any other interaction you had because you felt understood by another human being; it was a time when someone addressed your needs and worries — and treated you with great empathy.

To build empathy as a company, the contact center is the place to start. The contact center often is the turning point for any customer experience with a brand; the kind of consideration and care customers receive on the other end of the line can forever change their perception about the company — and it can negatively affect their loyalty forever.

Effectively building empathy into the customer experience requires you to use tools that are also built with empathy. This means choosing tools that put people at the center. Human-centered design, which is part of a design thinking ethos, follows a market shift toward empathy. It reaches an understanding of what companies need to improve to better serve their customers. Exercises like empathy mapping and customer interviews ensure developers put themselves in the users’ shoes and routinely check to be sure their efforts are on track.

Designing for Empathy

When it comes to actual customer-agent interactions, what does empathy look like? How can you measure empathy? And more importantly, how can you find opportunities to teach empathy?

This is where human-centered design comes into play. At Genesys, we’ve used strategies, such as empathy mapping and customer interviews to understand what customers need so we can then develop workforce engagement management tools that build empathy into the experience. Contact center supervisors should ask:

  • What does empathy look like when our agents express it?
  • Why is this important to us and how can we measure it?
  • How can we find opportunities to teach empathy?

There are two sides at the heart of every interaction — both trying to work together to resolve an issue. While there are many ways to do this, how both sides feel will vary depending on the approach each takes. In some easy-to-resolve interactions, customers who interact with a lifeless but effective bot might walk away satisfied because they felt they were treated well. In other situations, especially ones in which the customer is fraught and panicking, that same expressionless but efficient interaction might leave the customer feeling angry.

Joel Hickok, Senior Manager, Call Center Operations at AAA, summed up an empathy-based interaction like this: “[Our] customers are in a tough spot. Repeating it back, reassuring them, acknowledgment that they’re in a tough situation — using phrases that show empathy and using the correct tone.​”

Using empathy empowers companies to go beyond the standard measurements of handle time and disposition codes — they can move into interactions where sentiment, tone and how interactions are handled matter more. Ultimately, these factors affect standard measures like first-contact resolution (FCR) and Net Promoter Scores (NPS). However, looking closer at them shows what’s really happening in the interaction — and it can reveal opportunities for managers to mentor their employees. It can also reveal “teachable moments,” which could include:

  • How an employee responds when confronted by an angry customer
  • How an employee handles objections or the use of particular phrases
  • What the employee does to end the call on a good note

Presenting these back to employees, either with commentary or for self-review, can improve their ability to connect during interactions. While it’s possible to review interactions manually to find these moments, it doesn’t scale. Even when listening to interactions at a higher speed, you might only experience a small sampling of interactions from any single agent. Additionally, this can be exhausting for the reviewer.

Limiting a review to interactions or segments of an interaction is a better way to reveal teachable moments. To do this, you can search for interactions based on sentiment scores, especially the sentiment trend, and then review the call around when those sentiment events were detected. Another method is to use a topic-spotting tool to find specific phrases the agent said to a customer and then read how the agent responded. Finally, examine multiple factors, such as NPS and sentiment events, and then review segments around those sentiments — especially a positive NPS score tied to a call with negative sentiment scores.

By using artificial intelligence (AI)-based tools, you can even automate the discovery of those teachable moments. So, instead of going through lengthy review cycles to detect gaps in skills or knowledge, teams can quickly execute strategic analysis and identify variations in performance at the queue, flow and agent level. This allows supervisors to focus more acutely on building agents’ skills to further improve customer experience. To deliver truly empathetic customer experiences, contact centers need to embrace empathy with their employees – adequately supporting them throughout their journeys, understanding their needs, and embracing a culture of open feedback and continuous improvement.