How can AI help insurers build empathy across organizations?

Define what empathy means in your organization.

Ask 10 insurers what empathy means to them, and you’re likely to get 10 different answers. However defined, empathy for insurers is hard to convey convincingly while they are denying a claim. Yet every insurance company can readily identify those “top agents” able to deliver any message to a policyholder in an empathetic way, agents who possess a special ability to communicate with customers on a human level.

The first step for insurers in building an empathetic organization, then, is to analyze what these successful agents – the top 5%, for example – are doing and compile a list of empathy best practices that can be used to train all agents. However, these top-performing agents on the empathy scale are far more intrinsically skilled and not easily emulated. It is critical, therefore, that insurers bring in experienced coaches and consultants to build effective training and response programs.

Compounding the organizational empathy challenge for insurers is the trend toward providing self-service both to increase efficiency and meet rising customer expectations for convenience. Granted, policyholders don’t have a need for empathy in every situation, but when they do, they are not going to get it from an AI-based chatbot.

Use AI to build an empathy model.

The limitations of chatbot technology notwithstanding, AI can play an integral role in raising an organization’s empathy level by acting as a real-time coach for live agents, offering proposed responses and specific language to convey empathy to policyholders. In this context, AI acts as a skillful hidden hand guiding the human touch.

With input from management, data scientists, and experience from their top empathy performers, insurers should utilize AI to program or create an empathy model for other agents to follow. This is the second step insurers should take to build an organization that is consistently empathetic to policyholders.

Imbue empathy across the enterprise.

The third step is to implement the AI empathy model so that it is part of every conversation with customers. An AI model can generate suggestions in real-time along with information and feedback that managers can use to coach their teams and institute a practice in a way that better conveys empathy during customer conversations.

As an example, insurers can use AI and natural language processing (NLP) to conduct sentiment analysis in real time to gauge the emotions of a policyholder speaking to an agent. The AI model uses this information to suggest responses and specific language to human agents while they’re interacting with the customer. The sentiment analysis score then can be viewed by supervisors to gauge areas of improvement for the agent.

There is no doubt that some people have a gift for interacting with others and for expressing empathy. And though you can’t teach gifts, you can teach skills. With the right tools and training approach, human agents can upgrade their empathy skills enough to improve their effectiveness and overall performance. Not everyone is going to be in the top 5%, but if insurers can upgrade someone from the top 40% to the top 20%, and replicate that process through the AI model, that’s a big win.

Conclusion.

In the wake of the COVID-19 pandemic and as natural disasters continue to increase, empathy has become more important than ever for policyholders and, by extension, insurers. When a policyholder is in crisis, after all, he or she wants to talk with a fellow human.

Thus, while the benefits of AI, automation and self-service to insurers and policyholders are many, it would be a mistake for an insurance company to rely strictly on these technologies for customer support. The most effective approach is a hybrid human/AI model that can be customized to fit the specific needs of the organization.

This post was first published at Advisor Magazine.