For insurance agents, quick access to policy data is essential to helping customers faster.
With more than 55 million customers around the world, Zurich Insurance Group relies on strong customer relationships to stay competitive.
“Interaction is by far the biggest driver of loyalty with our customers,” Aleksandar Vidović, the CEO of Zurich’s AI and analytics company, ZCAM, told Business Insider.
Zurich has a network of thousands of agents worldwide who work directly with customers. Their main job is selling insurance policies, but the role also involves explaining complicated policy details.
They communicate across various channels, including phone calls, emails, and WhatsApp messages. The agents told Vidović’s team that juggling these channels, as well as ploughing through dense policy documents, could make it difficult to keep track of the details needed to give each customer a tailored experience.
For instance, Vidović said that the agents spent a lot of time trying to determine whether customers were covered for a specific incident.
“You can imagine going through 10 or 15 pages of insurance jargon. It can take a lot of time,” he said.
Previously, data on customers and company policies was scattered across systems or, in some cases, not provided to the agents at all. In January 2024, Vidović’s team set about creating a customer relationship management system that could connect the dispersed datasets to help the agents give every customer what they needed.
Simplifying customer support
One of the team’s guiding design principles was the “three-click rule”: No agent should need more than three clicks to get what they need. Whether it was pulling up a customer’s policy details or sending a follow-up message, everything had to be fast and intuitive.
Zurich, which serves customers in more than 215 countries and territories, is familiar with rolling out new technology; it has implemented more than 160 AI-powered tools across its operations. But when it came to developing a new approach to the CRM, there wasn’t a one-size-fits-all solution because different agents use different platforms. To tackle this, it built the CRM’s interface into existing tools like Microsoft Outlook and Salesforce.
The first part of the solution involved putting customer and policy data into one easy-to-navigate platform. For instance, when speaking with a customer, the intelligence platform allows an agent to easily find their current policies and records of previous interactions. This data then connects to AI features within other platforms, such as tone analysis and messaging feedback.
The CRM also helps each agent suggest the customer’s next steps, with the company taking inspiration from tech giants like Spotify.
“When we listen to Spotify, we get a playlist and then it recommends the next best song,” Vidović said. “We use the same modeling to recognize the most appropriate product for a customer based on their needs.”
The new CRM is now used in four of the company’s markets in Zurich.
“We see a significant increase in customer interaction because it takes less time,” Vidović said. “The agents don’t need to spend so much time drafting up specific messages.”
Based on Zurich’s initial calculations, the team could cut servicing times by over 70%, though Vidović said that they hope to better understand the tool’s impact by the end of 2025.
Transforming the insurance industry
Stephen Poux, the executive vice president of risk management at the Liberty Company Insurance Brokers, told BI that insurers, brokers, and clients were already seeing artificial intelligence’s potential.
“AI is transforming insurance agents from transactional intermediaries into trusted advisors and risk consultants,” he wrote.
“With AI tools automating many administrative and underwriting tasks, agents now have more capacity to focus on building deeper client relationships, offering strategic guidance, and delivering highly personalized risk solutions,” he added.
Vidović said that the most challenging aspect of the rollout had been managing agent expectations: Once they understand how the customer platform and AI work together, many of them reach out to Vidović’s team with new ideas.
“The challenge now is outlining the relevant features and continuing to upgrade the tool,” he said. “Because it’s a never-ending process.”
