AI in action: How AI-backed automation reduced the routine activities and tasks from my team’s workloads

When my son introduced me to generative artificial intelligence tools, I dove right in and started using AI tools like a personal assistant.

I was amazed by what the tools could do, especially when it came to creating communications. As a big Star Trek fan, I couldn’t help but think of the scene in the original Star Trek where Spock would talk to the computer. At the time, that seemed so futuristic. But that’s essentially what I’m doing now when I use AI tools: “Hey computer, help me.”

I’m not alone in my interest in AI. In the 2024 Agent-Customer Connection Study from Liberty Mutual Insurance, 64% of agency principals said they are interested in how AI can improve their business.

However, not many principals have taken the leap to actually implement AI in their agencies. The study found that only 6% of agency principals said they have already implemented an AI solution in their agency, but 36% say they are likely to be using AI in the next five years.

AI might seem like a gimmick or a toy right now, but it’s quickly changing the way people work. It’s important for agency leaders to take the time to explore AI and set parameters for their team about what AI is for and how they can use it.

For my small team at Top O’ Michigan Insurance, AI has been a great tool to boost our marketing automations and help us communicate quickly and efficiently throughout the client lifecycle.

Our approach to implementing AI

As independent agents, our brand is about the experiences that clients have with us. Clients want their agent to regularly communicate with them, educate them on insurance, and provide good service. These are key to driving retention and building trust.

Our team at Top O’ Michigan is small, but we value our clients and want to make sure we stay in regular contact with them. We wanted to find a way to communicate with clients in ways that are timely, consistent and personalized without adding more to our staff’s workload.

The solution we found was AI-powered marketing automations using our agency management system, Applied Epic, and generative AI tools.

We took a four-phased approach to implementing our new process for AI-backed automations:

1. Build the base

First, we made sure the whole agency understood the standards of data entry. You can’t really use AI to its full potential if your data is poor. Data integrity and consistency is key.

Fortunately, we had a solid base to start with, since data consistency has been one of our agency’s objectives for at least 20 years. Customer information is consistent across the agency. This has helped our team service accounts, and now it’s helping us build out automations and craft consistent messaging.

2. Build workflows

At Top O’ Michigan, we focus on five key “moments of truth” that clients have with the agency: quoting, onboarding, servicing, claims, and renewal and retention. To build out automations, we looked at what we want to communicate in each of those moments and what we want the experience to be for clients.

Then we started using our AMS to support the moments of truth. Our team identified the triggering events within the AMS when we wanted to communicate with clients, and we used generative AI to create the messaging around those events and build it into automated workflows. For example, if we get an e-doc from a carrier on a policy, our system sees that coming in, recognizes that it’s a renewal from the message class code, and pulls out AI-generated communication to tell the customer that they have a new document available in our client portal.

To get staff buy-in, we involved our team in the process. We trained staff on how to prompt the AI tools and vet the outputs. The team dove right into it, with team members taking ownership of certain segments, identifying tasks, and contributing to discussions about what we were trying to accomplish every step of the way.

3. Test and refine

Once we had a workflow built out, we ran and tested the automations, tweaking things as needed.

4. Maintain and keep building

We see this as an ongoing process. The market changes, so our communication will need to change. Or we may need to change certain criteria that triggers automation.

We also don’t want our messaging to sound stale or too robotic, so we use generative AI tools (like ChatGPT or Claude) to regularly update the communications to keep things fresh.

The impact of AI in my agency

Once we set up automated workflows, our AMS started taking over daily routine tasks that would usually fall to staff members. This reduced the amount of activity on our staff’s home base significantly – reducing roughly two-thirds of routine activities and tasks from staff workloads.

We’ve started using AI-backed tools to drive efficiencies in other areas of our agency, as well. For example, we used generative AI to create a new logo and a jingle. We also use an AI-backed tool to transcribe and summarize our meetings and identify next steps.

AI isn’t a replacement for people. It’s a tool to help them do routine tasks quickly and efficiently and save their time for problem-solving.

AI technology won’t solve your agency’s problems, and it won’t create good processes if you don’t already have those in place. But it can streamline your processes, save you time, and help you consistently communicate with your customers.