Why AI is Making Your Insurance Agency Fail Faster

AI has become a snake oil. It is so easy for someone to create something that looks like it'll do something, but the second you throw an expert to review what it's done, it's like, well, wait a minute, no, this is, this is not right. This is not how this is taught. This is factually incorrect. Too often people create something with AI that's a good test, it's a good theory, it's a good concept, but they don't know how to make it ready for the real world.

Hey, hey, welcome back. Welcome back to the Broker's Voice. Andy Neary here, CEO and founder of Complete Game Consulting. We're going to have a very interesting conversation today. I am joined by Corey Fraser. Corey, welcome aboard.

Thanks, sir.

For those who don't know Corey, Corey, how would you describe yourself as it relates to Complete Game Consulting? Utility man?

Well, because of you, I have accepted the moniker of AI mad genius. It is now all over all my social media accounts. It is on my Zoom account. It's just how I walk into a room now.

Yeah, no, I love it. Well, for those who don't know, Corey, you assist our company in many ways. You coach one of our clients, which you've been an awesome coach 'cause you've got industry experience, but you've also been diving deep into AI now for quite a while. So you have become my, as you just described yourself, AI mad genius, I might say mad scientist, on how what we're building behind the scenes with AI to really, may use AI to streamline the business. But where I would like to start, we're going to talk about AI, what it's doing in insurance, how agencies could be using it. But what I would love, where I would love to kick this off, you shared a story offline. You're doing some work with the Better Business Bureau.

And what, so the Better Business Bureau introduced me to, okay, to a university.

Talk about that, that story and kind of what it led to as far as how to use AI and maybe not how to use AI.

So I got introduced to the Better Business Bureau because of the work that I'm doing with AI. And of course, they said, we have got to introduce you to this group of people. And so, without naming names, they introduced me to a university and then their head of this particular program that they're running through. Said, oh my gosh, what are you doing? We need to talk. And so we started talking about AI in education, AI in training, and they started telling me about what they were doing, that they had had this company approach them and they were all excited. They were all gung-ho that this AI was just going to do everything for them. And they was— it was just, it was just great. It was all these wonderful things wrapped up in a pretty bow. Well, my first natural instinct was to, in the back of my head, think, uh-oh, we haven't even 100% told everybody about what we're doing, and yet here somebody is already being a competitor and they're already talking to universities with it. So my heart sank. Then he started talking, he's like, yeah, so we got it, we tried it, the professors tried it, They started doing all this stuff and they came back and said, this is horrible.

I was like, oh, thank you.

I'm fine. Come to find out what they were being sold in reality doesn't line up. I have had lots of conversations with businesses and business owners and a variety of different people on that very topic. Of AI has become a snake oil. It is so easy for someone to create something that looks like it'll do something, but the second you throw an expert to review what it's done, it's like, well, wait a minute, no, this is, this is not right. This is not how this is taught. This is factually incorrect. And so I started talking to him about how we were using AI, and they became very interested in what we were doing, to the point where now there's conversations at the national level about talking about what we're doing.

So at a high level, if you could kind of give the definition, kind of maybe the juxtaposition of what many are doing versus how we feel we're using it that people are saying, ah, that's what we need.

So AI makes it very easy to create just about anything. I know that sometimes I get frustrated and forget that I'm talking to a computer that doesn't know the difference between a 0 and a 1, really, other than it's either on or it's off. and I'll yell at it thinking, why did you give me this stupid answer? It's like, oh, wait a minute, you don't know any better. And so there are a lot of people creating a lot of stuff that really are not what I would call business or production ready. I've got almost 20 years in the business world now. And there's a difference between a— what I would call a test and something that's ready for prime time. And too often people create something with AI that's a good test. It's a good theory. It's a good concept. But they don't know how to make it ready for the real world. They don't know what business really needs or how business really even operates sometimes. So they create all these wonderful tools. They're everywhere. Just look at any social media feed and here's an AI tool that solves every problem that you need. Just pay me money and you can use it yourself. And then you go in and you try it and you come up and it's like, well, this is not anything like what they just demoed and I typed in the same thing. And so it creates this frustration and anxiety and fear around implementing and using AI. And what usually ends up happening is because they give it too much liberty and too much ability just to go and do whatever it wants, right? And I've— I was in that boat like a year ago going, hey, just do this. And it comes up with something, it's like, no, that's not what I said. But that's what it thought I said. And so started learning more and more about, well, wait a minute, no, we, we, we need to think about this in a very specific way. And somebody said this to me the other day, and I heard it and I thought, that's a great way to describe how to think of and interact with AI. And it's a 5-year-old with an IQ of 180.

That's so true. I was actually, as you were talking about that, I was thinking, what is the rule for becoming an expert? 10,000 hours, something like that, right? I think it's the 10,000 rule.

Something insanely high. Yeah.

So if you do that and you just divide it by what is a full-time gig for a whole year in hours, 2060, what is that? 2060 hours?

2040, I think, something like that.

Something like that. Well, if you divide that out, that's literally spending every hour of every workday on AI for 5 years before you can call yourself an expert. And how many people out there right now are calling themselves an expert within a year of understanding what AI does?

Or less.

Or less. And to your point, it's such a wild frontier right now. I think people are— it's kind of the same thing that went through coaching, like the online coaching business. All of a sudden everybody became an expert at something. They could build an online coaching business, charge whatever the hell they wanted to charge, and people pay for it, not knowing not getting any kind of result they expected. I feel like AI is heading down that path right now where it's a wild west. We're all still learning what it is, but all of a sudden now you've got all these AI experts that are charging a lot of money to create unknown results, yet people are paying for it. But at the same time, they're getting disappointed, very disappointed for what they're paying for. And I think a lot of it has to do, going back to what we're going to talk about here, why are you using it? What are you using it for? You know, I heard it said, and I love your thoughts, I mean, I read, I read the book The AI-Driven Thought Leader, and the number one takeaway I had from the book was AI— the goal is not to let AI be the thought leader. The goal is to let AI be your thought partner to help support your thought leadership.

No, yeah.

And as we look at the insurance industry, for example, My fear is you're going to have a lot of people use it to be the thought leader. And I just would love your opinion on that, knowing what you're, you know, and what you're building and working on. Like, what is the risk when you are letting AI be the thought leader?

Well, I think the, the best way to look at that is look at some of the lawsuits that are being happened when somebody let AI be their thought leader and they try to go to court with it. And it started making up stuff that it seemed so convincing that, oh no, this is a real case and this happened and it never did. And so the problem with, I think, with letting AI be a thought leader is it doesn't have any thoughts. All it is is an engine that's trying to determine what you want it to say. And if you turn around and tell it, hey, write me an article on this, and that's all you give it. Write me a LinkedIn post on just this, right? Just write it. Okay. What tone do you want it to be in? What, what direction do you want it to be in? Are you for it? Are you against it? There's so many things that we go through as people when we're writing something that we think AI can read our mind and it knows that when I say do this, this is what all of the details that I mean. And it can't. It doesn't know. And in most realities, AI doesn't know anything from the last message that you sent it. Now, there are tools and there are interfaces and everything that make it look like it remembers your entire conversation and your entire history. And that's just because it feeds it all to it every time you ask it a question. It feeds it, here's all the information you need to answer this question. But it sure looks like it remembers your name and who you are and everything like that. Well, it's because we feed it to it. All of the time. It has to read it somewhere. At the end of the day, it's still just a computer trying to process an algorithm to come up with an answer that it thinks that you want.

Yeah, I, I, you know, I look at what we're building, and we're going to talk about this in a second, um, with our online training and how we're using AI coaches, right? And one thing I think that makes the platform successful is we didn't go to AI and say, hey, We want you to be the thought leader and help us create a curriculum that's— we're going to turn into an online course, and then we want you to teach our clients this online course through a series of AI-generated coaches, right? We said, no, no, here's our thought leadership, here's our curriculum, we want you to enhance our thought leadership by building AI coaches who can distribute our thought leadership through modules that allow our clients to leverage all of our curriculum without us having to be there in front of them teaching.

Yes.

I feel that is a very different way of approaching AI, and I, and I share that because I'd love for you just to kind of share what we're building on that end and why we're building it. But I also believe it leads to a good conversation for agencies to consider how they could be using that same process maybe to enhance their thought leadership in the marketplace.

So if we go back and we look at everything that we've said about the beauties of how AI doesn't work and thought leadership should truly be and stay with the person having the thought, and AI can't have thoughts. And we recognize that early on, that the power in what you're teaching comes from your own experiences, not somebody else's. It comes from your years of already coaching people and knowing what they're going to say, what excuses they're going to have, and everything else that comes with what you've already done. Even if we were to feed AI, everything that you know, it still would not 100% think like you. We can make it sound like you pretty good, but it's still not going to think like you because it's still not able to— what I would say is just to be you. And so what we started to notice was there's nothing more powerful than reinforcing the content and reinforcing the training that you've already got, because that's where the power is. It's already been proven to work. Why recreate it? But what we, what we had to solve was how do we get a coach available to them to solve something anytime that they needed to help getting it solved? right? That's the, that's, I mean, that's every coach's dilemma. Coach has to trade time for money. Well, at some point, every coach hits a limit on either how much time that they've got available or how much they can charge. So there becomes a finite cap. And so knowing that AI has a memory problem, Knowing that the best way to have a conversation with AI is one topic at a time, so it doesn't start getting muddled and it doesn't start having long context windows where it gets lost in what you're talking about and confused. We said, well, why don't we just use that strength as a strength in how we train through an online training platform? And so now we have an agent for every individual task or lesson that they have to do. You have to know how to, you have to know how to create your, your, your sales pitch. That's one task. How to identify who your ideal client is. Actually, I think we've got that broken into 5 separate tasks just because all, all we want to do is reinforce one thing at a time. And AI is great for having conversations in that space.

So what you're saying— let me pause quick, Corey— in that, as we're building these coaches, what we're doing is taking every part of that one conversation and building an expert in a way for that one piece of that one conversation.

Yep.

Yeah, that—

and, and all it does is it talks about just that one thing. That's, that's all we tell it to know. You don't know anything about that. Don't ask it for a recipe on something because it's, it's not going to answer it. Even though, you know, these large, uh, large language models have all this information available to them, we don't care. We want it to fall within our guidelines and our guardrails of the things you can talk about and the way you can talk about them. And if there's something that we want them to know, we will tell them, oh hey, remember 2 lessons ago where you went through the exercise and you created this beautiful profile of This ideal client, if you were sitting in your room, this is, this is what they would look like. And it's just, it's just perfect vision. Well, now we have to create a message that attracts to them. Well, so we tell the agent, this is what you need to write this against. This is their ideal client. Help them come up with an answer that makes this match that.

Yeah.

So we're not feeding context or information into an agent when it's not needed. It only gets what it needs when it needs it, which is how we control the tightness of the conversations and how we make sure that they stay on topic.

I think that's a really good way of explaining our online training program. So if we have 12 modules, 4 exercises per, right? We didn't build one agent who is an expert on our entire curriculum. We built 12 times 4, 48 experts that all know one specific piece of one module to create one very specific kind of result. Yep. So, here's a great— I'd love for you to— because we've actually started to have these conversations with a few of our clients where you've gotten involved. Two, two real life examples that I think could come into play for an agency. When I think about where are they running a repeated process, where if you really broke the process down, it is a series of, of key points where an expert for that key point could be very helpful. For example, an RFP. Think about what you, I know you've had conversations with a client or two about this already. Where could an agency build something similar to the training we're building for an RFP where they could really leverage not just one expert who knows the RFP, but a series of experts?

So this is a, this is a long answer, so I'm going to shorten it up. It's amazing to me how many businesses, not just agencies. So this is not, this is not an agency problem. This is a business problem, have this same problem. And you can, you can change the, you can change it from RFPs to, um, how do I create pitch decks, to how do I do this onboarding.

Client onboarding was another one. Yeah.

And the first question that I'll ask, and I had this question, I had this conversation last night First question I asked, somebody asked, well, where can I put AI into my business? And the first question I always ask when they say, well, we need help with this. Great. That's a perfect place to use it. What's your process now? And I usually get blank stares. Well, we're not really doing it now. It's like, well, then all you're going to do is you're just going to do it wrong faster. If we, if we try to throw an AI workflow or an AI process into a process that doesn't exist, we're going to be building it from scratch. And it's like trying to teach a monkey to play chess. You're just going to aggravate the monkey. And usually that monkey is going to be the AI guy you hired to write this because they're going to get frustrated trying to get you to explain the process that doesn't exist. Having been there. I get to be a very frustrated monkey trying to play chess with a chessboard with no pieces on it. And so as an agency is looking at it, always start with, do we actually have a process for doing it now or do we just wing it? If you wing it, fix that problem first. Document, write down the process. Hand it to somebody that doesn't know anything about the process and say, can you follow this? If you can do that, you can put an AI workflow into it. It's amazing to me how many people think that AI does every single part of that workflow. And there's maybe one element of the whole thing where an AI agent is actually used. I've done RFP workflows for clients, and I think there's like 30 steps in the actual process of responding to it that most people do subconsciously. They don't even think about it. I mean, if I hand you an RFP as an insurance agency, if I hand you an RFP to go lay a new driveway, well, the first thing you're going to do is look at it and go, well, this isn't for me. But AI doesn't know that. It's— if you just tell it, create a response to this RFP, all right, I don't know how you're going to lay concrete, but I'll respond to it, and it'll write a response on how you would, you would create a driveway like no other, because it's trying to make you happy. It's what it's programmed to do. So it's amazing. How many processes, when people will describe that, won't even have that step? Well, we need to make sure that we're a good fit for it first. And I always like telling it to the story of when I was in elementary school, and even, even my sister brought this up the other day. When you have a classroom full of kids and you're trying to teach them processes and methods, especially in science, and you tell each one of them, Write the instructions for me to make a peanut butter and jelly sandwich. And then every kid writes down their process, and then the teacher's up the front of the room trying to follow this process, and the kids are all laughing because they're like, ha ha ha, not realizing that it's reinforcing the point that you have to be specific in the steps in which you're trying to accomplish. Otherwise, you are not going to get the results that you want. And AI is the exact same way. RFP processes, client onboarding processes— too many people see shiny tool, make things go fast, not realizing that they don't even know what process it is they're trying to solve.

That's a good— I, I hope everybody heard that. That to me was a bomb. If you're trying to use AI to build a process that does not exist yet, you're going to be in trouble.

You're going to, you're going to, you're going to create a process that just fails faster.

But it, it also made me— it kind of reinforced, Corey, why we believe our, our Online Training Essential program is going to be so effective. We didn't go to AI and say build a course from scratch, or build a course to build a process that doesn't exist. We've already got a process. It's our curriculum and how we teach it module by module. We went to AI then and said, all right, take an existing process now and enhance it at scale so people can use it whenever they want. Process first, automation second.

Well, if I remember correctly, we sat and we talked about how do we duplicate ourselves But in a way that's not detrimental, right? Because as a brand, I mean, you have a big brand, you have a big name. And the last thing we want to do is ever damage that. And in business, that brand is important. That name and that reputation is always important. We've spent, we're in March. 12 months working on this so far.

Yep.

This is not like, hey, I went to Lovable, said write me an app that does this, um, because it would not have been able to do it to this sort of scale and level and complexity that we've, that we've been working on for the past 12 months.

I, I know the, the phrase labor of love is often clichéd and overused, but it has definitely been a labor of love. But I think that the takeaway for agency, anybody in an agency that's listening to this right now is, you know, first challenge is getting your team to understand that what they do on a daily, not everything they do on a daily, daily basis is valuable. Because I think one of the challenges agency executives have right now who are trying to use more automation and AI is getting some of their team to understand. I understand you do that every day, and you, you might even place the value of your job on that, but that's not— I need you to be doing this, which is more useful. So it's getting the team to understand that I no longer want you doing this because we can automate it, but it doesn't mean I'm getting rid of your job. But then you've got to automate the things you do every way, every day, the onboarding, the RFPs, whatever it might be. But at the end of the day, going back to your point, if you don't have a clear and defined process from the get-go, going to AI and using automation to build it is pointless.

How many repeat— how many people did math without calculators? Calculators used every day now.

Yeah.

It reminds me of the movie Hidden Figures.

Okay. Didn't see that one.

Oh, so it's a story about NASA and they, the part that's relevant to the story is they brought in this large, huge IBM mainframe. That nobody knew how to use. And once they figured it out, they got to a point where it was, this output was important and it had two different numbers across two different days. Well, AI is going to do the same thing if you don't program it right. Yeah. Yeah. The conversations around The AI usage, how do I use it in my business? How do I use it to leverage and scale? Happens every day. And most of the time the conversation starts off with, I tried to get an answer from name your favorite AI LLM chatbot. and I couldn't get it to give me what I wanted. And the more I talked to it, the worse it got. And so I go in and explain just the exact same conversation we just had. It's not smart and it definitely can't read your mind. And so how do you use it to get the results that you want? Yeah, 99% of the time, AI is one little piece in a larger toolset that gets used to solve a business problem.

You know, I didn't, I didn't quite know what was going to come out of this conversation. I think you and I just hit record and said, all right, we'll see where this goes. We knew we had something to talk about, but I think the biggest takeaway for me is, you know, when I look at what we do for agencies At the end of the day, we're teaching processes, right? So if, if, if what we do better than anybody else is help agencies and their producers craft clear sales messages that get the right prospects to say, tell me more. The, the, how we get to that result is all through certain processes or frameworks we've built, right? Which allows us to go to an AI platform and build our curriculum on it because we've already got the process. Now we're using AI agents to enhance it as coaches. So when I think about how agencies can turn around, like, my big vision is, okay, we have now just taught them a process that is going to help their producers not only have a clear sales message but utilize it along across different prospecting channels to open doors with the right prospects. We can turn around, or they can turn around now and potentially use AI agents in their business to teach new producers how to repeat that sales message. So all of a sudden now what they have is an onboarding, a producer onboarding machine that when new producers are hired, they're already walking them through their process. Learn from us on how to craft a sales message that gets the right prospects to say, tell me more. And I envision our program a day where our curriculum is actually integrated in an agency, but designed where the agents are designed around that client, that, that agency's sales messaging. And the whole course, the whole curriculum is taught to teach the producer, is meant to teach the producer their sales message. Does that make sense?

Yep.

And now that agency has an— just like an onboarding process for their new clients, they have an onboarding process for their new producers that, um, literally keeps the guardrails around. This is the message, we're going to teach you how to say it so that you can go out and find the right prospects for our agency.

How many agencies actually have an onboarding process?

On two hands, I might be able to count that.

When, when, when you, when you think about that, it's like, what, what happens to an agent? I mean, I've heard it a lot. Um, I ran a mastermind this morning and the, the sentiment was the same. I was with another agency at one point, but this one acts like it cares more about me. And so when you have an agent that joins an agency and there's not that feeling of, oh, they've got everything together, I know what I'm doing, what I'm doing, the agency knows about me, they know what I'm capable of, I'm not just some afterthought that's waiting for some mentor agent to remember to take me on a sales call somewhere and introduce me to somebody. That there's actually a plan and it's being implemented and I'm held to a standard from day one. I've worked at enough places to know that if, if I'm working somewhere and I don't feel valued or if I don't feel important or if I don't feel like I'm contributing, why am I here? And I definitely think there's opportunities for processes and then leveraging AI in some of the parts of those processes just to help keep that momentum going. Yeah.

Well, we could probably sit on this conversation for the next 2 hours, Corey, but if you're listening in and you're still with us, one, thank you. Corey and I are literally just recording a conversation we intended on having anyways. But I think, Corey, I didn't know what the takeaway was going to be, but if I'm tuning closing in my advice to you right now. One, if you're still evaluating AI, because believe it or not, I have agencies who are saying that we're evaluating our stance on AI, please do something. It's, it's not a matter of evaluating, you have to use it. But Corey, I think the number one thing you said that I heard today is if you're going to use it effectively, develop, or at least uncover the processes you're already using, if not go to work developing processes so that you can use AI to enhance the existing process. Because again, it's not about having AI be the thought leader, it's enhancing the thought leadership in the, in this, in the processes you've already built. And I'll share this because we've been mentioning it throughout our Essentials program. I think what, what I'm so excited about around what we've built with our online training program, Corey, is we did it that way. We, we took an existing curriculum we call the Complete Game Marketing Playbook, took all 12 modules that have 4 exercises per module, and we have built 48 experts that have one job: to enhance that one small piece of our existing process so that a producer who goes through that training gets the desired result they want without us having to physically be there leading them. Yeah. Did I describe Essentials the best way I possibly could?

Oh, it's the perfect way to do it. It's the thought process we use to get there. It's everything, right? We didn't wanna reinvent the wheel, which was the playbook, because the playbook works. It's proven. It's been proven to work.

Yeah.

How do we make it more powerful and how do we get more engagement and how do we get more success and how do we get success faster?

Yeah.

Right. If you have a coach and you only talk to them once a month, I have people who will send questions, text messages, emails all the time. What about this? What about this? What about this? Well, now that agent's available whenever you want them to be, and they're only going to talk about one thing. If you get bored and you want to know what somebody's cookie recipe is, yeah, this isn't the place to go.

That's not the good one for you.

Now, I will, I will say this because it's funny you brought it up. And yes, we've talked about this. I talk about this for hours. So you're going to have to cut me off at some point. There's a theory or concept now for businesses who are thinking about implementing, and they call them AI studios. So it's an internal AI studio. And all it is, is there's somebody who owns this process, this workflow to evaluate and come up with solutions, but not in a vacuum. So you create your own place where there's one person who's responsible, or maybe even a group of people if you're big enough. But what you don't want is you don't want the Wild West where everybody is off doing their own thing. And it just turns into a cluster. It's a business process. It's a business tool. Evaluate it the right way.

Yeah.

Most businesses don't have experts that can evaluate it. So you need to create that.

Yeah.

If you don't have one, go find one.

Well, perfect segue because I wish I could say you are on my team and nobody can have access to you, but you also own your own business. If somebody wanted to reach out and learn how you could help them, maybe their agency, maybe they're in another industry, what's the easiest way to get in touch with you?

Ooh, I would say just go to the website, conifer.com, C-O-N-E-F-E-R, and start there because there's two tests that you take to see whether or not you're even in the right place. And guess what the first question is going to be? Do you have a process that you need to fix? Because if you don't have a process, you're not ready.

My takeaway from today, man, if you don't have a process, you can't automate anything. So get a process, then go to work using AI to automate. Well, man, it's been an awesome conversation. Like I said, I didn't, I didn't even know where we would go today, but I think this is going to be very valuable information for any executive sitting inside an agency right now who I hope is beyond evaluating, they're actually trying to figure out how to use it. I think your words of wisdom are very, very valuable in how to, how to best utilize AI to its fullest. So appreciate your time, man.

Oh, I enjoy it.

Everybody else, be good. Take Corey's words to heart and yeah, go to work using AI to enhance the processes you already have so you can use it the right way. Be good.