The AI Attorney: How Insurance Claim HQ Generates Leads with Artificial Intelligences

Jeffro (00:02.733)
What if I told you that one of the most aggressive property insurance attorneys in the country is using AI to generate leads and fight back against billion dollar insurance companies? Today’s guest is Galen Hare, managing partner at Insurance Claim HQ, a premier property casualty insurance law firm that’s recovered hundreds of millions of dollars for thousands of clients. Galen has been named a Super Lawyers Rising Star and one of the National Trial Lawyers Top 100, but that’s not why he’s here today.

Galen is actually on the cutting edge of using AI for lead generation in the legal industry, and he’s going to share his real world experience with what’s actually working. This isn’t theory. This is a practicing attorney who’s in the trenches every day using technology to find homeowners and business owners who’ve been screwed over by their insurance companies. So whether you run a law firm or any other service business, the AI strategies that Galen is implementing are directly applicable to how you find and convert clients. So in the next 25 minutes or so, you’re going to get

Hopefully a master class in practical AI implementation from someone who’s actually doing it and not just talking about it. So Galen, welcome to Digital Dominance.

Galen Hair (01:06.882)
Thanks for having me. I’m really excited to have this discussion today.

Jeffro (01:10.744)
Me too, and I love hearing people who are actually using things. But before we dive into the AI stuff, can you give us a quick background? What does insurance claim HQ do? Who are you fighting for?

Galen Hair (01:21.622)
Yeah, sure thing. So insurance claim HQ is a first party property casualty law firm. What that means is our clients are suing their insurance companies usually after something has happened to their home or their business, their actual physical property. like tornado fire, big disasters like wildfires, unique disasters like a home fire, hurricanes, things like that. And what we do is help those people, even if we don’t have to sue.

We help those people go after their insurance companies to figure out what they’re entitled to and hopefully get them paid in full.

Jeffro (01:54.391)
So what made you decide to focus on using AI for lead generation? Because I think the legal industry sometimes has a reputation for being behind in technology and maybe relies on referrals or networking or whatever. So how did you choose to go that route?

Galen Hair (02:07.446)
Yeah, so that reputation is completely earned. Law firms are very, and there’s something wrong with that per se, right? It makes sense. Law firms are really hesitant to jump into new technology period, much less AI, because what if it breaks? What are my ethical obligations? Data security is really big for us. We obviously have certain obligations that we can’t be, you we can’t fall behind whatsoever. And so I get it. And I understand the rep.

Jeffro (02:25.717)
Mm-hmm. Yeah.

Galen Hair (02:36.642)
the reticence, but I’ve always been really tech forward. I was always the person buying the newest, coolest computer. I had a Palm pilot back when that was a thing and then a Blackberry. Been stuck on the iPhone for a while now, but always was always kind of getting the new thing. it wasn’t, AI for me was kind of like a natural step, right? We always at our firm had the newest technology, the best technology after, obviously after we vetted that it actually worked.

Jeffro (02:47.733)
Nice.

Galen Hair (03:06.974)
But we did those things. So when AI became popular, know, popular, I’d say two years ago, but really known to the public four years ago, we were already playing with it. We weren’t using it, especially because of the data security issues that so many of the big AI providers still have today. But we were actually able to jump into marketing easier because we don’t have quite as much of the same data security issues. Because they’re not existing clients.

Jeffro (03:33.138)
Right, OK. is that right? So is that the main way you’re using AI to get those leads? Can you walk us through that process?

Galen Hair (03:42.134)
Yeah, I’ll walk through the lead, but today, no, we use AI across most aspects of the firm. But AI has gotten better. It’s gotten more secure and you can host things. And especially now with the stuff, this is super timely with the digital employee side of everything with the actual computer run AI models using things like OpenClaw. We’re even changing and I’ll cover that, but we’re even changing the lead generation side of that. So now we have actual digital employees that work for us at the firm.

Jeffro (04:12.222)
That’s amazing. And that probably saves a lot of money too.

Galen Hair (04:15.616)
It does. We have to be careful, right? So generally speaking, the digital employees are not typically allowed to communicate with outside world without an attorney signing off on that specific communication. So, you know, like Eva, for instance, she can receive a client text or email. She can summarize it. She can send it to the attorney. She can even draft the response, but the attorney has to click send, right, or edit. So that’s kind of how we provided those safeguards. But on the marketing side, we just…

Jeffro (04:25.214)
Make sense? Okay.

Galen Hair (04:43.884)
We still want to obviously make sure we’re doing the right thing, but we don’t need quite as many safeguards. Our safeguards are really just ethical. We still can’t solicit. So we’re not soliciting. So some of the things we’ll talk about today, a non-lawyer business could take that to the moon. We’re very limited in what we can do, right? So the first place we looked to use AI and marketing counterintuitive, considering what I just said, was with existing clients. We’re like, okay, we’ve got this database of thousands of clients that have had claims around the country but weather keeps happening and things keep happening. Why do we only get repeat clients when they think to call us? Some clients don’t even remember who their lawyer was like five years later, right? So one of the things we started using AI to do, and it was very clunky and manual before the age of digital employees, and you had to pay a lot of subscription costs. Now it’s much easier, but we were essentially using websites that allowed us to pull reliable hail and wind data and big weather events and then we were cross-referencing that across our address database with our existing clientele and then using that because we can solicit existing clients or past clients. We were using that to have our intake team reach out to those folks and just check on them after the wildfire, after the big hailstorm. And then if they came back and said, yeah, I do have some damage, we said, well, we’d love to help you again. Cause that was ethical and we could do that. Now it’s way easier.

Jeffro (05:50.692)
right

Galen Hair (06:07.414)
with the digital employees because building those APIs, think, you know, all those to get integrated into all those different sites. If you’re sitting at home trying to figure out how to do that, that was a pain and you really had to have a developer help you. And to some extent you needed the cooperation of the different systems you wanted to use and they didn’t always want to play nicely together. So now with digital employees, our folks just have logins to some of those systems are AI folks, I should say.

Jeffro (06:24.218)
Yeah.

Galen Hair (06:34.646)
and they can log in and they can point and they can click and they can just essentially do the API without the API, which has been huge.

Jeffro (06:41.412)
Yeah, so let’s walk through the actual tech stack. Can you tell us exactly which tools you’re using, which platforms you use? You’ve alluded to a couple, but I imagine there’s a lot.

Galen Hair (06:50.678)
Yeah, so right now, and I’m not the tech guy, maybe you needed the tech guy on here, right now we’ve got a computer running where the actual assistant, Liam, works. They have names because they have actual email addresses and they can use things. So Liam can access our entire database, knows where our clients are, and can make his own reports. He makes his own reports every night.

Jeffro (06:56.307)
Don’t have to go too deep.

Galen Hair (07:18.23)
You know, and what he does is he goes into a system called Hale Trace, which we have a subscription to. pay them for that, for that right. and he goes in and he looks for where the newest hail is above a certain threshold. We don’t want tiny little quarter sized hail. That’s not going to damage most things. That’s not worth it. But when there’s big hail, like just had six inch hail, which is insane, just outside of Chicago, he goes, Liam goes through says, okay.

Six inch hail in Chicago, that’s a market we need to look at right now. He goes and pulls our entire database of Chicago based people. Then he looks at what cases are ongoing. He flags those for the attorney that is currently representing those people to reach out and ask, hey, did you get through the hailstorm? Okay, last night. And then the others he sends to the intake team all organized and uploads it into our actual intake software, which is lead docket. So he actually puts those leads in the queues for the different intake people to call them.

We can have AI call these people, but we still want a real conversation to happen. So that’s kind of, that’s the process. Now, as far as tech stack, you’ve got OpenClaw, of course. I believe it’s currently communicating on GPT and Claude. Don’t quote me on that. And, you know, as I said, you have to give, if you’re going to go the digital employee route, you have to give them actual logins to all these things. So he’s got a Microsoft Office suite.

you know, he can get in and out of browsers and he can click around and do these things.

Jeffro (08:48.855)
Very cool. So, and one thing I want to highlight for the listeners right now, because people are in different industries, right? For what you do, there’s a very clear path from trigger or event to needing help with an insurance claim, right? So that’s a, can monitor hail, wind, tornadoes, whatever. And that’s the trigger that you reach out, you know, and make a touch point. Other industries have to think about what those touch points are. Instead of just waiting, you know, if you… make websites, maybe you think, okay, find a startup that just raised funding. Now they’re ready for a website or I don’t know, things like that. Look for those triggers, because then you can monitor that. Like you can pull funding data of new, you know, recent startups or something. So stuff like that, that’s just a really good insight that I think people need to pay attention to.

Galen Hair (09:37.846)
Yeah, you have to have your touch point, right? And for us, it is obvious, but look, we’re also constricted. We can’t reach out to someone unsolicited ever unless they’re an existing client. So that works for us, right? Theoretically, if I could solicit, here’s what I would do if I weren’t ethically prohibited. I would actually have him do the same thing. He would pull the hail trace and then I would have him use probably LexisNexis or some data source. We use a few of them.

But I would have him use something like that to go pull addresses, phone numbers, emails, ideally cell phone numbers. And then I would have him create a custom audience on Meta and the various social media marketing platforms using that group. I would have him start feeding them ads. And then I probably wouldn’t use a real person for this because there would be too many people. I would have AI agents, which we have a couple of voice agents. I would have them dialing all day. And I would also send hard mail.

I would send mailers out and I would send texts. So in an adjacent industry, that’s how I would probably capture those people. And I’d probably do a frankly much better job at it than we can given our restrictions.

Jeffro (10:49.876)
Yeah, no, guys, pay attention, take notes, do this. It’s not coming, this is possible now. The tools are there, you just gotta commit a little bit of time and effort to get it up and running and be willing to try something that might not work the first time. So, Galen, do you have like an aha moment where AI did something you couldn’t have done manually once you started using it?

Galen Hair (11:12.62)
yeah. I mean, I’ve had a bunch of them over time, but one for me that led to us doing the digital employees, right? Was I was on a zoom with someone. It was a kind of a weird referral call. They mentioned that they were really looking to connect with real estate brokers, but it was such a pain and there was really no one source to pick them up. And I said, okay, well that makes sense to me. Like I get that. And, but there was a website that basically lists all the brokers.

And he was like, yeah, so I think I’m going to have to pay someone in the Philippines to go do this. And I mean, this was probably, this was eight months ago. I, I think it was eight months ago. I used one of the AI platforms. I think it was just chat GPT, but, I had to, I had to work it with another one and I don’t remember the second one too. And we scraped the entire website, created an Excel, Excel spreadsheet with all of their info.

And he was like, oh my God, like I would have had to pay thousands and thousands of dollars to get this done. And I was like, oh man, there’s, I’m sure there’s so many things we’re doing at our firm. Cause I have a hundred live real people employees, right? I’m sure there’s so many things we’re doing at our firm that are causing my people to pull their hair out. And if we can automate some of this for them, there’s no reason not to that. That gets them doing what I want to pay them to do, which is to think and.

Jeffro (12:13.234)
do that.

Galen Hair (12:34.006)
not do data entry and boring things. And I would say that broker list, when I realized that and like saw literally AI save thousands of dollars, like it made all the messing around the prompts late at night worth it, you know?

Jeffro (12:45.607)
Nice. Well, that’s a good example. What about something that you’ve tried with AI that completely flopped? Do you have any expensive lessons you’ve learned?

Galen Hair (12:55.214)
Yeah, I mean, I wouldn’t say expensive, right? But some of these, here’s what I don’t like. And it’s because it’s where AI is moving. You know, some of the big LLMs are going to start offering ads. They just are. We know it’s coming. They’re already testing it. ChatGPT is definitely testing it. I mean, I know because I’ve been invited to test ads on ChatGPT. what is it doing now? It’s trying to keep its user happy.

And that has been a really frustrating moment for me because I feel like chat GPT and I now are in like a relationship that’s starting to go awry because I used to be able to have conversations with it and I felt like it would level with me. Right. It doesn’t do that anymore. And you know it tells you whatever it has to tell you to keep you coming back. It’s it’s as close to one of those like

sentimental abusive relationships you can have, right? I mean, but it really is, you know? I mean, it just told me the other day that my friend could probably still go skiing with an umbilical hernia. And then finally I had to push it six times. Well, what are the risks? And the risks were really bad, but it knew I didn’t want to cancel the ski trip. So it kept telling me, yeah, you two should just go. Well, this guy could have hurt himself seriously and ended up in an, in an ER getting emergency surgery in a place that

Jeffro (13:56.677)
Yeah.

Jeffro (14:14.425)
Yeah.

Galen Hair (14:23.168)
he didn’t live and then we would have had to get him back to Vegas because that’s where he lives on like a care flight because he wouldn’t have been able to sit up in a seat, you know.

Jeffro (14:31.513)
Yeah, I think a lot of people are starting to notice that, especially with chat GPT, it is more of a yes man than some of the other ones like Claude. So a lot of people have been switching and personally I use Claude and I like that it does push back. Because once I’ve given it the context of what I’m working on, I don’t have to ask it to push back. It will notice the discrepancy and say, hey, why are you asking me to do this before you said that? Are you sure? And so that’s been really helpful.

Galen Hair (14:56.95)
Yeah, no, and you have to, and with CHI-GPD, you can instruct it to stop doing that, but you have to know where to go kind of in the back end to do that. And it’s gotten frustrating. We have another good example of that, I guess, is we have a voice bot that does reception for us, including me personally. So if you call my cell phone and I don’t answer, an AI agent picks up and it’s really great. He remembers things about people. He takes notes. He knows my girlfriend, right? He’ll ask her things about her life, assuming I’ve paid enough attention. You know, he’ll…

He’ll keep up with her, right? But what he can’t do is hang up. And you just have to test these AIs because there’s this hilarious call where I guess she called, she was in the bath, she just wanted to chat. So her hands were wet, so she didn’t want to touch her phone. And she ends up on the phone to him for like 30 minutes while she’s in the bath, because he keeps saying goodbye. He keeps saying he’s going to hang up. And then he stays on the phone. And you’ve got to teach your AI to be honest about its capabilities. Because if you just assume…

If you go into one of these systems and say, Hey, here’s my business. Here’s how one of January leads. Can I do this? Some of the systems will say, yes, you can. So you still have to test it like six ways from Sunday. And then you also have to like ask it, be very honest, like double check everything. I mean, I will tell you, this is very rudimentary, but one of my most standard prompts when I’m doing something in an LLM, after I get the word product, after I get everything, and I’ve been through a bunch of prompts to refine it.

is tell me the truth. Is there anything in here that is inaccurate and is this document the best it can be? And sometimes I have to do that 17 times when I’m working on a project.

Jeffro (16:34.412)
Yeah. And that’s, that’s important to remember too, for like coding projects, especially if you’re not techie, like ask it, Hey, are there any security risks to this? Are there going to be performance issues if I get more than 10 users or like, you got to think that way. Like, cause yeah, it’ll spit something out that might work or look good, but you’ve got to press it to be better.

Galen Hair (16:56.172)
Yeah, I mean, even the digital employees, right? 12 little computers running, open call, all these things. Never thought the AI that helped us build it, the programmer that helped us build it, and me personally, never thought for a second, we need to turn off automatic updates. So, you know, one day it’s like Wednesday afternoon, our whole team’s rocking and rolling and they’re using these digital employees for everything. And it’s like they all walked out and quit on the job in the middle of the day. You know?

Jeffro (17:12.34)
Hehehe.

Jeffro (17:21.184)
Yeah.

Galen Hair (17:21.934)
And no one can figure out why and we go in and sure enough, like half the computers are off. The others are on like a login screen and you know, it was, it was crazy. So you just have to like, you can’t, you can do so many cool things, but the aha moments really come in the failures and, you have to embrace those and more than just, okay, chat GPT sucks cause it didn’t do this. You have to be like, what can I do to get the right, the right response that I need.

Jeffro (17:48.224)
Yeah. All right. Well, let’s get back to some of the things that are working. So we talked a little bit about how you’re using it for reaching out to existing or past clientele, but for finding new customers. You can’t solicit directly, but I imagine you can use AI to gather information, contact info to pass that to a team member, like drafting stuff. Are you using that for blog content and things like that as well?

Galen Hair (18:10.924)
Yeah, so for sure. So yeah, we can’t reach out to people we don’t have a relationship with. I will tell you what you can use AI for before I jump into that question. That’s really helpful is these days there’s so much big data out there, but the process of getting that data created in those lookalike audiences or even the direct audiences is so expensive. You can get the base level subscriptions to some of these data companies and use AI to actually pull all that info, create the database, upload it in to the social media marketing. And that’s something we’ve done really successful. As far as content, I mean, I don’t even know where to tell people to start anymore. There’s a million out there. You can get blogs uploaded. You could have them uploaded daily if you wanted from decent generators. You can have social media content created and posted. I mean, and the price is going lower and lower and lower. I think I was pitched something like $1,000 a month to get website, dev, SEO support blogging and social media posting from an AI vendor. I was quoted like, like five or $6,000 a month by an actual agency, right? I just got an ad fed the other day and I click on these and it was like $129 a month for all that. So with the token costs going down and everyone doing it, if you are sitting at home and like you’re doing all your own marketing, you are wasting so much time and money right now because you need to be building your business out.

Jeffro (19:21.694)
Wow, it’s coming down.

Galen Hair (19:34.05)
And you should absolutely, even if you do nothing fancy, right? Nothing like we just talked about with computers. Like if you think open claw is like a crustacean that you eat and you’re not comfortable with any of this, like you still go get one of these out of the box like Sintra or AO or marble, marbleism, know, where you get like 12 little AI agents for a flat fee. And that alone will start to do better than you’re probably doing current.

Jeffro (19:58.609)
Yeah. And for those of you listening who are wondering what OpenClaw is, somebody took Claude and basically turned it loose on a machine, gave it full permissions, access to do anything and everything so that it is the user of that computer. so OpenClaw, was OpenClawed, anyway, so whole thing. But that’s what it is. It’s an AI driving a computer with access to anything you give it. So that’s why I was talking about having accounts that it can log into.

It can literally open your browser, navigate to a webpage, click things, do things, interact with the file system and do anything that a human could do in that type of role that we were talking about. So it is very powerful. Obviously there are safeguards you need in place for that sort of thing. You have to be aware of what you’re giving it access to. And if it can delete things, maybe don’t assume it will do everything right the first time. So have some backups in place.

Galen Hair (20:51.468)
It will absolutely delete an entire database if you give it the ability to, because it’ll go crazy one day and think it’s doing the right thing. and, and AI does overwrite. on marketing, you got to be very careful with that, right? Your marketing history is really important because your ad, your ad account, whether it’s a Google ad account or Facebook or whatever has a lot of value in it. If you’ve been running ads for years, it’s not just that past campaign, that past campaign might’ve been good or bad but there’s data to be garnered there that you can also use AI for, by the way, to figure out what campaigns work, what campaigns don’t, and why. And if you just turn AI loose into those accounts without very specific safeguards or access limitations, it sometimes will be like, I did a new campaign for you, so I deleted all your other ones because they were trash. And you’re like, well, that was really helpful. I really needed that data.

Jeffro (21:40.347)
Yeah.

Galen Hair (21:43.662)
Because I don’t you think you did a better one, but I would like to actually see if we sign any customers, you know So you got to be very very

Jeffro (21:51.323)
Yes. now that, you know, it’s obviously getting better. Right now you have an advantage using AI because a lot of other competitors are not. So how long do you think this advantage will last before every attorney is doing this?

Galen Hair (22:08.014)
Well, I think where it comes is that technology will continue. The short answer is forever because by the time most law firms jump in, there will be something better. Right? I would say right now the gap is incredibly wide between like, I mean, it feels like the last two months AI like literally shot a rocket to the moon. So now there’s people on earth and like, we’re pretty far away. I do think over time that gap, like AI will advance a little bit but people that are already using AI will be here and then it’ll keep doing this. And you know, there will be a lot of people just kind of chasing AI. But for now, I think I suspect we’ve got a good solid knowing the way law firms work three to nine months where most law firms won’t be doing anything. They’re just now starting to delve into like, should AI help me with legal analysis or should it not? And judges are coming out with sanctions orders. So a lot of law firms are very terrified of AI, but.

Jeffro (22:36.923)
Mm-hmm. Yeah.

Galen Hair (23:04.962)
We don’t allow ourselves to be put in that position because we’re never going to do anything like that that would allow AI to file something false with the court. It’ll never happen.

Jeffro (23:12.463)
Right. Got it. So for somebody listening, they run a service business, maybe not legal, and they want to start using AI for lead generation this month. Where would you recommend that they start?

Galen Hair (23:24.874)
If you have zero tech knowledge about this stuff whatsoever, I would use one of the big current out of the box programs where you get like 12 agents that’ll help you. I mentioned those, it’s like Sintra, AO and Marblism. I think I saw recently, I’m not even endorsing those. I know nothing about those products other than I’ve read the ads played around on their websites and downloaded them like once or twice. That’ll get you started. If you’re already past that level, then you should be building your own stuff.

So I would suggest looking at OpenClaw. There are plenty of marketing AI companies out there that you can hire, but vet them carefully and check their pricing because the cost of the tokens is coming down. That if you sign up for an expensive Cadillac program for 400 bucks a month right now, especially if you’re in a contract, six months later, people will be getting the same product for like a hundred.

Jeffro (24:16.287)
Awesome. Well, Galen, I love it when episodes are super practical like this one and we get to walk through ways that people can use tools in their business right now. And I think this is so important, like you said, because AI is moving so fast, guys, the sooner you start, the farther ahead you’ll be than everyone else who’s waiting till next month to start, right? So the stuff that you’ve shared today, you know, it’s really helpful and I love that it is applicable to anything. It’s just changing the inputs and outputs, right? But the tools are the same.

You can use these principles to find, convert your ideal clients and just make your process more efficient. And if you guys are listening and you do want to learn more about insurance claim HQ or follow what Galen’s doing, where should they go Galen?

Galen Hair (24:58.882)
Yeah, so hit us up on our website, insuranceclaimhq.com. You can follow us or me personally on any of our socials too, and I will respond to you directly.

Jeffro (25:08.118)
Awesome. And guys, if you’re listening, here’s your last takeaway for today. AI isn’t coming someday. It’s here, right? I’m going to drive that into your heads. Galen has proved that early adopters are building real competitive advantages right now while everyone else is waiting and watching. So the question isn’t whether you should be using AI for marketing. The question is how far behind your competitors do you want to fall before you start? So get going. But thank you guys for listening to Digital Dominance. Share this with a friend, leave us a review and take care. We’ll see you next time.