How AI Is Helping Companies Redesign Processes

AI is just software. That's the way that I view it. It's a tool, it's an amazing tool and that's how I explained it originally to our team to ensure that everyone understands that this is not going to replace you. It's like putting an exoskeleton. Getting really geeky here. Putting an exoskeleton on you to be able to run faster and do things that you wouldn't normally be able to do.

Here we go. Another day, another systems and AI implementation case study. Today I'm joined by Chris Hunter. Hunter, he's from roofingsites.com they're our marketing company that helps support roofing companies. I love it when it says what it does on the tin. He's a long time Systemology supporter and has recently been re engineering some of his internal processes and systems and workflows to be able to leverage AI. And that's really what I want to talk about today. I want to understand his thinking, how he did it, the steps he went through and more importantly how you can do the same. So maybe first things first. Just like to welcome you to the call, Chris.

Awesome. Thank you so much for having me on. It's really cool.

Yeah, nice. Well, maybe as a place to start, I'd love to know how long it took you from seeing what AI can do. And I don't know if it was the ChatGPT moment for you to go from idea to execution. Was it something was like immediately obvious to you or did you sit on it a bit?

Yeah, so that's a good question. We first started using AI the first time I kind of came across it was. I had come across Jarvis, right was the name of the, the tool way back in the day. And that was because it was, it was, it helped us with content, right. And it helped us be able to create really good content, much better than sometimes than what our writers was creating, you know, and they helped us create all sorts of things like email campaigns and all, all those types of things. I think that they had to rebrand it at one point from Jarvis to Jasper I think, but which is what it's known. I think still right now we don't use Jasper anymore, but we did, we were heavily heavy users of that for a while. Chat GPT, you know, got released. What was that now, two, three years ago? Something like that. Two years ago seems ridiculous, seems like forever. But you know, it when it was released that kind of hit our radar and we started playing around with it. And when I say we mean my you know, team and, and we do a lot of content for our clients. You know, know, we have to, we're, we're mainly search engine optimization based, you know, and so that revolves a lot around content. We do some social media posts and, and some other things. But we started playing around with chat GPT and we're like, okay, well, this is kind of cool. And then when they started releasing some of the more advanced types things is, that's when we really started leaning more heavily into chat GPT specifically. Okay, when they released custom GPTs, that, that really made me sit up and, and go, huh, what, what can we do here? You know, how can we. And so basically I started with, with one, you know, custom GPT that I created specifically for the blog articles, right? And creating blog articles and creating blog articles that were lengthy in nature, that weren't robotic sounding, you know, was kind of the goal. And so I, I created it, let my team have it, and they went, they loved it, you know, and, you know, started using it quite a bit. Then I started listening to them and you know, in our level 10 meetings, you know, in this EOS type speak there, right? But in our meetings, in our weekly meetings, I started listening, okay, well, what else can we create, right? What else can we do with this and what other problems can we solve? And so we started doing that and, and you know, next thing I, I knew, I, I'd create an, you know, I, I'm, I, I, I just looked at my account, you know, right before this. I've got 18 custom GPTs that we've created, you know, for specific tasks to solve a specific problem. You know, anything from press releases to social media posts to actual assistance that are customized specifically for us. Like, you know, I was, I was telling you earlier that.

I, I have a, a custom GPT that I've created as a, as a personal assistant for myself, you know, so it's trained on me, it's trained on my processes, it trains it on my book, right? It's trained on this right here, 336 pages of my book here. And it knows a lot about being my team and so forth. And so I, I give it specific instructions. I'm like, hey, this is my task list, you know, for the week. Who should I delegate this to? You know, what tasks can I delegate off? And it knows the Colby scores for each of my team, so it knows who's really good at what kind of a thing. So that's kind of cool. It's really geeky, but that's what I'm into, you know, and I like, you know, kind of digging into specific tasks that I can delegate or, you know, offload onto my team kind of a thing.

So yeah, I love, I mean, first step there that someone listening to this should take is having a look at what tools they're currently using that are starting to have AI creep into them and then start to make use of that or maybe even seek out some specialized tools. You think of something like Jasper and it was tapping into AI paper back before the chat GPT moment and it was effectively putting a wrapper over the top of AI, creating prompts where you wouldn't necessarily know they figured out their recipes that were going on. You'd populate the different fields and then it would give you the result by tapping into the AI, which was quite magical. Before people kind of really saw what was going on and now you kind of migrated and thought, well, we might as well go straight to the source here. And then you can start to control your own recipes, control your own training and then start to build it out for different use cases. So I know you're a bit of a systems thinker, so you think in terms of the different processes that are going on in your organization, how did you start to think about prioritizing or where did you start? What were some of the first processes that you kind of thought? Oh, that, that's an obvious move.

Yeah, so that's a good question. I. The, the biggest thing honestly is what's driving us most nuts, you know? Yeah, and when we first started, when we first. So we're EOS organization and, and we set up in 2018 with EOS and when we mapped out the whole process and what it's going to look like in five, 10 years on where we're going, we were going to have to have like 50 writers and editors and, and, and I just, as we did that and, and you know, hit our, our big hairy, audacious goal, I just started thinking, how in the world are we going to manage 50 people right at the time, I think we had four, you know, on the team and maybe five actually, because we had four on the leadership team. You know, we were very leadership heavy. So.

Anyways, when.

The biggest thing with me on any kind of system has been what's driving us most nuts? You know, what do I need to offload of me specifically, what do I need to offload of my project manager? What do I need to offload, you know, and, and think of those things in those kinds of terms. And really, that's according to your book and what I, I read in your book too.

So yeah, it's definitely the first step in moving to AI is getting very clear on what the core processes are, how you're currently doing what you're doing, because that ends up becoming the programming for the machines anyway. And then you look at re engineering and taking off certain chunks and say, oh, okay, well that whole step now we can replace it with a prompt and go draft it up over here using our custom GPT and then the writer or whomever can step in and do the polish. And you know, we just saved 80% of that team member's time. So I think that idea of identifying the processes and what's going on and then thinking about where the pain is and then thinking, well, let's go to work on that and see how AI can help you sort of go faster, cheaper, better. I know you touched on earlier. I'd love to dig into a little bit more with an actual use case. Let's say the blog articles that you mentioned earlier. So being a marketing company for roofing companies, one of the services that you offer is to provide them with blogs that can help with their search engine optimization. So that's obviously a big chunk of what it is that you do. Is that something like when you went through the systemology process and thinking about systems, had you, have you actually documented out a process that the team was following pre.

Applying AI 100?

Yeah. So and kind of backtracking. We've been in business since 2003, right. When we first started implementing search engine optimization services for our clients in around 2005. So we've been doing it a really, really long time. And, and I'd like to say yeah, we do have it systemized. And honestly the content piece of, of our, our organization is because it's the one that we've been doing the longest. So we had SOPs, we had tasks, we had checklists, we had all of these things lined out. Everyone was trained on exactly what to do for those things. So implementing it and with AI was just simply okay, well what task can I pull out? You know, and the writing tasks seem to be the most logical piece to start with. The other task, I mean, and, and you know, we didn't really talk about this yet, but like just coming up with titles. Right. My editor is the one who comes up with titles and that that was taking up like 80% of her time every single month. And so we systemized that using AI to come up with kind of a. And what's now known as this topical authority method of, you know, finding out all of these really great titles that we could use to start with for that, for the articles. So yeah, it's just been one piece at a time.

Yeah, yeah. And how do you think about, let's say specifically with the blog writing, you've got a GPT that is trained for that particular task or do you have multiple, like you've got a headline GPT, you've got a body GPT or is it just one GPT?

So, so I have a friend that helped me with this, honestly. His name is Jonathan Mast. Super, super smart guy. He's in the same mastermind group that I'm in and every time we met we meet up like three times a year in person and, and quite a bit more online.

He's super deep into AI. He's created in a whole AI agency that does nothing but content for other agencies kind of a thing. Right. So he helped me kind of with that prompt of how to create that custom GPT to create the, the article assistant itself. I then took that assistant. Well that, sorry that I'm using the wrong terminology. The, the, the custom GPT and created in the open AI back end. Right. And in the API area and assistant. Right. And now my programmers tied that in with our software that we use to manage our entire business that we've created over the past 10 years now to, to help with that entire process. So it's literally almost like Jasper inside of our own software, right?

Yeah.

And it's able to, you know, generate the articles. Generate the titles. Well, titles first, then the article and then it, it, we have editors and they go through and, and edit it. But I think we're down to like five minutes per article, right. That we're editing. So it's, it's, it's, it's. This time savings is just massive, just absolutely massive.

And how long would an article pre AI take you?

So, so we paid our writer per article. You know, it was like, you know, every single article that she wrote and she's been with us like 15 years.

And by the way, I, we trained the AI on all of her articles that she's written over the past 15 years. So there was like 15, 000 articles or something ridiculous like that that she had written. But we, she was, she, you know, that was a per article thing that, that she was, you know, charging us. Our editor on the other hand, right. Was spending probably an hour to two hours editing, you know, and, and I think that's just because of her process of what she does, you know, on in and making sure that the facts are checked, that, that you know, what we're publishing is actually factual, that it. And that it reads correct. So now going from maybe a 10 minute per article process and I'm just paying, I've elevated our, our writer now to.

Content manager is what we're calling her. Right. And so last time I checked it was about 10 minutes per article that they're spending on those things. So yeah, massive, massive difference.

Ah, hey, I know you're tuned in right now, so I'll keep this super quick, but you like awesome YouTube videos, don't you? Well, there's not really that many YouTube channels that create great content about how to systemize your small business. So I want to make sure that you catch every video that I publish. So if you haven't already, firstly give this video the thumbs up and then go ahead and subscribe to my channel. You can even hit the bell notification to make sure that you're the first to know when I publish anything. So go ahead, do it right now. I'll wait a sec and then I'll transfer you back to the video.

Yeah, and then, and that's going for a couple of hours and then if we multiple or an hour. Yeah, yeah. So we're kind of talking, you know, 1/6 the time or something like that. And then multiply that out across all of the articles that you're writing. Yeah, and then that pretty much drops right to the bottom line.

Absolutely. And, and more importantly, it allows us to scale because that was the one piece as we were looking to scale that is not scalable in agencies. Right. And until AI came out, it's not really available. I mean it just, you just have to hire more writers, train them and manage them. And the management piece is the hardest of of them all. But you know, and now it's, it's, it costs us literally dollars per article, you know, maybe less than that. Honestly, I haven't looked at the cost savings. I, I do know that it cost us about, in the open AI end about 12 to 15 cents per article. Right. When you're paying the API fees.

So, and then prior, like it's probably might be hard even if you kind of ballpark it, writing it before and with the editor. Do you have a ballpark how much an article was costing you?

Yeah, so it, it was costing us about $50 per article, roughly, you know, around that range. And then we would charge $150, you know, per article and then cost now like you probably negligible. I mean it's, it's ridiculously low. And those two people, the editor and the, the content manager. Right?

Yep.

Can, I mean we, we can easily handle a thousand clients I think with those two people.

Yeah, yeah. So I mean there's a couple of things you mentioned there and I think it's important to pull them out the steps, like you got very clear on the process, like you know, systemology thinking, critical client flow departments, core process. Great. Now where creates, where's the pain? Okay, well let's go in on those and what a high volume, high, you know, frequency where savings can be gained. Okay, well let's re engineer that and then start off at the moment. You know, when you start off, it's a custom GPT and you manually doing it and loading it in and then you've gone the step beyond that, which is then. Well, now let's have a custom coded solution on our different prompts that's tailored specifically for our use case that now becomes part of your 1111 herbs and secret spices. It's the special tool, it's something that's proprietary to you guys and now becomes that real competitive advantage. So I think that, yeah, that's key. And then you just multiply this out across all of the different tasks and functions and going on in your business and then, yeah, those efficiency gains just stack and it compounds and then you have significant impact. So one question you mentioned about with the writers and you've got the same writer that's been with you for.

Should we say, about 15 years. Yeah. So.

How has the team gone embracing AI? Like is this something that they were hesitant about? Were they wary that they were worried that this is going to take their jobs? Love some insight on that.

Yeah. So, and that was never our intention with, with AI was never to replace anybody. It was to use it as a tool to make that person way more efficient. Right. And by a multiplier of, I don't know, probably 50 to 100.

Right.

Kind of efficiency, which is really just amazing once when you get down and think about it. But you know, AI is just software. I mean that's all it is. That's, that's the way that I view it. It's a tool. The way that I explain it to my roofers is that it's like the air hammer is right for, for putting shingles on, On a roof versus a, a regular hammer, a 16 oz hammer, you know.

But it's, it's, it's a, it's an Amazing tool. And, and that's I think how we should all view it. And that's how I explained it originally to our, our team to ensure that everyone understands that this is not going to replace you. This is to augment you. This is to, you know, it's like putting an exoskeleton, getting really geeky here. Putting an exoskeleton on you. Right. To be able to run faster and, and do things that you wouldn't normally be able to do. And it's, that's the way that I view it.

Yeah. Yeah. I think that's a great way to explain it through to the team and it makes them stronger and more valuable. And I actually think team members that embrace AI will outperform those that don't tenfold. So you want to get on board this train and that will help you become more valuable because I think the AI role, the systems and AI champion is such a critical role for someone to understand process, understand how to re engineer and introduce AI. The, the savings and the benefit to the business are so significant. And we've just taken one little piece, one little example. But yeah, think about every role, every department. There are these wins just waiting to be taken advantage of. I mean in the tail end. Yeah. I don't know if there's any additional things you maybe other use cases. I know we touched on a couple. You talked about using a, an assistant or training up an assistant GPT for yourself and that helps you to delegate out tasks because you've trained it on Colby scores and understanding your team. Are there any other things where you're just like, oh, that's been a real winner for us.

So lately. So I sit in the sales seat, right. In my organization, that's my last major seat to give up, I think before I elevate completely to the visionary seat. But I've honestly been using it to evaluate my strategy sessions and discovery calls that I'm having with clients. Right. And everything is recorded so it's easier. And I use Read AI, of course, AI tool. I use Read AI to record all of my meetings with clients or potential clients. And I basically, I take that transcript of that down, you know, download that transcript, feed it into my assistant, right. That I've trained to be to act as Jeremy Minor is who I'm using lately, right. For, for sales coaching. And I, I tell it, critique this. What should I have asked? What should I, you know, how should I presented this? How what should I say next time? What are my follow ups? What should my next follow UPS be and it, and it's giving some really insightful information and helping my entire sales process there. So. And additionally helping my assistant. Right. I've got a sales assistant that is doing the discovery call. So I'm helping, it's helping me to help coach her.

Way better. Because I'm not trained, quote unquote, in sales. Right. Even though I'm, I'm like, that's my role.

I view myself as more of the back end kind of guy, but I've been thrust into that area because who else is going to do it? Right?

Yeah.

And so I, it's been really helpful to me to be able to, to coach my assistant and hopefully to be able to create SOPs based off of all of these trainings. Right. For my next replacement, which is going to be a closer, you know, type person.

Yeah. Yeah. I love it. The more that we talk about AI and the direction we're headed, the more that I realize this, how critical process is. It's really about identifying those critical processes. And then as I say, the systems that you document today will become the programming of the machines tomorrow. I'm curious if you see yourself as a systems person or if it's something that, you know, you've, you've just come to realize how critical it is in business.

Yeah. I think that the longer that I've been in business, the more that I learned that businesses are simply a collection of systems.

Right.

And so the more systemized that you are with every bit and piece of your business, the better that the business overall is going to function. Right. It's easier to train people when there are systems in place already before you, before you bring them on. And so I'm not perfect by any means. You know, if you talk to my business partner, he is way more systemized than I am. And he always, constantly asked me, why did you do that? That doesn't make any sense. You should have done it the systemized way. And I'm like, oh yeah, that's right. Okay, so I'm, I'm not perfect by any means on, on, on being completely systemized. But I, I, and I've got friends in that mastermind group that I'm in that are at a way different level than what I am like 10x the level that I am. And the more that I talk with them, the more that I realize how systemized their business is, you know, and to be able to ask them, hey, how did you get there? How did you do that? You know, and to be able to get those answers of hey, let me give you the SOPs. This is what we did. These are the systems that we used.

So. Yeah. Is that the same mastermind where you first came across Systemology?

Yes, I, I, I'm pretty sure. Right. That this book was, was handed out at one of our intensives at, you know, and I probably should have inscribed it in there of where I got that. Right. But I'm gonna, I'm in a mastermind group called a Seven Figure Agency with Josh Nelson. So it's one of the things that I've been looking for, honestly for the past 20 years of being agency owner. Right.

Yeah.

And I came across it in 2020, I think sometime in 2020 when I came across Josh Nelson's book and I read it and where he talked about how to become a Seven figure agency and agencies are super hard to scale, is super hard to scale any service businesses, but agencies in particular are super hard to scale. And he had the, he had the roadmap for that. And so I read the book, started drinking the Kool Aid and joined the mastermind group. I've been a Mastermind group member for the past four plus years. I think, I think I just hit my, my, my birthday or anniversary, you know, with them.

And just an amazing group of, of individuals that I'm, I'm with there. We're going in less than a month to our next intensive in Miami. So.

Yeah, love it. Yeah, we, the Systemology message tends to resonate so well with agency owners and I think it's because, I mean, I used to run an agency. We had Melbourne SEO services that I ran for 13 years. So the whole idea of Systemology was birthed around that and the struggles and the frustrations of scaling agencies. I get it. And I've been there. Yeah. With no surprise that, yeah, our message sort of reaches the agency world, but might leave it there. Look, I'll pop a link through to your site roofingsites.com for people to check out. As I said, I just appreciate you hopping on and sharing some key insights and strategies on how you're leveraging AI. Hopefully it's inspired other business owners to do the same.

Thanks so much for having me on.