Why We're Building Our Own Software for Home Service Businesses with AI

More and more plumbing companies, HVAC companies—they're just vibe coding.

The mistake people are making is that they're building AI for their team members and themselves, whereas what you should be doing is building AI for AI.

You can build a website with AI.

You can go check out my vibe coded website, servicegabage.com. So this was built on Claude?

Yeah. This is crazy.

You don't need us anymore. You don't need a marketing agent. Just go do it yourself.

What can AI build today that like is surprising you? Yeah, I think the thing that's surprising me is. Okay, welcome back to Owned and Operated, a top 150 business and entrepreneurship podcast. On this show, we talk about how to grow your home service company, how to buy them, how to market, how to just do all the cool shit. I'm your host, John Wilson, and during the day I run a 4-location home service company out of Ohio. And for funsies, I talk to my friends on this podcast about home service. Today I'm rejoined on the show by my good friend Sam Preston, the CEO of Service Scalers. And today we're talking vibe coding. Welcome back to the show.

John, what is vibe coding?

Vibe coding is when you're coding on a vibe. I'm gonna go get on ChatGPT or Claude or Lovable, uh, maybe a year ago. And I'm just gonna put together a program that I have no business having the ability to code.

That's vibe coding. Mm-hmm. Mm-hmm. Gotta have that, right? Like that perfect Spotify playlist that just kind of gets you into trance. It's late at night.

Got your phone on in the dark. Totally. Just vibing. Yeah. Just going. Yeah. You're just vibing. But yeah, that's vibe coding. It became more and more of a thing. Obviously, like, as time has gone on and like AI has become more and more like out there, what's been really fun is like we've had, I don't know, like 5 or 10 events this year just like at our office. And consistently more and more like plumbing companies, HVAC companies are, they're just fucking vibe coding. Like they're just out there like making shit. And it is the funniest, it's just the funniest thing. I'm imagining someone from like, like a tech bro from San Francisco, like walking into this group of plumbers and like all of the plumbers are just like, here's the 4, here's the last 4 things we shipped. And it's, it's, I don't know, dude, it's, it's funny. It's a funny time to be in the trades.

It really is. Like websites used to be and still are, you know, relatively a difficult thing to do. But if you have just enough patience and a little bit of Claude, ChatGPT experience, you can go create something pretty amazing. Um, and you don't need to spend, uh, tens of thousands of dollars on a developer or an agency to go build you out a massive website. Yeah. Uh, you can just go do it.

Yeah, that is kind of wild. I mean, we, we did, uh, we worked with you guys a couple years ago to do our programmatic And like, is anybody building that big, like a 10,000-page site, like vibe-coded? Like, is that a thing? Or like, are we not there yet?

I haven't done 10,000.

Could Roto-Rooter go rebuild their website tomorrow vibe-coded? Like that, I guess, was the question. Like, so, so Roto-Rooter could do that?

Yeah, Roto-Rooter, if you're a bigger company right now, national brand, you absolutely should be trying to get as many pages out there as possible. And to try to expand as much as possible, cuz you absolutely can. And so, but no, the biggest site we've done so far was a 200-page site locally. So it wasn't for like this national brand.

Yeah.

Just like one, you know, small little company that wanted some coverage. And so, but most of the ones we're doing right now are probably between like 50 and 100 pages.

Hiring great leaders shouldn't be a guessing game. At the Military Veteran, TMV, they connect home service businesses with proven leaders, military veterans who know how to execute, adapt, and inspire. Their clients scale faster, hire smarter, and build stronger teams. One HVAC company even grew from $50 million to $3 billion in just 5 years with TMV's help. They move fast, handle the heavy lifting and deliver exceptional talent you won't find anywhere else. If you're ready to hire leaders who actually drive results, get a free 45-minute TMV executive coaching session with a successful home service entrepreneur. When you mention Owned and Operated, click the link below to get started. Yeah, over the past— so this has come up a lot. We've actually had a couple like group sessions on it and you said you had some internal company meetings on it, which honestly we do not do, but I really like the idea of. But we've been creating a ton of shit on the back of AI. And I'm curious, like, how you guys are thinking about, like, what are you vibe coding? I think we're supposed to be talking about websites specifically, but I just think in general, a lot of people— a lot of people, yeah, going off script. John told me I was allowed. You're good to go. So, I think a lot of people are, Just like wondering like, hey, what can I, what can I do here? What can I like code or put together? Um, so I'll give some examples from what we've done. Uh, but I also think like we can give a philosophy on how to think about it. Uh, so we rebuilt our fleet software. Um, and we got into this mindset and you actually said this off camera and it was something like, what was your, you have this build meeting. Can you wanna talk about what the build meeting is?

Yeah.

I wanna pick on one idea.

Yeah. Yeah. So on a weekly basis, we have a meeting called The Build, and it's broken into 3 different pillars. One is how much work can we automate? Like, how much can we just have AI do for us? Think like admin work. Think like small, like things that someone's just like literally dragging and dropping. We don't need that anymore. We can just literally have that. So our team is focused on strategy versus admin. So pillar number 2 is how do we never hire another human again? Obviously we will be hiring more people, but the idea is like you either are going to prove out that you can't do it with AI and you're going to hire somebody, or you're just going to do it with AI. And then pillar number 3 is how do we take our external software costs and bring them internal? We were just spending $1,500 a month on an out— a third-party dashboard, marketing dashboard. agency or software, and we've now just pulled that in-house. So it's $1,500 a month more in my pocket.

Yeah. So what we've been focusing on is that. So like those last 2, as we've like approached the software market, we're really just like, hey, does this thing provide— if it's more than, I don't know, let's say $5,000 a year, We consider building it ourselves. 5,000 of like subscription cost, which I was talking to, I was, so we're interviewing for CFOs right now. And what has been really interesting is we run this discipline, we called it, we call it a fuck-up list. So again, this is why I'm explicit on Apple, guys. I don't know, I don't know. But so we call it a fuck-up list. We're like, we're gonna go fuck up those costs. Like we're gonna go negotiate that software, vendor or whatever. And, but I found out there's actually a formal name for this, which is called a cost-out list. And I was like, oh, that's my fuck-up list, is what I said. I was like, no, no, no, no, no, you misnamed it. Yeah, you got it wrong. So apparently that's a formal thing. It's called a cost-out, which makes total sense. And what was really interesting is the guys I was interviewing are like, Like they're eligible for CFO. So like, you know, talented individuals working in much larger businesses, like one was like $900 million and one was like a PE firm with like billions of revenue. They're running a cost-out structure and at any given time they track, hey, here's like 50 things that we could save money on.

Yeah.

Maybe it's 50, maybe it's 100, maybe it's 20. And then they just prioritize it by the amount of money they can save. And what was all that to say? What was really interesting is in my mind, I have a, I have a mental list and it's probably like a million freaking dollars of shit. Like it's always so much money, it's kind of insane. Uh, but they track all the way down to $5,000 and that was almost a billion-dollar business and they're tracking cost-outs down to $5,000, which I thought that was pretty interesting. Um, so all that to say, if it's—

That's awesome.

If it's more than $5,000 a year to pay for a subscription, we are strongly considering building it ourselves. So the very first thing we did was a fleet management software to track repairs and mileage and VINs and like repair tickets. And because everything we looked for on the market, that was gonna cost us $20,000 for our size of a fleet. And $20,000 to like basically for a souped-up spreadsheet. Sounded insane.

Right, right.

Yeah.

Yeah, that's insane. And so how much, how much have you so, uh, like saved total with replacing that?

Like that? Well, we didn't have to buy software, so $20 grand.

Okay, okay.

Then what we started doing was we started looking for like, we piloted a couple softwares, and I think this is part of the thing that gets really powerful with it. is we tested a few softwares that had very niche needs in our purchasing department or in marketing. And when we couldn't find on market what we were looking for, we just went and made it.

That's cool.

And that has been really powerful. We made our own marketing dashboard to track attribution because we'd been doing it manually on Google Sheets for 2 years. So yeah, like, so you can ship a lot of different stuff.

I mean, as simple as like, I have like my own little dashboard that I built out because Stripe is terrible at telling you, projecting what you're actually going to make over the next couple months. And so I built my own little dashboard where it takes in all of our financials and then just projects like, hey, based on this, based on what you did last year, based on your current growth month, this is where you'll be in a couple months. Love that. Um, it's just something that Stripe, as great of a company it is, just does not do that for us.

I like the build meeting philosophy. I think we're probably going to add that into our business. So that was really good. So it's like, hey, what can we automate? How do we prevent employment creep? And how do we remove current items from our P&L? Uh, I like that a lot. So that honestly, that should probably like, we should just end the episode there because that's probably like the most valuable.

Yeah.

I think there's so much stuff, um, like you can start automating permits, you can start automating down payments, FP&A, like financial analysis, uh, like what you just said. Like you should be, I should be able to download transaction history and say, hey, give me a budget. So you should be able to reduce your FP&A time, which, like, FP&A software is like $3,000 a month. So what can AI build today that is surprising you?

Yeah, I think the thing that's surprising me is the mistake people are making is that they're building AI for their team members and themselves. And I think that's the biggest mistake. Whereas what you should be doing is building AI for AI, right? Like, why do you care In a lot of ways, how the sauce is made, as long as the sauce gets to you. And so the way we're thinking about it in so many ways where you would normally have a task done, sent to an employee to then like do the task, but like, how do we, how do we take that task? And it just like skips that part where it goes to doing the task and then just goes to approval.

Yeah.

So somebody's just reviewing it to make sure it's done correctly. Because, I mean, obviously the Claude does hallucinate. These engines will— my sister's actually with her company, it had basically been just creating random data in their CRM, just blasting out. And literally, like, 6 months later, they thought they had all this data and it was just a complete lie. So you do have to check it. You have to figure out the failsafes there. But I think the biggest mistake people are thinking of with AI is like, How do I get an employee involved in that versus like, how do I take it from like end to end done? And then if you want a human approval, get to that human approval, that's fine. But how much can we just like take off instead of displaying this for— instead of displaying the data and the content for human eyes, how do we display it for AI eyes?

I think it's been, it's been honestly, so I went down this experiment like maybe a month ago. Because we were— I was trying to figure out— we have a pain point inside our business, and the pain point is like our time to get a permit is too long and it slows down revenue. So we're like, okay, this is an acute pain point. There's a revenue outcome tied to this. So like we do have to get this right. We have to solve this problem because this is becoming a bottleneck. So we, um, so I get on ChatGPT and I, I try to go do the same thing. So I get on ChatGPT or Claude or one of them I'm like, hey, I want to build out this process. Like, how do I do it? And it was pretty interesting because the way I imagined it going was I want to record a screen of this process happening, and then I want to input that recording, or maybe like a transcription of the recording, into the LLM, and then the LLM spit out like Here's the steps, and then I take those steps and I turn that into an agent, I guess. And so that was sort of like how I was imagining solving it. And for some reason it didn't like that, and I don't remember why, like the LLM didn't like that. But I do think I agree with you, you can figure out just how to completely remove friction, uh, from the business. Like I think purchasing A big one is like data entry and accounting. Right now, we're trying to figure out data entry and accounting because we bought three businesses so far this year. We're about to buy our fourth, and like accounting just got very, very complicated, which is good. Like we're doing a good job. We're growing all that stuff. Whatever. We're figuring it out. But I do think like we could remove so much friction just by automating like job costing. You know, that feels like a very, like, A to B task.

Yeah, so my encouragement, at least as far as how, like, one of the things that we're doing when we build stuff, is obviously the problem with the accounting thing, or just like the numbers things that I was just talking about, is they, you know, Claude will hallucinate. So it will just agree with itself and give you numbers. Yeah, what happens when you give it bad data? When we're doing, like, websites, one of the things it wants to change, I don't know why, but it wants to change phone number. And so, you know, we built out like a little design for someone to go print out stuff for their GBP so they can get their location verified. And stupidly, I just didn't check the phone number. I was like, oh, everything else looks so good.

Yeah.

And, you know, we vibe coded, we had the design. I didn't have to get a designer involved, which was epic. And then the phone number was like the last, last 7 digits were off. So the first 3 were on. I was like, okay, cool, this is great. And I sent it over to him. He's like, this is great, but wrong phone number. I was like, damn, why? And so that happens a couple of things. So when you're doing accounting, the thing I would tell you to do is get both Codex and Claude Code in it. So if you don't know, Codex is ChatGPT's Claude Code version, and you can actually have it work together in the same terminal. So when I'm coding something that's big or inside of our new operating system, I basically have it both build out a plan and then Claude take over to combine those plans together and then go build it. And then both have to QA it before it can spin it out and push it to live situations. Obviously, we want to avoid situations where we take things down. We want to avoid security issues. We want to avoid so many different things. So having both of them play out works really well for us. And I think in something like an accounting for 4 different businesses, you would want both to check each other.

That is good. Um, all right, so let's get back to building websites. So like, you can build a website with AI?

Yeah, yeah. You don't need— you don't— you don't need us anymore. You don't need a marketing agent. Just go do it yourself.

What— what do you think like the big benefit is here? Because I would imagine like maybe SEO, but I— I almost feel like the cooler thing of like developing like websites with AI is you can just do faster landing pages. So like if I'm running Meta ads or if I'm running Google ads or whatever ad I'm running, I can just go like test a bunch of landing pages and iterate really fast. And that feels like the best use of this technology or like, what do you think?

Yeah, I know. That's, that's a great one, right? Just because offers are generally the thing that's going to help with, you know, Meta working correctly is by having multiple different offers. And so if you can test ad design creatives, as fast as you can test landing pages. Like, that's pretty cool. The other thing is like locations, like people, you know, historically when they've come to us or any other marketing agency and they're like, hey, we want to build out this website. Well, how many locations do you want? Well, I want, I want all the locations. Well, that's going to be $100 grand. Just, you know, that's expensive.

Yeah.

Okay, cool. Well, like, no, let's just do 7. Now it's like, well, what are the locations you want to be that you're at right now? And then where do you see yourself in 3 years? So that we can start building up those locations. What if we start driving leads from a location that you weren't like ready for because the market isn't as saturated in that one city, you know, 2 doors down? So I think that's a really cool aspect of it. As far as like content, you can create a lot of really good, unique content. One of the things I'm playing around with my, my own is like a personality to our content. And so like you can create like a, you know, the way you want to sound, the way you want to come across very uniquely. And I think that's really cool. Yeah, there's, there's a lot that you can do. Website. Website also, it's like you've, you've kind of been, you know, I remember, man, 10 years ago I built a website with an agency. This is before I owned my own agency. One of the reasons why I got into it was a $30,000 website. And man, it was just so bad. Like the design was just rough. It was a home— it was a commercial office supply business, like furniture business. So like desks and stuff like that. They made it like dark gothic black. I don't know what happened.

Oh yeah. And I was just stuck with it.

And if I wanted to change it, I had to spend more money. Like now it's like, okay, well, like if I don't like it, I just keep working with Claude Design. I just give it more examples to come up with for websites. And that's, I mean, quite frankly, that's what we're doing on the backend now. We don't sell, we used to sell on top of like WordPress or Webflow. Right now we don't do that anymore. We're just coding those designs exactly the way you would want and giving it to you to choose from which design direction you wanna go. And then we're building out the backend for it. So this is literally what we're doing. You could just do it yourself if you wanted to.

Most marketing agencies will show you clicks, impressions, and maybe even traffic, But none of that really matters if the phone's not ringing. And that's why we partner with Service Scalers. They are built specifically for home service companies, and they focus on one thing, which is driving real high-quality calls and book jobs. This is a no-brainer. They're offering a 60-day money-back guarantee on LSA management, Google Business Profile optimization, and website builds. If you don't get more visibility, more calls, and better leads, then you don't pay. If you want more booked jobs without the marketing headache, click the link below and book a free strategy call with Service Scalers. It's, yeah, it's interesting to see how fast, um, we did, yeah, we did this huge SEO project like 2 or 3 years ago and we're trying to figure out what to do now because when we did that project, we went from like, uh, a small website. Well, we went from like a Scorpion website cuz we had Scorpion. Uh, this was like 4 years ago.

Mm-hmm.

And we went from a Scorpion website to like, uh, programmatic, which, uh, like service area, I think, I don't know, we had like thousands of pages. It was crazy. And like that worked and that really propelled us for a few years. But now like the big gap that we have is we're on Webflow and Webflow is really good for programmatic and like massively deploying content. One of the downsides is you run into a limit on number of pages, and we hit that limit. And so we haven't been able to do any more. So then the big challenge is like, okay, well, what do we do now? Because we have Nashville, and we have 2 businesses in Fort Wayne, and we have Medina, and we have— so like, we have a lot going on, and we're going to do more. And we're planning on rebranding all to one singular brand. How do we do it? So, this is like an active discussion that Jesse and I are having of like, okay, well, hey, we're capped on Webflow. What do we do? Do we rebuild the website again? Do we do all this again? It's 10,000 pages or 8,000 pages or whatever the fuck it is.

Yeah.

Yeah. Yeah.

100%.

So, have you migrated—

We're rebuilding the site. Let's go.

Yeah. Yeah. Let's get a couple beers. So, have you migrated a site from Webflow or something too— like, how would you handle a site like our site? I don't even know how you'd do that. Like, it's so big. Because there's so many URLs, you don't want to lose any of them.

Uh, you don't want to lose any important ones, you know?

If you got a bunch of them there. Yeah, yeah, yeah. If it's getting traffic, yeah. I mean, out of like 10,000 pages, it's probably like 200 pages that like get real traffic.

Yeah, yeah, yeah. And so, I mean, the first thing you wanna do is you wanna crawl your site. You can use Screaming Frog, there's plenty of different softwares that you can go use, and then you wanna scrape together—

What's it called?

Screaming Frog.

Really? That's the name? Okay.

It's actually a nice little software. That's hilarious. But it'll crawl your site and kind of break it down to like, this is your, like, every page on your site that you've got.

Okay.

And from there, you want to build out your sitemap. So all the URLs that you have, all the— every single page, and figure out what that sitemap actually looks like. And then you want to probably run that through some kind of like SEO software to figure out how much traffic each page is getting. Some pages are going to get a lot, some are going to get a little. And then you just, the first thing you would do is make sure that there are no pages that you don't wanna lose out on, right? Like it would just be horrible if we did.

Yeah.

When you are migrating, the thing to think about to keep as much of that traffic as possible is keep the content as similar as possible, keep the H1 tag exactly the same, and then keep the URL the same. If you do that, those 3 things, you generally don't see a big dip in traffic. So Screaming Frog is also gonna pull in any, any pages that you are having trouble with. You can do a lot of cleanup, some redirects, lots of things that you can kind of get rid of. And then the next thing I'm gonna have you do is build out a design guide. So basically, you know, that really expensive creative work that you would go and get from a design company.

Yeah.

You can now work with Claude to actually build one of those. I've got one I can send you that gives you—

Here's the, like, websites that we find inspirational or whatever. Is that how that works?

No, more like, this is my colors, this is my typography, this is my messaging, this is who I want to be.

Okay. So like your brand guide.

You know?

Yeah.

Yeah. Make sure you swear, but make sure it's tasteful, you know, things like that. Yeah.

Don't say fuck. Only say shit. Got it.

Only, yes. So, uh, as much of that as possible, you want to get down because once you're building at scale, like you can—

what? Yeah. You like that? John just like, John just gave me like quite the look. He was like, really, dude?

Did he flip you off?

This is why we can't, this is why we can't have nice things. This is why we're explicit on Apple. Oh, good. Sorry. Sorry. Continue.

Um, no. Yeah. Yeah. So get that, that sitemap, get that brand guide. And then the next thing I think you should do is I should— I think you build out a custom coded site that you're going to have on GitHub. So we're moving away from WordPress and Webflow. And one of the biggest reasons why is because it's just useless code that's in between your code and a consumer, and it just slows down your website. And then I would start having, having the team build out all the pages, making sure that it's based off of— obviously you have a lot more locations now and you get yourself a pretty awesome website pretty quickly. I would 100% do that. There's so many different things you can do.

Yeah, that's really— that is interesting. I need to get back into Honestly, I haven't looked at it that much for like the last year. I mean, obviously, like Jesse's running it, but like I ran the like project three four years ago, and then these these days it just hasn't been top priority. But like it's becoming top priority because we're prepping to do a rebrand. We're prepping to like add all these new location pages. We're prepping to like redo all the GMBs. I think we have like or GBP's. I think we have like twenty five GBP's now. like across the platform and we're just like doing a big sort of re-smash. But website has become a bigger and bigger topic because we're literally out of location pages. So we have to figure it out. How do you think— I have not been paying attention at all to like what the larger agencies like Scorpion, like, are they still doing like 2 blog posts or like what are they doing?

Here. That's a great question. I honestly, I've been so heads down. I've not been paying attention. Yeah, yeah, yeah. To what they're doing. I hope not. But my—

yeah. Yeah.

Like 2 years ago, I was selling, like talking about this podcast going like, look, we're doing 10 articles a month. That's so cool. That's like table stakes now. Yeah, yeah.

What you'd imagine, but somehow they still get contracts. It's kind of honestly incredible. I think they just like have the franchise relationships or something. I don't, I don't know. I mean, I would imagine that they would be under stress. I would, I would have to imagine that they're under strain.

Yeah. I feel like the big agencies are really going to struggle because they can't move as quickly as we can. And they are going to be more nervous on using something like this. You know, kind of like Amazon had such a big problem with Claude. Because they brought it in and developers were just using it and they started having issues with it. So they have pulled back the type of usage. Now, the great thing about us is we're smaller, so we can test things really quickly, see if things work. I'm like in those rooms having those conversations on a weekly basis directly with the people that are seeing the results inside of the marketing. And so like, it's just a smaller team. And so it's scarier when you're Working with like such a large company and such a large team. So I don't, I wouldn't even know if they were using it at this point. Whereas like on our end, we're like, how do we get each one of our team members a chief of staff?

Yeah. Yeah, that makes sense. So what do you think like a $1 million company should be doing today with this? I've got some thoughts.

Yeah. I mean, $1 million company, 100%, you know, it obviously just depends on your time and where you're at, but like you should definitely have a website. Like even if it's just a one-page Go use one of the, uh, the cheap tools to go build that. Um, you can also— I mean, we, we have a pretty, uh, low-cost offer at this point for it. Um, but you can, you can go do this yourself, um, and see where you're driving traffic. I think I mentioned I've got kind of like a secret project on the side where I'm trying to build a, a home service business. I haven't mentioned what it is yet, but, uh, and, uh, it's been up since April and I've done my own—

One of us, one of us, one of us.

Uh, I threw up a website. And doing SEO and GBP. And I'm like, you know, 45 days in, I've got 5 leads so far. Um, and so pumped about that. And I probably couldn't take too much— too many more at this point. But like, you know, this is, you know, vibe coding at night, just a little goblin, uh, working, trying to get traffic through it to it.

What do you, uh— I, I think something that's interesting, it feels like I mean, it feels a lot like a decade ago. It's like a transformative moment where this is a moment that scrappy companies can beat legacy, which like I'm on both sides of. Like, we want to be the scrappy company, but I also like have to acknowledge that like, hey, like we might be legacy or we might be approaching legacy. Right? So like a scrappy company could meaningfully compete. And I think the smaller you start, like how do you get rid of some of the base function? Like some of the use cases I've seen people building purchasing softwares, I've seen a lot of like estimation softwares, like hey, easier flat rate type things. I'm seeing some like bill.com replacements. But I think the way you should be thinking about it is like, hey, what work can you take off yourself as a small business owner? What work can you take off your team completely? What can you automate? What can you get rid of? And then what costs can you cut? Because I do think, for better or worse, like customer acquisition is becoming more and more expensive. So if you want to scale, like you have to be able to scale your customer acquisition, like your marketing. So the more you can cut out of unnecessary costs, like fleet software that you can just go make in an hour, the better, because then the more dollars you can shove at customer acquisition. And I like the next question was, what would a $10 million company do today? And I think it's basically the same thing. You just get more power from it. I know for us, we're driving mid-20% EBITDA margins right now, and that's very high for the industry. And I think a big part of why we're having that success is because of how much we are deploying into automation and AI. I just think we are trying to like bite-size it. Like, hey, let's do this task, let's do this task, let's look at this position, let's look at this. Like we haven't done like a massive overhaul of really anything.

Yeah, so if you're a $1 million company, you should be trying to figure out how do you like not go take on a bunch of costs that you don't need to. Again, like I built out a little proposal design and told Claude what I wanted to come out with. And then when you're at the, you know, the prospect's house, I'm taking photos of it, I'm uploading it, I'm using my voice to dictate, hey, this is what I want to see on the proposal. And it's giving me basically a custom proposal for that person I can just send to them, you know, instead of spending $60 a month. That's what you're doing if you're a million-dollar business, you're a smaller business, right?

Yeah.

If you're a $10 million business, you're hiring somebody and their job is to either make you more money or cut costs by what they're able to implement via AI. And that's— it's an easy ROI at that point. It's like, figure out your biggest cost basis, all those— what'd you call it? Not the fuck list.

What's the thing? Oh, my cost outs. Yeah, yeah, yeah, yeah.

Cost outs. Yeah. Find all your cost outs and then have them just tackle that as quickly as possible and then go do the opposite. Like, where else are you not you know, tackling something because you've been limited. Uh, you know, where are— you know, are you slow in speed to lead? Okay, well, let's start automating that to get faster. Are you having issues with leads? Go, you know, figure out what— where you can deploy AI to get more. Um, that's— that to me is the name of the game.

Homeowners want your services, but they do need a better way to pay for it. And I'm hearing more and more contractors talking about Comfort Connect. Comfort Connect gives your team access to powerful payment options all in one place. It includes traditional financing, lease-to-own options, and their Premier Program, which bundles everything from equipment maintenance and covered repairs with ongoing support all wrapped into one predictable monthly payment for the customer. Contractors using the platform are seeing higher close rates, bigger average sales, and stronger long-term customer retention because homeowners stay connected to their brand long after installs. So stop treating financing like an afterthought and start using payment options your customers need as a part of your sales strategy. In, in the old days, a 10% net profit business was like good, and 15% was great. And I think what this does is this allows 20% to be good and 25% to be great. And I think that we're gonna start seeing You think you can get to 25? That's where we are.

No way.

Yeah, 25% EBITDA, bro.

Yeah, that's amazing.

Yeah, I know, it's ridiculous. Yeah, but, um, yeah, I think, uh, I, I think like it should change your expectations for like what's possible.

Yeah.

Sweet. Well, honestly, like this conversation was really good. This helped me a lot. I'm gonna talk to Jesse about Like our website? Because we were— yeah, we were having this conversation last week on like what do we do? Because we're trying to like figure out how to add in all these other locations. And it is like a— it's a thing. And we're a little, you know, obviously concerned about disrupting our current site.

Yeah. I mean, have Jesse reach out to me and him and I can jump on a call and just kind of talk him through how I would do it. I'm happy to Talk them through exactly what to do, 'cause, uh, it's, it's one of those things where you should do it.

Um, I think you'll see a lot of people do it. Well, I think we have to. I, I think, like, I think, like, we're probably not optimized for AI the way we need to be for, like, AI search, which I think I'm noticing personally that I'm searching more on AI. So, like, I'm sure other people are too. And then they're also starting to run contractor ads on ChatGPT now. So, like, we're trying to figure that out. So, like, yeah, we're trying to— and we have to add these new locations. So it's kinda like one big project of, like, gotta put this all together. Yeah. Yeah. Yep. Yeah. Sweet, man. All right. If people wanna learn more, what should they do? Google vibe coding and then check out— yeah. What were you gonna say?

I was gonna say you can go check out my vibe coded, uh, website, serverscalers.com. Uh, oh yeah, yeah, yeah, yeah, yeah, yeah.

Did you put on there like, we vibe coded this?

I don't know. We should though.

That would be funny. Oh, dude, this looks beautiful.

The greatest part about it is like the person that did this had no experience in Claude before I had them do it. And I was like, okay, let me just walk you through it.

That's actually crazy. They did this.

That's actually crazy. Right?

That is really interesting. So, this was built on Claude?

Yeah. And probably a little bit of ChatGPT and We're pushing it out to, again, GitHub as your repo and then hosting it.

This is crazy.

Yeah. And honestly, like, it's kind of like it was pushed live last week and we spent like 2 weeks on it.

Yeah.

Like, it needs a lot more, but we just were like tired of the website that we had and like, let's just get something up.

Wow. Okay. Yeah, this is amazing. This is amazing. It's really cool. All right. Yeah, check out servicescalers.com. That is an example of a vibe-coded website. That's fucking awesome. Cool. Well, if you like what you heard, make sure you give us 5 stars, like us, sub, and comment below whatever you're shipping. I'm really curious what other people are working on inside their business. Thanks everybody.

We'll see ya.