← Owned and Operated - A Plumbing, Electrical, and HVAC Business Growth Podcast
Owned and Operated - A Plumbing, Electrical, and HVAC Business Growth Podcast
AI Is Creating a MASSIVE Gap Between Contractors
00:48:07
Loading summary
Transcript123 lines
- [00:00]
A
You are fighting with two hands tied behind your back. Go get the technology. Every single person is actively building like an AI tool inside their business.
- [00:09]
B
It's mind boggling how much it's progressed, really in just the last three months.
- [00:14]
A
Oh, yeah. Like, I see people trying to code CRMs right now. The first one we're actually like working on and prepping to ship is like a cost comparison tool. Like, are you driving best in class pricing?
- [00:24]
B
You know, we're getting to the point where that is possible. The tricky part is always like the details and the nuances.
- [00:30]
A
I mean, I think the potential advant is you could probably drive like 20% material savings. What are you seeing for like these large consolidators? What's the pace of technology adoption?
- [00:40]
B
Yeah, I would say that.
- [00:49]
A
Welcome back to owned and operated. I am your host, John Wilson. During the day, I run a plumbing H VAC and electric business and out of Ohio, Indiana, and soon one more state, which is pretty awesome. And for fun, I talked to my friends about how to grow home service companies and what they're doing to do it. Today I have my good friend Tyson Chenon, the founder of Avoca. And we've been partners with Avoca on the show, but also inside our business for two years now. And they help power our call center, dispatch, speed, delete, and a bunch of other capabilities. So today we're diving into AI inside your business. Tyson, welcome to the show.
- [01:28]
B
Pleasure to be here, John. Good to see you again. And happy belated St. Patrick's Day.
- [01:32]
A
Yeah, yeah, yeah. Were you in New York for anything interesting? We had a few people in Chicago and they have the Green river thing.
- [01:39]
B
Yeah, there was a parade. I was actually visiting a customer, one of our customers in midtown. And there's a giant parade that, you know, delayed. I luckily took the subway, but one of my colleagues got delayed almost an hour by this massive parade.
- [02:01]
A
That's funny.
- [02:02]
B
You're Irish.
- [02:02]
A
Oh, yeah. Irish as hell, baby.
- [02:05]
B
I thought you were Irish.
- [02:07]
A
Well, I'm just very pale. Yeah. Today will be. Today will be fun. We're dive. AI is sort of like there's AI is obviously the big conversation.
- [02:16]
B
So, like, it's so crazy over the last couple.
- [02:20]
A
Yeah, it's crazy. And I'm not even like in the trades, but, but you know, like on the macro, you've got like AI, there's job risk. Like Meta is doing their 20% thing. Like, there's like a lot going on externally and inside the trades. I think it's kind of funny how much like how people are deploying AI inside the business. It's hilarious. So we had this event at our office two weeks ago and we had a dinner and the average contractor there was like maybe $20 million or something like that and every single person in the room, it was 20 operators, every single person in the room was actively building something or actively deploying something in from like an AI tool inside their business. And it was just, it was like you sit there and you're like, I'm talking to people running plumbing companies, like running roofing companies and like everyone is shipping something inside their business to help automate some process or do something. It, it felt like surreal. It feel if.
- [03:32]
B
What was. I, I mean I. In my podcast I just started called Trade Secrets from the Ground up, where I interview kind of founders and operators, home service businesses, learn about their stories, their trials and tribulations. But recently my recent one, there's a guy, Mark from, from Golden Rule and he was telling me about he. He's been acquiring a lot of smaller businesses, doing tuck ins and the process used to take him, you know, like a week to find all the companies that were interesting and then reach out to them in a personalized way to gauge interest. And he Claude coded it up in like an hour and it just runs now he understand the workflow in his mind. He just had to write it in cloud code. It integrates with Gmail, it integrates with Google Maps and it can research. It's mind boggling how much it's progressed really in just the last three months.
- [04:32]
A
Oh yeah. Well as they've dropped more of this autonomous stuff like yeah, Claude code is it Perplexity Computer? One of them that's openclaw. They have the openclaw computer where like controls your computer and it is wild. Like yeah, what you can ship. So like yeah, that's an awesome one. There's like a BD like business development component. I'm seeing stuff as simple as hey, I'm launching something to like help me do quotes, like I need to help estimate. So here's my labor, here's my material. I like. I see people trying to code CRMs right now which like, you know, I don't know if that's a good idea or a bad idea but like they're trying. Have you ever bought enterprise level software and realized that managing it really just became a full time job? Well, that's pretty much exactly why my restoration business switched over to Field Pulse. We were tired of software that promised efficiency, but came with endless training sessions, onboarding and frustrated texts and using it began to feel like it was a job all on its own. Fieldpulse was built for owner operators who don't want more dashboards. They need scheduling, invoicing and job tracking. And they need it all to live in one place without the chaos or the learning curve. FieldPulse is simple to roll out. It's easy for your team to actually use, and it was 75% cheaper than the other titans of software we were using right now. If you book a demo with FieldPulse, you'll get an exclusive partner offer. It's 20% off an eligible annual subscript and 50 off premium support for your first year. If your software upgrade has turned into a time suck, it might be time to make the move. Book a demo with Field Pulse and see if it's a better fit.
- [06:17]
B
I can't fault them for trying, but hey, maybe in six months. It's, you know, who knows? It's. It's pretty insane. I mean some of the top. I was just having this convo with my co founder the other day and, and we're talking about this at all hands but it was, it was like even like six months before the difference between someone that was good at using AI tools and could kind of use it was like 50%, maybe 30% productivity difference. Now it almost feels like it's 5x. You know, there's an anthropic post recently about how a marketing team.
- [06:52]
A
Oh, the one. Right.
- [06:53]
B
All their. With one person and we have a guy at our company and he's this is this. I just realized this two days ago and it completely blew my mind. He's a guy on the product team and instead of. He doesn't even write code anymore. He doesn't even use an ide. What he has set up is this environment where he's controlling eight AI software engineer agents for that write code two that you know, kind of check and monitor and then two that kind of relay the information and it's kind of like an AI scrum master. And you know, every once in a while they'll come to him when a decision needs to be made or a judgment call needs to be made. But more or less he's just able to develop these products that I mean would probably have taken a team like three weeks in. In a day just by interacting and iterating rapidly with his team of A8AI software agents.
- [08:02]
A
That is, yeah, that is interesting. The, the, you know, we, where we're at with AI is like I feel like we were probably Considered an early adopter. Like we started pushing pretty early. It's like two or three years ago or whatever.
- [08:18]
B
Yeah.
- [08:20]
A
Now where we've sort of stopped is we haven't gotten over the autonomous, like we haven't done that yet. Right. Like we haven't got gotten over that autonomous hump. I think we have a couple, I, I have a few ideas on like how to do it. Like one would be purchasing, I think could be kind of interesting where like how much of that process can we make autonomous? I think another one that would be cost comparisons would be another like job costing basically. Like can we, you know, shave off some labor from that? I'm thinking some of our work inside our accounts payable department because like accounting is like the big nut that not many people have cracked yet with, with AI because I think they're afraid of like the data manipulation.
- [09:10]
B
Our AES would absolutely love if we brought in an AI accounts receivable agent because people and a lot of, a lot of contractors are not good about updating credit cards and stuff like that.
- [09:25]
A
That's fine. Yeah. An AR one would be a good one.
- [09:29]
B
Yeah.
- [09:30]
A
And I want to do something with marketing. I think there's some alpha to be gained and obviously the anthropic thing is a great example. You guys are a great example. Where I struggle is like budget control because like what I imagine happens is like some, you know, freaking AI bot spends a hundred thousand dollars in an hour on meta ads and we just didn't even know it happened. But I, I do think there's like a real advantage to like what can we, how close to anthropic can we get?
- [10:02]
B
It's fascinating. I mean we are, are our roadmap. A little bit of kind of insider information there is that we are looking into the building AI marketing agents more end to end. We already actually have success with some, it's not specific marketing and you guys are kind of using this, but it's an autonomous way to think about capacity management and filling your board up based on weather patterns.
- [10:33]
A
Okay. Yeah, yeah.
- [10:34]
B
So let's say like, you know, that it's going to be, you know, pretty moderate temperatures for the next three days. You can use, we can use that to then anticipate what the demand levels will be and then backfill how many jobs need to be made and then we understand how many texts and calls we need to send out to your existing customer base in order to, you know, create those many maintenances or jobs. And we, and we just kind of do that in a Continuous way. And so that's more so leveraging, you know, members, existing customer base, any pre historic customers in order to keep your board full. And so that's kind of an agentic workflow that, that's been working really, really well for some of our larger customers recently.
- [11:25]
A
Yeah, yeah, that's interesting. I think I heard someone the other day talking about like LSA management or GBP management, which would be kind of interesting. Yeah, I think there's something to do there and I don't know what it is but like we're trying to work on it. The first one we're actually like, the first one we're actually like working on and prepping to ship is like a cost comparison tool. We've tried to find these things out of market. But what, what we and I have like five friends that have built this now, which is kind of funny. But every invoice that comes in, compare it against another invoice and match the sku. So like did a copper fitting from this supplier cost the same more or less than this supplier? And like are you driving best in class pricing?
- [12:10]
B
That's so funny. I remember his name, Eric. I think it was Eric Maxson from a PHCC conference that I went to two years ago, had that idea and we, we were almost thinking about building a like a workflow for him at that point. Maybe we should reconsider it. But his perspective was like for any RFP for parts, like he wanted an agentic process whereby we would effectively conduct the RFP for him. Like say here's what we need for a project, here's the parts. Reach out to email 10 suppliers or call, get their quotes, see who has the best quotes, narrow down to three and then have them like bake off against each other and just run that entire thing agentically. So yeah, seems like, you know, we're getting to the point where that is possible. The tricky part is always like the kind of the details and the nuances so that you know, you don't, you don't make weird like mistakes and you're actually and also able to like connect all the systems together.
- [13:23]
A
Well, the challenge, the challenge we found is the SKU numbers or like the written description. So like the AI has to be able to pick up the context that hey, this is a, they're saying this is SKU number 1234 and these guys are saying a skew number 5, 6, 7, 8. It's the same thing. And so that we're finding is the hard part is like how do you connect different Descriptions, different SKU numbers and use context, but.
- [13:53]
B
Exactly. And two things that sound similar might also be slightly different.
- [13:57]
A
I mean, I think the potential advantage is you could probably drive like 20% material savings. I think it can be very real cost savings.
- [14:07]
B
Rfp, like, no, I feel like people don't really think about it that much. But I had a case in a BCG with this company called Restaurant Depot. They're a giant supplier. They're like the Costco for restaurants is the description. But we ran some RFPs with their suppliers because they haven't changed them for three years and easily 20, 30% or more. Just by running an RFP.
- [14:36]
A
Yeah, yeah. So we run ours regularly. What we probably should do is maybe even the simplified version of that is like when was last time we ran an RFP against this SKU bucket, water heaters, faucets, furnace, like whatever it is. Because we're probably on a quarterly basis. But some of them are extremely time consuming. We just ran a five month process for our vmi and I mean it took a long time and then we ultimately ended up switching one of our VMI vendors. But it was a real process. But the savings are probably gonna be like 10 or 15%. Like it's meaningful.
- [15:16]
B
Yeah.
- [15:17]
A
And you know that on millions of dollars is a lot of money. It's a lot of money. Yeah.
- [15:21]
B
Yeah. But yeah. How's. Yeah. Business has been booming. How.
- [15:27]
A
Yeah. So what we're doing this year and it's been kind of fun, we're bringing on these small tuck in acquisitions and I haven't started using it for deal sourcing like you said. Mark from Golden Rule was doing like I like that idea a lot. That's funny. So what we're doing is like we're, we're going into these small businesses, we're trying to find companies, two to five million dollars of revenue. And what's really interesting, you, you take, you forget the. You. What's the right way to put this? Like, we've invested so much into our infrastructure and our tech stack and like our internal capabilities. You forget how much, how far you are until you go look like take over a company that has not done any of that. And you're like, yeah. You're like, holy shit. So that has been kind of fun because we're walking into these businesses and it's like, oh my God. You know, we don't have, we don't have 24. 7 phone answering. Like what are we doing? We don't have a speed to lead, so I can't Even use like most of our lead partners because they would just, it would go nowhere. Dispatching is still very manually. Like no one's touching dispatch yet. There's no like, you know, SMS campaign to just schedule appointments.
- [16:50]
B
So really utilizing their previous customer base.
- [16:53]
A
Yeah, it's, it's, it's really interesting to. So we did two acquisitions so far this year and we're prepping for our third one right now. We close in like a week. And it's you. It's wild to walk into these businesses and see the amount of opportunity that you have. And that's like kind of like the low hanging stuff to like hey, well we can just add 247 call center and like your board will fill. We can add speed to lead and I can use more leads.
- [17:22]
B
By just being 247 on Google you're gonna get an extra, you know, however many leads a day.
- [17:28]
A
Yeah, yeah. So the, the company we took it over, we took a small one over. It was a million and a half of revenue in January 1st and in their first month like a million and a half of revenue is like 110, 120amonth. In their first month we did 240. So doubled was wild. And it was like we turned on more leads. We responded to them faster. The, the call 247 call center took a minute to launch. Actually I got to find out when that launched but. Because some of it takes time. But it is funny, you walk into these small businesses, you're like you are fighting with two hands tied behind your back for, for basically no, no real reason at all. Like you can just go get the technology to do it. Yeah.
- [18:18]
B
Yeah. I mean it's really a great. I mean for P especially I think that's something that is, is just like a very easy, easy win. I mean we, we've launched now throughout all of of Sila like the entire country.
- [18:35]
A
That's crazy.
- [18:35]
B
And they have a requirement now. Every new acquisition they make, they'll just have us onboard them in one of the first month, all that stuff. So let's get that taken care of. And yeah, I mean for you I imagine it's amazing. You're getting to buy these smaller companies at low multiples and then tuck that into the main business and also double. It's just like stacking wins on top of each other.
- [19:06]
A
Yeah. I mean it really has been we so and like we're still figuring out the. You know we've done a ton of acquisitions over the years but like figuring out how to centralize is its own journey. Because when we've done acquisitions over the years, we didn't really centralize, we just sort of combined. But these new acquisitions are in new markets and we're running them as new locations. So centralizing is, is a whole topic now. So, hey, are we centralizing call center, dispatch, install coordination? Are we centralizing Tech stack? Are we centralizing into one service titan or multiple? And just like trying to work through all the nuance of that. And like, is this company TCPA approved? Because that's like, I mean, that's a thing. That's a big thing right now.
- [19:48]
B
Yeah.
- [19:50]
A
But yeah, just trying to like figure out our center. So we, how we started it was, hey, we're going to run these things a little bit separate. And after a month it's, it was like, hey, we actually have to centralize this. This. If we can combine Call center, if we can combine Tech stack, if we can do all of this in one thing, this is better. So the sila thing makes sense to me. They're all in separate instances. Right. Of service titan. Okay. So it's not like one giant instance. Okay.
- [20:22]
B
Multi tenant. Yeah, exactly. They got multi tenant architecture going.
- [20:26]
A
Yeah.
- [20:28]
B
But it's, yeah, it's. It's really fascinating. I mean, how do you determine when you do these tuck ins, do they all go under, you know, the Wilson brand? Or like, how do you, how do you determine like what goes under versus like stays its independent brand?
- [20:46]
A
Yeah, I, if it's close historically, we've always rebranded just a Wilson. We're trying to decide. We're going through that exact same question right now. So we, we closed on one in Indiana and we're about to do one in another state. And both of those, it's a little bit more of a question mark. Right. They don't know the Wilson name. That doesn't mean anything to them. We have no brand equity there. These brands do have some brand equity, but they've been under invested in. So I don't know the answer. I mean like normal pe, they tend to not rebrand because there's a lot of risk inside that rebrand. So the way we've thought about it is how it is. I think there's some efficiency to be gained on the rebrand building. One brand is probably more like valuable as a business over time, I think, I think there's nuance. So one of them will probably rebrand the other one. We won't. The larger one because it's like 7 or 8 million. And I feel like there's probably some Brand equity there. The other one's 3 million. And I think we could rebrand that pretty safely, but I think the answer is it depends. Horizon seems to. I mean, they rebrand all of them. Are any of those launched or are they all acquired or kind of a mix?
- [22:14]
B
I don't know. Actually, we are. Yeah. I think we're. Yeah, we know those guys. We're. We're not like live with any of them.
- [22:27]
A
Well, I mean, maybe you talked about Golden Rule a second ago. Do you know, like, is he multi location?
- [22:35]
B
Yes.
- [22:36]
A
And is he rebranding up to Golden Rule or is it just sort of like.
- [22:41]
B
I think that there. I believe so. There's also. And there's. He has some other brands too, like roofing, but I believe the. The plumbing H Vac is. Is Gordon Roland. Yeah, I think he's just like built a pretty kind of unique brand that, that he's. Yeah, exactly.
- [23:02]
A
Trying to add on to.
- [23:03]
B
But yeah, with Sila, it's also. It's very. It's very situational. Like, sometimes he'll roll them into the. The Sila main brand, like have Sila in it. Like Sila something. Sometimes like Jackson Comfort will be. Remain Jackson Comfort.
- [23:21]
A
Yeah, yeah, yeah, that makes sense. Yeah. I feel like maybe it's size, maybe it's geography. I'm not sure what. I'm not sure what it is, but it is interesting how much like as you something I think is kind of interesting. I was talking with a friend of mine about deploying A.I. i don't know, yesterday. And he's. He's a small business. He's like a $3 million plumbing H vac contractor. And this dude shipping like, like we talked about earlier, this dude is just like fucking every tool that he can. Yeah, he's just like, you know, vibe coding. Yeah, it's hilarious. So, like living off of Red Bulls and so like, you have that like, contractor. That's the small contractor, that's just like pumping AI tools into their business faster than they can move. And then you have like the, the PE. The large. Large. It doesn't even have to be PE, but just like the large contracting business. Maybe it's 100 plus million. I don't know what we want to consider large here, but. And that seems to be a much slower process, which obviously makes sense. There's more people, there's more jobs, there's more layers, there's more. All that stuff. But what I, what I kind of find interesting about that is, is the level of what could be gained is so different so, like, another example I have, I have my friend Sam over at Service Killers. He's shipping shit like crazy. And like, one of the conversations we had is like, scorpions. Not like no one could tell me that score. Like the CEO of Scorpions, like, vibe coding his fucking way through, like, whatever they want to build for Scorpion. And, but like, if Scorpion did, they'd probably save $20 million. So I, I, I don't. How are you, like, what are you seeing for like these large consolidators? Like, what's the pace of technology adoption? Are they getting what they were after, like this cost savings or are they not moving fast enough? Because my perception is the small guy is going crazy and they're getting good benefits, but the big guy could save like 5 or $10 million by deploying the right stuff. Customers today expect pricing online. And whether we like it or not, homeowners want transparency before they ever even pick up the phone. So we decided to lean into that. And that's why we use Contractor Commerce inside Wilson. It costs us about $1,600 a month. And we run it across H Vac Plumbing and Electrical. We use it for things like generator install estimates, water heater estimates, full H Vac replacement estimates, and it's producing real revenue. There are months where we've driven 20,000, 30,000 and even 40,000 in book sales directly through the platform. We've had double digit jobs booked in strong months straight from our website. What this really does is it turns our website into a guided sales process. Customers can see their options, they can see pricing ranges, and when they contact us, they're way more educated. If you're trying to modernize your sales process and capture demand that would otherwise bounce, it's worth looking at. Go check out Contractor Commerce and book a demo at the link below.
- [26:29]
B
Absolutely. I mean, what we're seeing, we do work with probably now 10, you know, billion plus revenue home service businesses.
- [26:42]
A
That's, that's like such a ridiculous statement. Just, just off like that is. That's crazy. Like, that didn't exist ten years ago, but. All right, keep going.
- [26:51]
B
Yeah, exactly. I mean, I, yeah. Mark said when he first joined nextstar, the big cat in the room was Horizon at the time, which was like, like 40 million or something.
- [27:05]
A
Oh, 40 would have been gigantic.
- [27:07]
B
40 was exactly. That was the elephant in the room. A $40 million company.
- [27:12]
A
Yeah, yeah. Like I'm 40 this year and I still feel like an infant. Like. Yeah.
- [27:20]
B
But yeah, I would say that, you know, to kind of your, your kind of the two parts of it 100. I do think that companies that have the, you know, quote unquote, like AI native like way of, of being built, even home service companies, you can just ingratiate that into the DNA. Right. Instead of having this huge call center or this, you can have one person that is, you know, kind of an agent whisperer or like an agent manager. That's really good and just like being able to do a lot of things. And if you start it with that DNA in mind, it's a lot easier to grow without, you know, grow efficiently and. But, but exactly like these larger companies, I would say there's. It's usually fall in two camps. Some that are very AI, like, like forward and tech forward. Generally those folks, they, they're down to go fast and they have alignment. Like all the executives are aligned, the board is aligned. Like, hey, this is. AI is an existential threat. You have to, we have to incorporate into the business quickly. We typically work with more folks like that. And if everyone is aligned and kind of rowing in the same direction, it is very tough. There's still a lot of change management. You have to fundamentally change a lot of operational processes in the business. But you can go very quickly. Exactly. We've seen case studies, tens of millions of dollars in terms of revenue uplift and also like, you know, cost savings as well. But then there are folks that are a little bit more traditional, careful. And so yeah, with those folks, they type, they typically like to just, you know, go a little bit slower and just like kind of dip their toe into it a little bit like. And to, you know, some degree it's, it's fair as well. I think the worst case you. You can be is someone that doesn't have like the processes in place. But. And you're really gung ho about AI, then you're just gonna fuck yourself.
- [29:37]
A
Oh yeah, yeah.
- [29:38]
B
Like a lot of businesses that don't use AI and are still like extremely successful, a ton of them don't use any type of capacity planning. Like, and just like, you know, stick to the, you know, the old ways and do things really well. Like those businesses can be very successful too.
- [29:58]
A
One of my, one of my favorite examples of that, which I think is like is. Is funny. Are you working with Four Seasons?
- [30:06]
B
Yes.
- [30:07]
A
So they, I think they're like a funny. Obviously they built an incredible business and they've like, they've led the charge like for the industry. Like it's an incredible business.
- [30:18]
B
Dave is also, Dave is also just an incredible person.
- [30:22]
A
He's a Great guy. I've only talked to him once or twice but like he's, he's awesome.
- [30:25]
B
No, you know, he's a Porsche race car. Like a Porsche race driver.
- [30:30]
A
Oh yeah, Dave. Like yeah, pop off dude. That's hilarious. But yeah, it's an awesome business. But I think something that was really interesting about them is I think they're adapting now. Obviously like they're using AI, but like two years ago or three years ago whenever Cortech got involved, like I don't think they were using Service Titan yet. They were using like a legacy CRM
- [30:53]
B
that they, they're still not using service.
- [30:55]
A
Okay. So yeah, so they were, they were using a legacy CRM that they self designed in like 1990 or early 2000s or something like that. And I think so. I mean honestly that's kind of interesting that they're using AI when still using that tool. Like what, how? I mean I guess I don't need to know how that works. But that, that is interesting.
- [31:18]
B
Yeah, it's kind of interesting. Like the way I see it is I use this analogy a few times. It's like China actually a lot of countries, they never use credit cards but now they've kind of just skipped forward or like for example they didn't use Apple Pay or something infrastructure and it's like they skipped that entire generation and now they've just kind of skipped into the future. It's kind of like a similar thing. Funny enough, if you never got onto some of the modern tech stack of home services, some folks we're seeing is they're going from their own self made or on Prem CRMs right into the future of agents. And that weirdly enough, because they definitely know they need to switch off the legacy platform is almost sometimes a smoother transition because they've essentially gotten to skip over an entire kind of tech generation.
- [32:22]
A
That is wild. I'm trying to imagine. And this is because like when I think about my dispatch board, I think about my service lane dispatch board. So how much deployment can happen inside someone that's like on a homemade CRM? Because this is becoming a bigger conversation now. Like more and more companies are actually like I made a joke about it earlier in the show, but more and more people are attempting to save 100 grand a year and not do a conventional Service Titan route. So how does that work? What does deployment look like?
- [33:01]
B
Yeah, so with agents again for us, and I'm sure other folks too, we're pretty CRM agnostic. Like you can use a CRM but at the End of the day, all we're using to run the agents are APIs into the CRM. And if you think about it, it's just like the CRM is a UI on top of a database. And if we're using APIs, we're obviously bypassing the UI layer and directly just talking to the database layer. So really any CRM that has a database is, is applicable. And even some people that like just have a database like we could also you know, work with technically. So I would say there's large companies
- [33:48]
A
are you seeing that are just database.
- [33:52]
B
Not many by and large. You know, most folks that we work with are on a service Titan or Salesforce or an i360, you know, accurate. You know, there's a lot of these different ones.
- [34:08]
A
That's interesting.
- [34:09]
B
But yeah, the beauty of agents is that they can work as long as there's API keys or some way to kind of speak to an underlying data source, the agents can work on top of it.
- [34:22]
A
This came up as a topic on the show I think about a year and a half ago. And if we think about like what could AI do? I don't, I think this was, I don't even think this was me. It was, Chris Hoffman said it and he was talking about the, the dream scenario is one day could his company run and it's 150 or 60 million dollars or whatever it is, 4 or 500 tax. Could it run off of Jobber or Housecall Pro and could they reduce their OPEX by half a million dollars a year? If it, if we're like basically just referring to a database because the, the big problem that you have to solve, like you don't even need reporting because you could have an AI bot like you could pull the data out and manipulate in whatever way you want.
- [35:12]
B
Well, I mean even for us, like we don't even, you know, for. We have, we use HubSpot and HubSpot reporting is notoriously ass like just completely trash. And so we have just started to create views. Funny enough. Like a lot of times we don't even use decks anymore. Our, our like VP of sales has gotten so good and you know, soon I think everyone will, we're spinning up like super, super nice decks completely on Claude. So to your point exactly like these, all these things can be spun up now with Claude.
- [35:49]
A
Claude.
- [35:50]
B
It's interesting. I mean there's still a lot of value to a nice CRM in, in a lot of these like you know, accounting and, and all these workflows. Yeah, it's still important to have one.
- [36:02]
A
But, but like at what price? I think is the. I think is the. What price? I think is the question.
- [36:06]
B
A lot of people, a lot of people do feel, feel that I think the. Yeah, like, like as the value is. Is. Is kind of going more and more to the agent and AI worker level. Yeah, it's a question a lot of people are asking.
- [36:26]
A
Yeah. Chris's thoughts at the moment, I think were really the only problems I have to solve are what's my technician interface, what are they entering invoice and price booking and payments. And those are the three core problems. And then you just have to determine how much is that worth paying? Because it's a lot.
- [36:48]
B
I mean what would happen to you if you, if your business, let's say like in three months had to move to, I don't know, Jobber, since you mentioned that one, like what. How would that impact things for you?
- [37:01]
A
I mean it'd be probably a shit show. But I think what we would have to figure out is exactly this problem where like hey, we do rely on reporting, we rely on payroll features, we rely on accounting features. We would have to, we would have to like deploy AI. Yeah, we would have to rebuild those because housecall Pro is not made to run us. And so I think we'd have to figure that out. But I also think someone's going to figure it out. Like, I think that like the cost savings is worth figuring it out.
- [37:36]
B
But you can imagine a person starting out with like a 1 million, 2 million. That's a little bit tech savvy. They could literally, if they started building their business on top of Snowflake and then just, you know, connected Snowflake to Claude code and built a set of dash like all the dashboards they need in ui, you know, that could be. And then they just scaled it like that. That's interesting.
- [38:03]
A
Yeah, I think. How far away do you think like. Well, I guess on the PE side, the one thing that I think keeps big CRMs having market share aside from long term contracts is who's going to buy the business. And if Apex is on Service Titan, then they're going to want you to be on Service Titan. And I think that that's the. It's almost like the. Because it's the legacy infrastructure for some time until somebody giant does something different. It. I mean maybe, I mean Salesforce is probably just the same price as Service Titan. It's probably like.
- [38:45]
B
Yeah, exactly. There's some folks that have, I know are, you know, thinking about it but with all the implementation, it's probably, I think, a little cheaper. But with the implementation costs of having to kind of retrofit Salesforce for a service business, it's. It'd be like the same or more expensive.
- [39:03]
A
Yeah, yeah. And I, I don't suspect. I mean, I guess you tell me, like, sounds like some people are like talking to you about this. Is Big P talking about this or is this just the contractor that's like looking at this line item?
- [39:17]
B
There's a few, there's some big P's. Obviously I won't name any names like they're thinking about it for sure. I mean, obviously we all know that there's one Big P that's not on Service Titan. But yeah, there's been chatter. There certainly has been. But I do think, I don't see, I don't see, you know, in the next, you know, three, four years, like a, you know, of departure from service titan from the top 20 PS.
- [39:47]
A
Yeah. Too much. It's like too much switching costs, too much switching risk. Yeah. But it is interesting because I think in the next couple, in those three to four years that these companies won't move. They're going to have to figure out how to integrate people that custom built their own fucking CRMs, because people are starting to do it like it is. It's funny. There's a, there's a guy in Pennsylvania and He's got a $5 million shop and he custom built his own CRM. He vibe coded it with Claude, like two weeks ago he put it on Twitter and I'm like, all right, dude, let's go.
- [40:22]
B
Yeah, it's, it's, it's certainly, certainly crazy. And, and who knows what's going to be possible, you know, six months, one year from now. Yeah, that's gonna get.
- [40:31]
A
That's funny, man. So what else is coming up? What, what else is coming up for you guys? What's next for. For Tyson and the Evoke team? When I first started growing Wilson, I was obsessed with the numbers, books, podcasts, and checking the bank account pretty much nonstop. But as the business grew, I had less and less time to obsess and I lost clarity. That's exactly what CFO made. Easy solves. Tyler works directly with home service owners doing a few million a year and helps you understand your numbers in a way that actually drives decisions, things like pricing, hiring, equipment purchases, all of it. I know what you're thinking. This is just another consultant. But Tyler is different. He built and exited a $25 million home service business in himself. So you know that this isn't just fluff. His clients routinely improve profit by 5 to 10% just by tightening up their pricing, labor and overhead and finally getting that oh so juicy control over cash flow. If you've got a bookkeeper and a CPA but no real financial partner, this is what you're missing. Go book a free 20 minute strategy call@cfomadeasy.com owned and get the financial clarity that you've been craving all along. And do me a favor, let Tyler know that Oceansha.
- [41:41]
B
Yeah, well we just did a pretty big feature launch a new product called AVO because it's kind of like the Jarvis for it's kind of like the Jarvis for being able to control your business in one interface. So John, imagine being able to. We got to show you and Lori this soon. Imagine like a chat GPT like interface, like a terminal where you can ask it any question about, you know, plugs into all of your, you know, data conversations. You can ask what was my booking rate, you know, last week? What was my revenue in this location during winter storms? And that's interesting. You know, what were the top reasons my calls weren't unbooked, were going unbooked. So you can ask it all these interesting questions directly. Get insights instead of needing to comb and filter through a bunch of these different views. And the coolest part is it has write functionality as well. So you can say hey, what are all the outstanding unsold estimates? Filter, sort that by day of estimate. Okay. For all estimates that are over four days old, do do this to them or do that or like send an SMS campaign, send a drip campaign with number one, number two, number three. So just a really cool thing I think for. That's really cool to be able to control their business.
- [43:11]
A
Yeah, I think, yeah, I mean that is, that is really cool. It reporting is tough. I was actually and maybe, maybe I'm just, I'm in like a weird zone where we're not small and we're not big. But I was trying to like access data today and honestly it was kind of a pain in the nuts to get the information I was trying to get because like I, I touch service titan reports once every three months. Like I, you know it's at this, they've been auto emailed to me for years so I haven't had to build something in like a long time and I'm trying to like fumble my way through this and it was, it was kind of ridiculous versus like yeah, it's like just A query makes total sense. And that's what everyone's used to now anyways. Like just ask ChatGPT, like, hey, what. What happened here? Like, walk me through it. So that makes a ton of sense. That's sweet.
- [44:01]
B
Yeah, exactly. So, yeah, I think it'll be then another nice part about that is it's a, It'll be a way to directly control your agents as well and like update them. Like, hey, we need to update, you know, before you would need us sometimes either go into this portal or, you know, ask your CSM to make updates. Now you can just ask AVO to make this update to the bot.
- [44:27]
A
Yeah, yeah, that's really sweet. I'm. I'm into that a lot.
- [44:30]
B
Yeah, that's. That's one thing other than that. We opened a Santa Barbara location around 2, so that was. Yeah, it's our first, our first new office and yeah, really excited. I'm going to go down and visit those guys next week and then, yeah, I think I kind of mentioned this very briefly, but I'm starting my own podcast, Evoke, a podcast called Trade Secrets from the Ground up. And that'll probably premiere, I think in early April.
- [45:10]
A
Oh, nice. Well, yeah, probably by the time this goes out, you'll have an episode or two out. So we should be able to link to with the show is it's on Spotify, Apple, YouTube, like pretty much everywhere.
- [45:17]
B
Yeah, yeah, we'll do, we'll do it all. And, and the premise is really to highlight, you know, really outstanding home service founders and operators that have built businesses from the ground up. And so for, you know, all the folks out here listening around, you know, wanting to start a home service business, thinking about taking the leap, you know, you can hear, hear what it takes and how to, how to, how to move in this new world of AI. First world.
- [45:44]
A
Yeah, yeah, that sounds awesome, dude. That sounds awesome. Do you want to tell me about this test that you guys did? Like the sort of like head to head?
- [45:52]
B
Yeah, yeah, that was actually a really interesting one. So Dave had this brilliant idea on. He wanted to know just how good the outbound AI actually was. And he was early, I think this was over a year ago.
- [46:07]
A
Oh, the. Oh, this is the out.
- [46:08]
B
This is the outbound. This is the outbound.
- [46:10]
A
Okay.
- [46:11]
B
Like SMS or sms, we do both. But he just wanted to know how good the SMS outbound agent was. So what he, so what he did at Four Seasons was a. It was actually a pretty cool. It's probably one of the best because a lot of times when you pilot it at different locations. There's so many variables, it's hard to get a true test. But he did this pretty well. Like according to the scientific method and everything. I think he had 50 of his CSRs. And then versus our outbound AI agent team do a split sample, random split sample of like 10,000 contacts each to try to book, you know, maintenances and inspections jobs for shorter season. And we were able, over the span of two weeks, we were able to, I think, primarily through understanding the right messaging, getting that down, timing, we and. And just speed, we actually won by, I think, one percentage in terms of the overall. Oh, my gosh, that's funny. But here's the kicker. All our AI agents could do was just text. His team of 50 could do calling, texting and emailing. So, like the whole, you know, trifecta. But yeah, it was great to see and he was very, very happy to find the result of that.
- [47:34]
A
Well, thanks for coming on today. This was a fun conversation. If people want to get a hold of you or contact you or find out more, where can they find you? You're on Twitter now.
- [47:44]
B
Yeah, I have a Twitter. I'm still not big on it. My email is just TysonVoca AI. And of course you can find us get a demo of the product just by visiting the website.
- [48:00]
A
Awesome. Thanks, man. Thanks for coming on. And everyone, if you like what you heard, make sure you like and submit.