How AI + Humans Are Building the Future of Lead Qualifying (Reworked AI Founder Interview)

Brandon Parker00:00Let's nail it. All right, Fred, welcome to the Legends of Land podcast. Here to talk to you today. I'm the co founder of reworked.ai. I'd love to start this conversation. Let's hear how you're using AI, machine learning, what you think the impact on business is. I love the quote.

Fred Castonguay00:18AI is not going to take your job, but somebody that knows how to use it more than likely will. And there are so many facets of our business. And I love this little story. We were talking with a collaboration partner about some marketing materials that we'd put together. And this gentleman we were talking to said, Oh, you can get some stock images and maybe change this. And he said, I like it, but maybe there's some different options out there. And our team said, No, these aren't stock images.

Fred Castonguay00:49We used AI to create the images. We told AI exactly what we were looking for. And this gentleman who's I don't want to say he's old, he's my age. But he was blown away. He's like, Oh, we should really be doing stuff like that. And it was just so poignant. And we're obviously using machine learning. That is a cornerstone of our business, how we use machine learning to understand dreams and gigabytes and terabytes of data. As you and I spoke the other day, I think the applications in voice are the most exciting.

Fred Castonguay01:20I've worked with companies and talked to some companies with what you guys are doing. It's just so scalable. And as you convey to me, you're using people to create, using people in the process, but where it's scalable, it's AI. And that's what it's designed for. It's not to replace, it's to augment and allow for the ability to make it better.

Brandon Parker01:44Leverage, Naval talks about the whole human leverage, technology leverage, money leverage. I think the leverage that AI gives us is just fantastic. And the first single person billion dollar business is now possible because of how much we can do just as an individual. So it's the people who are using AI in their daily workflows and moving five times as fast as they used to, I think that's the hidden benefit that a lot of people overlook.

Brandon Parker02:16They think, okay, there's gonna be like these cool tools that come out and we'll be able to do this and that. For me, it's just I type in 140 words a minute because I use an AI tool. I write my emails in two minutes instead of ten minutes because I use AI. All these things are just what I do on a daily basis. It's just starting to integrate into my life. You said it when we were talking. You're moving fast.

Fred Castonguay02:44I'm always saddened when people say that they're scared or what can AI do? Just look at the technological advances over the course of human history. The facts. Now, I could get a piece of paper in front of you, in the computer, email, all of these different advances, they didn't make things worse. They didn't put legions of people out of business. They created new opportunities for things to happen that people had never considered before. I agree with you. Just this morning, was working on a user case, taking some disjointed information.

Fred Castonguay03:20Two years ago, I would have worked on that for hours or had some team members work on it for hours. Ten minutes. Done. You would have had to pay somebody, Hey, these are my ideas, several meetings, concepts, back and forth. Now you have a couple key points that you want to get across. Give it to an LLN. Bob's your uncle, you have a press release that is perfect.

Brandon Parker03:47What do you think the highest leverage thing you do on a daily basis? Like, just could be small that you use AI for?

Fred Castonguay03:57Like you said, immense. Right? I'm always letting Gemini help me. Used to be autocorrect was you just tab forward and do the autocorrect. But now all of my emails are essentially and it's wrong to say they're written by AI. I'm still writing it. It's still my thoughts, my ideas, my concepts. It's just using a tool. It's no different than a pen. We're from pencils to pens or chisels. We're just using another mechanism, tool to help us communicate our idea that much more effectively to ensure the other person gets it.

Fred Castonguay04:33Because sometimes I can write an email and the other person is like, I don't understand exactly what he's trying to say. Gone. Miscommunication.

Brandon Parker04:43Lost in translation. An example of just tactically how I send messages just across the board almost at this point. I have a tool called Whisper Flow where there's a hotkey on my computer. So any text box on my computer, instead of typing, I hit the hotkey, and then I just speak, and the text box populates all formatted in what I said. And so I don't really type.

Fred Castonguay05:08That's how I can go 140 words every day. So this looks really cool tool. I've gotta try it. I've just learned something new. Right? I mean, I do it on my phone for text messages and email, but how stupid of me not to think that there's something for my computer typing. What do you need a keyboard for anymore?

Brandon Parker05:24Yeah. And I think the novel part of this is that it goes in every chat box. So if I'm using ChatGPT or if I'm in Notion or if I'm in a Google Doc or if I'm in Slack, I just hit the hotkey, I talk, and it pops up. It is interesting because it'll format. I'll put bullet points in, it'll put numbers and form out it really well, so it's a hot tip. But what I'll generally do is I'll whisper flow into ChatGPT. I'll do a quick brain dump like, hey, tell Mar we need to do this, Jan we need to do this, and this and that. And then I'll put it in ChatGPT, and then I'll say, write this in plain indirect language and simplify it.

Brandon Parker05:59And then I'll hit, grab the output, throw in the message, then in less time than it would take me to write the message, I've spoken it, formatted it, and then sent it.

Fred Castonguay06:13And your teammates appreciate it because it's easily digestible. I wasn't sure they got to call you to follow-up. Is it this? No, it's right there. Clean. Boom. What's the deliverable? What's the timeframe? Everybody's happy.

Brandon Parker06:30Something else funny I do, someone sends me a wall of text, I'll just copy it and put it in chat. Jsdp and say, Hey, simplify this for me. So I'll reverse the process for people's messages to me. I'm not going to try to repeat that.

Fred Castonguay06:43I just don't read them.

Brandon Parker06:45I used to not read them. I used to read the first sentence and the last sentence.

Fred Castonguay06:49Yeah, come on. It's too much.

Brandon Parker06:52That's funny. That makes you sound like I'm not paying attention. Yeah, that's something there's an interesting phenomenon there where it's like we get so reliant on AI where we start getting stupid. I feel like when I say something like that, it's like, that's happening to me slowly. I'm conscious of that too.

Fred Castonguay07:09What I mean is it the opposite? Our parents' generation were taught cursive and all these things. My daughter's 12 and she's learning cursive. I haven't written a cursive in forty five years. Does that mean I'm stupid? Does that mean I have a lack of intelligence? I would counter no. It's just it's moving. Ability to communicate is moving and getting better. But when someone sends me a blast of a text, it's like, come on. Originally, there was an SMS, a short message service, right?

Fred Castonguay07:41Emphasis on short, please.

Brandon Parker07:43Yeah. And on another note, something AI has done for me has unlocked a massive amount of great creativity. I'm a hyper creative person, but also because I move fast I've never become like a pencil artist or a good guitarist or these things that I might do but I'm not very good at. With AI, it's opened up this whole realm of I make music with AI, I make a lot of art with AI. If I can tap into this creativity that I wasn't able to in the past because it would have taken me too much time, there's too much of a barrier to entry.

Brandon Parker08:19We're talking about all the good things with AI, right? We're

Fred Castonguay08:23working with this new client and he said something to me the other day which made perfect sense. I had not heard it yet, but I don't think he came up with it. Because of AI overview, the traditional click, right? Think of it, right? You type something in Google, if people still do that, and the first thing that comes up is going to be an overview of what you just ask, with some insight and maybe some companies that are there. And Google and Gemini are going all in on that now.

Fred Castonguay08:54But their business, who's been around for ten, fifteen years and would get X amount of traffic from their website. Someone clicks on it. They can track the cookies to see who's there and reverse advertise for them. But their click traffic is down 60% because people don't need to go to a site to get the information that the AI overview just gave them. Unintended consequence for the business, but a benefit for the consumer and a challenge for the business to say, okay, the goalpost has shifted.

Fred Castonguay09:27It has moved. How do I move with it rather than just being upset, sticking my head in the sand? How do I create a better process to get to where the customer is so they can still use AR overview? But I make sure I have the right information there. Yeah.

Brandon Parker09:45And so some tactical information. I've thought about that also. There are some tactical information for people who have a business, and I'm sure you're doing some of this. I mean, you're on a podcast right now. I think that content becomes king. I think that having information for AI to go grab, here's Rework, here's Fred from Rework on a podcast, here's a bunch of things he said about Rework, how it works, And having more information out there for the LLMs to go ping becomes more of a benefit than the standard optimizing your website to rank higher on Google.

Brandon Parker10:28That's more value information out there. Yeah. And we're hearing this from people now about the SEO methodology

Fred Castonguay10:35of making sure all your pages are indexed. When the spider's crawling them, it's easily they can get in and out. I really think the content. Yes, I agree with you. Content is king. And it needs to make sense. Right? What's the Achilles heel of an LLM? Hallucination. So they're spending billions of dollars when they're providing output from an LLM to how many places can I authenticate this? Fred is saying A, B and C.

Fred Castonguay11:05That's one data point. AI does not like one data point. Okay, Fred said it. Joe said it. I read it over here. It was multiple points. So having content, even rudimentary SEO would want this. Are you linked? But it's less of a structural and more of a presence of the information that's verifiable.

Brandon Parker11:27Yeah, that's a really good point. The links, the backlinks, but that concept now extrapolated out to just information in general. And so it's almost like for reworked, if there's clips of you and sections of you basically answering FAQ questions, because that's going to be someone's typing a question into Google and then they're getting a summary. And that FAQ question that you've answered, that's like a backlink to reworked or it might be a direct link to rework, depending on digestible information.

Fred Castonguay11:58I had a watershed moment a week or two weeks ago now. A potential client reached out and one of the questions we ask everyone is, How'd you hear about us? And he said, I was having a conversation with he said, I was having a conversation with ChatGPT about how to use AI to help them find and go to where their consumers are in the roofing replacement business. So as he's going back and forth and ChatTP offered, you should look up and talk to Reworked AI and EagleView because they have a product in place that is using AI in several different instances to create and solve this problem.

Fred Castonguay12:40And so he told me this and it was so simple because it's content, it's verifiable information, it's multiple sources. And ChatGP said, Go talk to these guys. They're doing something that is along the lines of what you're looking for. And since then, we've had several different companies come into us like that, but it's the name of the game. Ignore it at your own peril.

Brandon Parker13:01The caveat there is that it's not necessarily... I was picturing you and me who seems like I'm thinking forward so often. I was still picturing someone typing a question into their computer, but that's not even the input. The input is a lot of times even maybe a voice conversation with chat.

Fred Castonguay13:20You were talking about whisper flow, right? I mean, my wife doesn't type, right? She does, but she'll have conversations with chat on her phone. My wife's first language is Spanish, so she's having it in Spanish And it's back and forth. They're literally having a conversation, which is great when you talk about the advancement of information, the availability so quickly. As long as one has enough ability to understand where something could be wrong.

Fred Castonguay13:53Because there's plenty of instances where large language models give bad information. So consider the source, verify the source, take a moment and say, Wow, I shouldn't drink Clorox. No, you shouldn't. You should check that. But it's with anything. It's when newspapers went from The New York Times to USA Today. You're getting all your information in a snippet. You've got to verify.

Brandon Parker14:19Yeah, there's the learning curve historically with new technology, how we get information specifically like advertising and marketing. It's almost like a cat and mouse game. And so, I like to think about there was this Rolex watch commercial back in the day. Not Rolex, but Timex. And this guy jumps off a cliff into the ocean and pops up and his Timex keeps a ticking. It's a famous commercial because maybe that watch wasn't even waterproof or something. People just saw something on TV and it was the first time that there was this betrayal of the truck.

Brandon Parker14:57They said all this stuff and everyone just believed it wholeheartedly because they saw it on TV and then realized, oh, it's just marketing and it's not even far referred to. And then everyone kind of slowly learned what TV marketing was.

Fred Castonguay15:07It keeps on ticking, right? Oh, there it is. But think about this in the traditional Googling of something. You're looking it up. I am looking for a place to get a pizza. It gives me the top picks or Yelp and I'm going there and I'm making deductive reasoning on my own of what type of pizza I want. I may look at reviews. It takes a little bit of effort. But if you're conscientious, you probably get to the right place at the right time. Now with the large language model, as that gentleman said to me, he was having a conversation with a large language model.

Fred Castonguay15:39That large language model has taken on a human because he felt he was having a conversation, he felt that the large language model was telling him, Call rework dot ai. These are the guys you need to talk to. So when he got to me, he was ready to sign up. He said, yeah, I'm going forward. So that's great. I love it. Thank you. We believe in our product and we know it works and we stand behind it. People still it's a large language model.

Fred Castonguay16:11It's a computer. It's a massive collection of algorithms and data. Do your own homework. Make sure you're asking the right questions. But he said, No, Chad CPT told me to do it. I'm ready to go. Okay, great. But take things with a bit of skepticism and make sure you validate it before you drive off the cliff.

Brandon Parker16:32Yeah, so we're in that blind trust phase here. Interesting. That's a really interesting take. When those people call in, what are you actually doing at Rework? Why don't you give a of a Let

Fred Castonguay16:45me just give you the overview. We started rework.ai four and a half years ago to help land flippers drive efficiency in their direct mail outreach. I was an ammo land flipper. I do less these days, but I was sending 12,000 to 15,000 mailers a month. Good deal flow. I was doing well. Ended up sending this gentleman a mailer. He's now my business partner, computer engineer, data scientist. We put the company together. We started with land flippers, got into all facets of real estate, looking to expand, started in solar business.

Fred Castonguay17:21Solar quickly led us to roofing. And as part of what we're doing is we're using machine learning to look at the underlying demographic components of a homeowner or a landowner to determine their buying propensity. That's our Betti score, skeleton structure over into solar and roofing. We had to train the models based on existing data that we bought. The problem that we had in both solar and in roofing was we developed a computer vision model that could look at an image and determine if the condition of the roof was warranting a replacement or there was damage.

Fred Castonguay18:00The problem we had was the imagery was inconsistent. We did many pilots and worked with several companies and sometimes it would work really well to identify the bad roof, sometimes it wouldn't. That led us to EagleView. EagleView is a massive company. They've been in the aerial imagery and technology space for many, many years. They have timely high resolution, high fidelity aerial images taken by a plane, not a satellite. And our collaboration with them has solved the roof condition.

Fred Castonguay18:33Right. So what we're doing in the roofing space is bad roofs, good people. Right. We can find the roofs that need attention, whether they're about to fail or failing and people that can afford and are look like people that have replaced roofs. And it answers the age old question, where are my customers? And it's being adopted. We've rolled it out three or four weeks ago officially with EagleView, our collaboration. And we're really pleased because there's a lot of roofers in The United States, they're all trying to get to if the person already knows they have a problem.

Fred Castonguay19:12If you're in your home and it starts leaking, you know you have a problem. You're going to start calling roofers. But if you're sitting there and a mailer or a NOC or some form of outreach comes in and says, Dear Mr. Parker, our records indicate, our imagery shows that your roof is about ready to fail. Okay. You might go take a look. You might talk to your partner and you didn't know that before that showed up. And then you look at the window or you get on a ladder, you're like, wow, they're right. I better call somebody. So it's discovering who's got the problem that doesn't know they have the problem.

Fred Castonguay19:44And that's that you talk about trust from the LLM, right? You're demonstrating credibility to somebody. That person will respond for that credibility with, come and take a look, come and help me. I'm curious what they are actually looking for. Like how can you tell what a bad roof is from an aerial view? Hey, I that's the short answer. Yeah, it's it's why we're here is completely right. And it's the same. You know, I've told this story to so many people when you're teaching a baby the difference between a dog and a cat.

Fred Castonguay20:19Right. Dog, cat. The baby is starting to pick up, oh, pointed ears, whiskers, all these nuances. There's no one thing that someone could say, all dogs have this, all cats have this. It's not one single data set that is going to train that baby. That baby's going to learn of what a dog and a cat is. How would you explain to somebody that's never seen the color red, Has never seen color? How would you explain that? There's no one thing.

Fred Castonguay20:49So when you're looking at a roof, because their computer vision, which is phenomenal, has been trained on literally millions and millions of images, It knows even the computer vision that we created, which is not as good as theirs. We were able to train the computer vision to see things that a human couldn't Because the human eye gets tired, you're looking at it. So it's using AI to zero in on it goes down to one inch on their imagery and it sees irregularities.

Fred Castonguay21:24The same thing that we were looking at, discoloration, pooling of water. Something looks out of place and they can get down. One of the outputs from EagleView for us is a simple red, yellow, green. This roof is red. That means it's at or near failure. Then we're looking at the homeowner, their credit score, their income range, their marital status. So it's the right roof and the right homeowner to say, yeah, this person is more likely to have a conversation. And then we can tweak our dials, right?

Fred Castonguay21:55Our AI is dials, not switches. We've got customers that say, I don't necessarily need roofs that are at failure. I'm selling a product that is a it's an upmarket. Right. I'm selling a metal roof. So I want to find people that are in asphalt shingle roofs that have the wherewithal or in a house that is of this value and look like this to upgrade to a metal roof. Or I want to sell a composite roof in Florida that mimics and looks just like a Spanish tile, but not everybody can afford it.

Fred Castonguay22:31Instead of being a $35,000 roof replacement, it's a sixty, but it's a fifty year roof. So the way to look at the homeowner and the roof really gives a significant amount of insight to these roofers, manufacturers and distributors that just is not on the market. This ability to look at that is not on the market today, except for us and EagleView.

Brandon Parker22:52And there's a you have quite a moat because of the data. The amount of data that you're collecting over time, like you're giving the example of the dog versus the cat, things just need to be learned over time. And the people who hold that data have an exceptional competitive advantage. I the current gold rush almost feels like, yes, of course it's AI, but I think it's also data. A lot of the large language models you hear about have capped out on the amount of information.

Brandon Parker23:25They read everything on the internet. And it's like now, the next gold rush for large language models is where can I find unique data sets? And so, they might come to reworked AI at some point and say, Hey, you have all the information all about roofs and you've looked at billion roofs at this point. You're the guy that you want my roofing information from. There's a second layer of opportunity collecting data. Yeah. Data is key.

Fred Castonguay23:53But I would say even more fundamental to data is money. Right? And if you look at the large language model leaders today, there's no startups. The funding that is going into these companies, right? Gemini, Claude, you know, with Authentic, these are highly capitalized organizations that it's very difficult for a startup that would get maybe a pre seed of even $100,000,000 which seems like a lot of money.

Fred Castonguay24:25They're not going to touch it because it's compute power, right? Data is the Internet is there except where it's up behind a paywall, which is becoming more and more prevalent. Everybody wants to keep Reddit now. Nobody can crawl Reddit for this information. So everybody wants to keep their data for themselves. But yeah, it's it's having that data in. And that's the when we look at EagleView, we look at them as a big brother. Right. First of all, they're a fantastic company. I have not met anybody within the organization that's just not a really good person and really committed to what they're doing and committed to driving change through using technology to drive change and help their customers.

Fred Castonguay25:07They really care. But they're big. And their moat is they've got like 65, 70 planes. I mean, they own the plane. They have software. They've shown me the software. I obviously don't have it because I don't work for them. But they've got software that they can see where all their planes are at any given time and they're flying. I hope I'm not going to misquote them. The 50 largest metropolitans every six months at 6,000 feet. Right. You want to talk about a moat, right? The ability to have that underlying data as proprietary with their cameras.

Fred Castonguay25:41Right. All of these things that they've developed is phenomenal. We're just tickled to death to be collaborating with them because we're bringing the opportunity to connect on a retail basis with roofing customers and hopefully in the future other type of customers. But yeah, it's money, data and create a moat.

Brandon Parker25:59Yeah, I love thinking about this stuff. That's something that is unique.

Fred Castonguay26:05Think I moved my AirPods. Okay.

Brandon Parker26:08That's something I think about a lot at Land AI because what we're doing is we're having these professional level sales conversations with an AI acquisitions manager. So, we're doing is we're getting motivation, price, and timeline out of people. Why they want to sell, how much they want to sell for, and when they want to sell. We started off very tactical with these are the types of questions we answer. We want to be like Chris Watts.

Brandon Parker26:40We want to mirror. We want to open ended questions. We want to use calibrated questions. Over time, as we've collected bigger amounts of data on how these conversations go, then it starts to become more like what you're talking about where a cat dog and the AI just begins to learn, here's what I say and this actually works. And this a better flow of the conversation to get that information. And so, right now we're pinging back into a vector database and we're just starting that iteration of V2 where it's going to start self learning on the thousands of calls that we've had.

Brandon Parker27:17And that is just going to be such an epic moment to see it start learning and see the places that I can only think so far as to how to structure the conversation. Yeah. When you were explaining to me last month we were talking about what you guys are doing,

Fred Castonguay27:31I think it's wonderful. I think the application, particularly how you're doing it. I like the marrying of scalable technology and human beings. I think that's a good secret sauce. What I would love to know in a period of time when you've got the data is we all as humans have bias. And when you guys were putting this together and I know you use many different experts to create scripts and develop the underlying starting point.

Fred Castonguay28:02At some point you'll start to see your AI start to do things that is data driven, which is non bias driven and is correct. There'll be a divergent from what you guys were your original, oh, this is the way we should do it. There'll be a divergence and it may be slight at first, but it will happen. And that's going to be your moment when you realize that, okay, we've got enough data. This thing is in. It's not a bad thing.

Fred Castonguay28:33It's not a bad thing that this AI that you build is learning. It's a good thing. Who's going to benefit? The consumer, your company, your customer's ability to get to a deal to drive revenue. That's the value of all this. But I can't wait to talk to you and learn about that divergence because it doesn't mean that it was set up wrong. It's human bias, right? And the data will drive the decisions and drive where it goes.

Brandon Parker29:03Yeah, I'm really excited to see what starts happening. We have these breakthrough moments with the AI, which is another fun part where it gets twice as good. It's like one day it was like this and now it's twice as good. Woah, that was epic. That's the fun thing about all this AI that when you have an infrastructure built, sometimes it's just a model upgrade. We could upgrade chat GPT or one of the language models and all of sudden everything gets better. Specifically, we're doing just to catch people up if they don't know is we do cold calling. We ran a call center for years and got really good at cold calling.

Brandon Parker29:37The issue with cold calling is it's just too much garbage comes through the pipeline. We get 100 leads and only one or two of them is actually going to be a contract. And so how do you sort through those leads? So what we started doing was doing the cold call with a human and then live transferring to conversational AI. And conversational AI has a really high level sales conversation. So you get the best of both worlds. You get a cold call, which is massive outreach, but then you get the lead qualified instantly so that you don't have to deal with a bunch of tire kickers.

Brandon Parker30:09And so thinking through the benefits of AI, I just had this moment where I thought, oh, you know what? This actually benefits the tire kickers too. Because what happens is they talk to AI, we realize, hey, this person's not gonna sell, and then we stop calling them. Usually, those would get the triple dial from a land investor who's trying to do his job, trying to get them back on the phone and see if they wanna sell. And they were just kicking tires and now they get 23 phone calls from some random person. That doesn't happen anymore. Stop calling me. I will stop calling you.

Brandon Parker30:41Talk to Maya. She sorts you out, but we don't need to talk to you.

Fred Castonguay30:45I have a question for you, Brandon, on that. I'm a big supporter of what you guys are doing. Share with me about your controls that you're building in there. If it's following the data, it could go off track. Are there guardrails on this conversation? How are the engineers and yourselves making sure that directionally the conversation is what you would want to be proud of as part of your company? I'm just curious because it's not my area of expertise.

Brandon Parker31:10Yeah. So first we started off with a more strict conversation flow, so it kind of followed. And so, our guardrails were very specific, like you couldn't knock it off track. So, that's where we started. And then we started as it was working well, we basically started taking guardrails out and giving it more end states making sure so it's getting to the same place without as many guardrails, but removing them safely. At a high level, that's the way that we've approached it.

Brandon Parker31:43The final state of this being where we have enough data over time that we can kind of set it free and it won't need guardrails. It's semi structured now. It's kind of in between state, if that answers your question. Yeah, absolutely.

Fred Castonguay31:57I think it's important to note, I mean, like our primary AI is machine learning, which one of the key characteristics of machine learning is it's supervised, right? The input is known. The end state, the output that we're looking for is known. We're simply asking it to analyze a significant amount of data in a very short period of time. And I think for you guys, you're not you know, you're not pure unsupervised LLM. You've got to start a couple of starting points and gates that you're asking the machine to go through to get to where you need to go.

Fred Castonguay32:29So it's really encouraging to hear that you've got guardrails in place because they'd always need supervision of some degree. We all see the stories that come out about questions coming into an LLM. And of course, that's one of 17,000,000,000,000 inquiries that gets on the news. Yeah, It's something to be aware of.

Brandon Parker32:47Yeah, absolutely. So you can put the guardrails in place. Also, a key part of it is a fast feedback loop and a human in the loop, which we always have. That's something that we press hard on. What that looks like is every time, literally every call that comes through, transcript passes to a human who reviews, looks for tweaks and improvements to the model, looks for mistakes.

Fred Castonguay33:12The human loop is very important with AI. Think people overlook that often. For every conversation that's going to go to it's getting set up for follow-up.

Brandon Parker33:22Yep, everyone that comes through. Exactly.

Fred Castonguay33:25It's impressive.

Brandon Parker33:30I would say we're not saving money by using AI. To some extent, sometimes we spend more money, but in the end we will be more efficient. You're investing in your product.

Fred Castonguay33:42You know where you want to go and you got to get there. In order to keep from driving the car off the cliff, it takes investment.

Brandon Parker33:51Really interesting thing that I found, and this is just an unexpected fun fact, and it'll resonate when I say it, is there was this question that we had in the beginning. Are people going to want to talk to AI? Or are they going to be like, I don't want to talk to AI? That was something that we did sentiment analysis and we started measuring. We're going to measure how many people drop off. It'll probably be like 10% or 20% or something like I don't want talk to them. The first surprising thing was was under 1% who said, I don't want to talk to AI because we tell them before we transfer.

Brandon Parker34:21We're going transfer you to this really cool AI tool. Under 1% said, I didn't want to. So, that was surprising. The next thing that was surprising is people are just exceptionally candid with it. They just say things to it and just Have a conversation with ChatGPT. I knew you'd feel it. It's just like talking to ChatGPT, just tell it what you're thinking. There's no filter. And this is kind of a secret hack for land investors because it's like, let's say, I think it's worth this much. Oh, my brother was kind of dick or whatever.

Brandon Parker34:54Moved out of state. They'll just tell you.

Fred Castonguay34:57You were telling me that before and it was a revelation for me because before you said it, it made sense. Major corporations in The United States, and I'm assuming globally, have set the stage for voice AI success because right now it's a binary choice. I've been dealing with two fairly large package delivery services in The United States. I won't mention their names. We know who they are. Their customer service and their ability to get to an answer via a phone or a website is deplorable.

Fred Castonguay35:34Right. So if I could call in or chat with a real intelligent AI, not a supercharged FAQ if then statement bot, but if I can have a conversation with my package is lost. My package, I'm getting a message. The package address needs to be updated. Oh, please give me a tracking number. Read the tracking number. Oh, I see here. Please give me the update. Okay, I've got it. I would love it. I would be tickled to death. But it happened to me just this week and it was an act of Congress to get and I never sure that I did it right.

Fred Castonguay36:13And then, you know, the package did go up. So the end result, it was okay. But I think the corporations have set this up and trying to drive efficiency and get them press 9 for this. No, just have the phone answer and say, You've reached ABC Corporation. I am a large language model that is here to assist you. What can I do for you today? And you don't have to get transferred. You don't have to say book a flight, change a reservation. No, just tell me what you're looking for. And I think a lot of people would find that quite refreshing to get away from the, please listen to this menu, but yeah.

Brandon Parker36:48It's true, though. I mean, that resonates. That era, even when I hear those now, I go, how is Chase Bank not as far ahead on conversational AI as I am right now? Why not? Because I shouldn't be beating Chase Bank to the punch on how to get someone nice to talk to on the phone. Citibank, Chase Bank. Right now you've mentioned some of these banks. I get my

Fred Castonguay37:12wife and I were both members of American Airlines, whatever it's called. Right. Once a week we get a mailer for Citibank American Airlines card. And it's amazing the lack, the analog approach to mail us. And it may be something as simple as every time I take a flight with American Airlines, a trigger pops, send them a mailer. It's every week because I'm usually flying somewhere. So it's that.

Fred Castonguay37:42Here's a giant corporation with terabytes of data that should be able to think, okay, I've sent them the same piece 70 times. Maybe, just maybe, I can mix up a little bit, send something different. I don't know. Or maybe stop because these are not cheap mailers. These are large, glossy, they're big, they're thick. These are not 60¢ postcards and they're just burning through cash. And I guess some people sign up and that's all they need. But to your point, who drives innovation?

Fred Castonguay38:16Not large companies. There's two things that drive innovation in my opinion, war and small, agile companies that are seeing a problem. And yeah, the fact that you're leaps and bounds ahead of a large corporation, it's not surprising. Not surprising.

Brandon Parker38:33Do you see reworked going a couple of years down the road? You're doing the good customer, bad roof right now. How do you see that developing?

Fred Castonguay38:43Yeah, the short term roadmap is the whole home services space, particularly focused on exterior window siding, lawn, tree care, pools, patios, because they're all visible. We're working on some ability to look at internal aspects, HVAC, water heaters, particularly HVAC because that's a bigger ticket price. But we lose the vision at that point. Not going to be able to then we're relying on if there's permit data there.

Fred Castonguay39:16We would love to, I mean, we've got a couple insurance customers and they're very demanding. But the fact that if you can get them on board and that's why I talk about my frustration with these large banks, right? Love to be able to get inside of a large bank and help get the right information to the right people at the right time. But just in roofing alone, there's ninety, ninety five thousand roofing companies in The United States. Many are being bought by PE firms. There's significant consolidation.

Fred Castonguay39:47It's an industry that we're you know, there's enough runway here for some time to really drive innovation. So right now, of our focus is on roofing, ensuring that our real estate clients are getting updates. Matter of fact, we're doing a webinar in a week or so real estate products. Our team continues to grow to make sure that we're not losing any focus. So the future is very bright and we're really excited of what's going to be out there. As we said before, I'd like to see where you're going and how as you grow and continue to get very good at real estate.

Fred Castonguay40:19Is there a home service play for your product as well?

Brandon Parker40:24Yeah, it gets really interesting a couple of years from now or even just a year from now. It's a lot of development for the land from betty.ai, which is your land sorting tool or the data sorting tool. The way that I understand it, it is removing most of the people who won't sell and just the remaining people who are there are the people who have a good possibility of selling. A good explanation? It

Fred Castonguay40:52doesn't tell you where the needle is in the haystack. It just says there's no needles in that portion of a haystack. So don't waste your money on mailing, whether it's in land or just real estate. Betty will go through and she's looking at the characteristics of the owner, not the property. And she'll come back and say, this half, this third, this two thirds, they're not going to be interested in a deal. Don't bother reaching out. Don't call, don't text, don't mail. It's not going to work. Just focus on these.

Fred Castonguay41:23So that 50,000 line list, whatever amount of deals you add in there, five, ten, fifteen, twenty, five, thirty, doesn't matter. Instead of having to mail an action all 50, you're now actioning twenty, thirty, 25. So your costs go down. It's an efficiency play. It's using the demographic characteristics of the owner to predict their propensity to take an action.

Brandon Parker41:47I think that probably a mental block for land investors using that is like, what if there's one in that other 1,000 I'm throwing away? How do you think about that? Yeah, I do. It's FOMO, right?

Fred Castonguay42:03And I was talking to a new client this morning and he was tuned in. He got it. He's like, I'm not worried about missing one. I'm not near 50. And if I can drive efficiency and scale and be capital efficient, great. But the funny story I was talking about was a client that says, well, I don't want to miss one because it could be very, very lucrative for me. I said, look, look at the properties that you're mailing to. This is a landflipper. These properties are anywhere from 5,000 to $10,000 in tax valuation.

Fred Castonguay42:38And you want to mail X number. And if you missed one, which Betty is not perfect, there's about 98.3% accurate, it does happen that she misses one. What are you really looking at? If you want to make sure you don't miss the one, then double your data and mail the same amount instead of mailing 5,000 or instead of getting 10,000 of data and mailing 10,000, get 20,000 data, run them through Betty, mail 10. So if you were going to have five deals in your original 10 that you were going to mail, now you're going to have nine deals because you did miss one.

Fred Castonguay43:16Possibly you missed one, but you doubled the number that you were starting from. So instead of having five, you go ten minuteus one, you're at nine. So you're way ahead of the game. You're still four up on what you were going to do before because data's cheap, mailing's expensive. That's what we're seeing. Get as much data as you want, but just be careful on how you mail or call.

Brandon Parker43:38I think that's a great take. It'd be interesting to connect Maya and Betty together, Maya AI and Betty AI.

Fred Castonguay43:48You're to put us out of work.

Brandon Parker43:50Have them connect as little AI friends and for land investors.

Fred Castonguay43:55Well, and Betty can talk via API. Maya, here's a list. This is Steven's list. Here's the ones that you need to action. And then the beautiful thing for both of us is Maya can give feedback instantaneously on something. So a lot of our feedback now is some of it's coming via API on a monthly basis. Some of it's coming via email spreadsheets. Hey, this is our inquiries and deals. But that instantaneous feedback of we just had this call transaction, boom, we're elevating it or moving it down the funnel.

Fred Castonguay44:30Boom. Betty, okay, good. Got it. I learned. So yeah, that technology is here. It's just a matter of finding the customer that's using you and using us and get out of the way. Rather than getting a file and getting it through us and then mail it to you, it's like, no, here it is. Boom, Betty, send it to Maya. Maya, send it to Betty. Everybody do it. You just tell me who I need to write the check to buy this property.

Brandon Parker44:55What are the main things about AI or your business or business operations you think we've kind of glossed over? Where are you going to dig into The

Fred Castonguay45:06one thing I always tell current and future customers is if it doesn't work for you, if you use it and you don't get the results we're talking about, let us know. We're going give you your mind back. We're not here to get $149 from Joe. We're not here to get $2,000 from Steven once. We want to be ingrained in their process. We want them to subjectively and objectively know this works and they are getting value.

Fred Castonguay45:37It is helping drive their business. It's helping them save money. It's helping them spend more time with their children, play golf, do whatever they want. It's an integral part of their business and they wouldn't dare go to market without it. So we stand behind it. A couple of times this happened. People say, I didn't see it. It didn't work. No problem. We'd like to know why it didn't work. Give us some information, but we're happy to be passive. You don't want to share it. You're upset. You're having a bad day. Okay, we'll give you your money back. We stand behind our product, both in terms of being there for people as well as if it doesn't work.

Fred Castonguay46:07What bigger picture patterns are you seeing in the land investing business? I'm seeing a bit of panic right now. There's three types of land investors, right? You've got your core, nucleus, that have been doing this for years and they're scaled and they've got processes. Then you've got the next tier of people. I consider myself in that next tier that it used to be bigger for me, but they know how to do it. They understand that the real art in land investing is not buying the property.

Fred Castonguay46:38The real art in land investing is selling it. Right. Because any arugatana can buy it, but can you sell it? And then you've got this big cadre of people that want to get in. They see people like us talking about it. They see all these podcasts. They do it. Quit your W-two. It's scary. Right? And there's a lot of money to be spent along the way. And it pains me to hear when people I talk to them like, Well, they bought a property, they can't sell it. They're $50,000 in and this is not what they expected. But it's not a difficult line of business, but people have to be careful and there has to be a lot of common sense.

Fred Castonguay47:14And the core people are still doing well. One of the gentlemen I was talking to this morning, real pragmatic, he's like, there's so much land out there and there's so many people that aren't. We're not even scratching the surface on this. But I think there's a bit of hysteria in our country as it is. So it just bleeds into everything. Everything becomes a little bit more difficult and a little bit more dramatic than it really needs to be.

Brandon Parker47:38I think that the way that I picture land when people talk about, Oh, it's getting saturated, I think we're so far away from saturation. It's not like there's this much land and if it gets bought, it's gone. It's a fluid market. People are buying and selling land like the stock market. You can't buy all the stock market. I sell land to land flippers.

Fred Castonguay48:01I always have four to five, maybe six properties in my quiver and maybe one's not moving as quick as I wanted. So I try to get into 30% and get out at 60 to 65. But if it's not moving, invariably a flipper will contact me and I always tell them, hey, I bought this on flip, but this is my price and I'll come off a little bit to leave a little juice for them. And I sell it to a land flipper. And maybe they're better at marketing it than me. Maybe they're just better than me. And that's okay.

Fred Castonguay48:32Right? I tell them land flipper, this is the price I'll sell to you. It's wholesaled. It's not wholesaled in the definition, but I'll leave something for you as well. Then I'm fine. I've covered what I've put into it, made a couple of bucks. But to your point, this stuff's changing hands all the time. People are dying all the time. People are getting divorced all the time. Life is constantly changing and that's one of the key things that Betty is looking at is all those life events. Credit score movement, credit score overall, but credit score movement is a huge indicator of what someone could be thinking.

Fred Castonguay49:04Good or bad. Won the lottery. Bankruptcy. Right. Where you were a hold three months ago. Now you're thinking I might need to get rid of this. So, yeah, I agree. We're not even close and there never will be because someone will always want to sell and someone will always want to buy.

Brandon Parker49:26On the business side, I think there's always room for good operators. If you come in and it doesn't matter what business you're running, if you just do the basics and the fundamentals really well, then you're going to be successful. So, maybe there's not just money laying in the street for people to fall over in land investing.

Fred Castonguay49:49Let's work. Some of the hardest days ever put in are buying, convincing, getting ready to sell, it's work. Well, let any of the talking heads confuse you that it's an hour a day and it's simple? No, it's work. But you're the boss. It's your own business fundamentals. Keep true to yourself. Put a budget together, understand what you want to do, create some small goals and move on and be careful of signing up for too much coaching and all this other stuff because the only person that gets rich at a gold rush is the person selling the shovel.

Fred Castonguay50:27Even the workhouse sells shovels, but ours are good shovels.

Brandon Parker50:31Yeah, big time. What I usually end the show with is a it's a tweet. But if you were to send out a short message, like a tweet that everyone in The US would see, let's say every land investor, every business person in The US would see this tweet, what would you say?

Fred Castonguay50:52Bad roofs, good people equals rework dot ai. I love it. It's

Brandon Parker51:02a hot one to end on. It was great to talk to you today, Fred. Thanks for stopping by. Thanks, Brandon. It's good to see you as always.