Latest / Real Investor Radio Podcast / Ep. 98 | Brad Bone: From Real Estate to AI with River X

Transcript

Ep. 98 | Brad Bone: From Real Estate to AI with River X · Real Investor Radio Podcast

• Craig (00:14) Hey welcome back to Real Investor Radio everyone. I'm Craig Fuhr coming to you today with a rare Friday Ravens jersey. I'm not going lie Jack, I did forget about our Friday recording today so we're recording on this is September the 19th, don't know when it will air probably shortly. But yeah it was Ravens Friday around the office today Jack so I wore the jersey. Good to see you my friend. Jack BeVier (00:37) Absolutely, absolutely looking forward to the game on Monday. Craig (00:40) We've had a little bit of a summer hiatus, Jack. I actually had an actual listener yesterday reach out via email and go, hey, I got to have my real investor radio. Where you guys been? Like, what are you taking a little break? And you know, everybody takes a break, Jack, even the best. Jack BeVier (00:55) ⁓ our bad, our bad. We'll get back on the horse. Craig (00:57) What are you doing? some great guests coming up and today we have an awesome guest that we'll introduce in a second. Brad, you can feel free to jump in anywhere. Jack and I usually start the show off with like a quick little news story. I had two today. We'll only cover one. Jack, we need to talk about the TOPA, the rental act that's happening in D.C. but we'll do that on a smaller, another episode. But Fed rate cut. Jack, we've waited all year for it. And just like he did last year in September of 24 on the 18th, the Fed came out and we were taking bets around the office. We actually had guys that were believing that it was going to be 100 bips. And I told them they were all smoking crack, ⁓ which is not condoned at all here at Dominion Financial. So in any event, it was a 25 bips cut, Jack. What's your take? I have a take. And how that all affects DSCR. Brad Bone (01:47) Happy. Jack BeVier (01:50) Yes, so the long awaited cut, There's been pressure from the White House for many months now to get this rate cut. And ⁓ I think that I was looking at some of the future markets, poly markets, stuff like that. And there was even some pricing in a potential for a 50 basis point cut. But I don't think anything above that. to the extent, once it seemed like a couple of weeks ago, that jobs report, Brad Bone (02:10) Thank Jack BeVier (02:19) came out weak and ⁓ that seemed to like really close the door on the debate as to whether there was going to be a cut at all. And everyone agreed like, yeah, it's time. And now the debate is, you know, whether it was going to be enough and ⁓ were they too late? This is not enough. And are we actually, ⁓ do we actually need more than just this 25 basis point cut? So. Craig (02:44) Well, you know, the second thing that I think a lot of people are looking at, Jack, was sort of the softening housing market around the country ⁓ and, you know, just a real reluctance on on most home sellers to sort of go from what might be a three and a three percent or minus 30 year mortgage up to over seven. We saw some rates kind of, you know, going under seven ⁓ as far as a 30 year ⁓ note goes, Jack. ⁓ Brad Bone (02:54) those four slides. Craig (03:11) Brad, you can jump in here anywhere. I'm sorry. You have to stop us when we close our mouths. But Jack, so softening housing market, we're seeing it in lot of markets around the country. How much of effect do you think this rate cut will have on the, which generally the 10-year bond, the price is all 30 Brad Bone (03:21) much much Thank Jack BeVier (03:28) Yeah, I think that once that jobs report came out, you it was several weeks ago, the mortgage market priced in, you didn't wait, right? They didn't wait for Powell to, you know, for anyone to actually formally drop the rate. The, I think the thinking there was that, and it's not, so the mortgage rates came down as in the wake of that jobs report, not because people anticipated a Fed rate cut, but rather because that indicates that inflation is no longer or less of a factor, no longer a factor. And as a result, the long end of the curve can come down if the market believes that inflation is no longer persistent and pervasive. so people sometimes ⁓ connect the Fed with moving mortgage rates. there's really, that's not quite right, right? Like the Fed moves rates down if inflation is under control. or is coming down and they're worried about a softening economy. And that lack of inflation may move, expectations may move the long end of the curve, but the Fed's only operating on the short end of the curve. those decisions are correlated, ⁓ where inflation comes down and the Fed may drop rates and mortgage rates may come down, but it's not correlated, not causational. The jobs report is what's causational, right? Like how the economy is doing is, Brad Bone (04:40) Thanks. and I'll you guys later. Jack BeVier (04:53) the driver, these other two market, know, what the Fed sets short-term rates at and what the mortgage market sets long-term rates at are derivative ideas off of that jobs report. In this case, the jobs report. Brad Bone (05:00) you I think, I think what's, yeah, I think what's interesting is, is on the day the fed cut the rate, ⁓ mortgage rates actually went up and it's actually, we had two days of, of increased mortgage rates. I think it might've dropped just a little bit today, but, ⁓ really that was all priced in a few weeks ago. And then all of sudden I think we got a little bit of a hot, report or no, I guess it was the comments that Jerome Powell made. Craig (05:07) Chump it. Brad Bone (05:33) made everybody a little bit nervous. And then we got a little bit of a report that was a little hotter yesterday or maybe it was today. Anyway, we've had a little bit of an increase. I'm hoping the trajectory is still down. We need it to go down, ⁓ which I think it is, but, but yeah, that that's what's always interesting is people think fed drops rates go down, not necessarily, you know, cause that's Craig (05:53) Even in your first comment, you're proving to be my favorite guest ever because you've just literally teed up exactly what I've been writing about. And honestly, talking about what Jack, around the office here for at least six weeks, we saw a softening in the five-year Treasury Jack over the past six, seven weeks. I'm saying six, seven here. And I went back Jack last year. And I took a look at this time at sort of July, June, July, August of last year. And what we saw during that time was a softening, a fairly significant softening of the DSCR rates. And frankly, the phone started ringing off the five year and obviously the DSCR after that. And then the phone started ringing off the hook with investors wanting to refi with DSCR bread. Everybody was waiting, waited, but there were still a lot of guys on the sidelines waiting to do the refi until they heard pow in September, which was actually September of 18th, September 18th of last year. And literally the next day rates started ticking up. And over the next couple of months, it was generally above four. The yield was above four, got up to probably four and a half. And then, you know, so we saw a tick up in the same ⁓ five year treasury yield. just after the announcement a couple of days ago. So I still think, Jack, that right now is a real sweet spot for guys who are looking to refi and get into DSCR loans. It's still the best we've seen in probably a year, frankly. But I don't know how long those rates will stay down because I still think that there's a lot of uncertainty with bond buyers over the overall economy. ⁓ deficit spending and the ever rising debt. Jack BeVier (07:42) to that point, I got a hot take for you, okay? Craig (07:44) ⁓ contrarian. Brad Bone (07:45) Thanks for watching. Jack BeVier (07:46) Maybe, I don't know. So I think that this is just a theory, right? This is just Jack crystal ball in here, okay? But I think that if you look at the shape of of the yield curve right now, it looks like this, kind of bottoms out at like the two year and then starts to tick back up, but it's still like a U shape. So it's not a normal shaped yield curve. I think that we're though at the beginning of a, you know, rate tightening cycle where the short end of the curve is now going to get pushed down by, by the Fed. And so I think that, you know, over the course of the winter here, we're going to see, you know, normalization back to a normally shaped yield curve. But then I think that we're going to have, I think that we're going to have some labor market, continued labor market weakening. And the reason I think of that is I think that is because of AI. ⁓ I think that none of the AI stuff that anyone's talked about, none of the investments have really, they've really, just starting to make it into like production at different companies. And, ⁓ you know, people have started to talk about how entry level jobs are, have pressure already because people were like, don't hire some 22 year old, just get AI to do it. ⁓ And so those people are getting a little, you know, have, are having a harder time right now. But I think that's going to move, you know, as AI is more more adopted, that's just going to move up into ops at America's companies. And there's going to be white collar, there's going be pressure on white collar labor, ⁓ the white collar labor job market, you know, eventually the pie gets bigger, right? But in the meantime, it's a bumpy ride, you know, with new technology, you know, adoption. And so. Brad Bone (09:22) Thank you. ⁓ Craig (09:30) Yeah. Jack BeVier (09:33) I think there's going to be a 20, 20, I think the story of 20, 26 and 20, 27 is going to be AI's pressure on the labor supply or on the labor market. And that'll, you know, decrease consumer confidence, which will be bad for, you know, bad for the economy, know, the GDP and, you know, stock prices probably take off. Brad Bone (09:42) All right. Jack BeVier (09:58) because profits will be up because they're not spending as much on people and they're getting efficiencies of this new technology. So the stock market will rip, but Main Street will be upset, not happy about it and there's gonna be pressure on that. And consumer confidence and consumer spending will be down in 26 and 27. That will lead the, and in the meantime, the short end of the curve has gotten Brad Bone (10:05) Life. Life. Thank you. Thank Thank Jack BeVier (10:24) you know, below even the two year. And then, you know, in the five-ish year range, so people won't know how long this is going to take, you know, to like play itself out, but they're going to be like, well, it a year or two. This is like a five, seven year readjustment that the economy is going to have to absorb AI and create new companies to expand the job pie to, you know, to make the economy bigger. In the meantime, we've got this like maybe five years of disruption. So, I think you see the yield curve from like the one month to maybe five, maybe seven years, let's call it to the five year sitting low. But then we still have these structural issues in the economy about overspending deficits. Are we going to need to print our way out of the deficit? And so why would I want to invest in the 30 year? you know, like this. So and so I'm to demand a premium for those longer duration loans and a bigger premium than exists right now. So like right now, I think the difference between I looked it up the other day, the difference between like the two year and the 30 years, like 120 basis points or something like I think that's going to be a wider gap. You know, like the premium for two plus is going to grow. It's going to expand because we haven't fixed any of these long term issues. Brad Bone (11:19) Thank you. Thank Craig (11:34) Right. Jack BeVier (11:45) But in the near term, we need to like pump the economy to things moving along. So I could see a refi opportunity is my point. I can see a refi opportunity in a year or two because the five year gets pushed down because of the labor market challenges. So anyway, that's my hot take complete crystal ball, just me talking shit prediction. Brad Bone (11:47) Thank Thank you. Craig (12:07) Well, you know how we love to time capsule all predictions. So producer Kyle, make sure we hot mark that point of this episode so we can come back to. Brad Bone (12:15) I would think, Jack, that scenario could actually push the housing market up a little though, right? Unless the economy's so bad that everyone's losing their job. But if you have rates come down and there's somewhat of a recession, housing turns around, right? Jack BeVier (12:32) Yeah, I agree with that. Yeah, I agree with that. ⁓ Yeah, depending on, I guess, depending on what the like right now the DSCR loans still trade off the five year, right? Like even though they're 30 year fixed rate loans, on average, they pay off or they have in the past paid off over five years. So they price us off the five year because they think it's a five year loan that they're making us even, you know, on an investor by investor basis, you know, like I ain't never refinancing, you know, a lot of our, you a lot of the stuff We're just never going to touch it. And so I wonder if there'll be an opportunity where the market does still prices us all off of the five year, but I'm like, no, I know I'm going to keep this thing for 20 and that, and if the premium between the five year and the 30 is like significant, it may be like, I may be like a buyer of debt, right? Like, it may be like, Hey, I want to borrow that money that's being priced off of the five year. Cause I know that I'm getting a good deal there. Cause I'm to keep this asset for 20 or 30 years. Brad Bone (13:26) Yeah. No, that's it. Yeah. Good take. I like it. Craig (13:27) Well, yeah, go ahead, Brad. I don't know if you had anything. Jack BeVier (13:32) Anyway, yeah, we'll see. We'll see. Craig (13:33) Well, always a good conversation. ⁓ I'd encourage folks to definitely read more about sort of what happened last year versus what's happening right now in terms of the five-year. And keep your eye on that rather than Fed rate cuts in terms of DSCR. So let me bring in our guest officially today. We've got Brad Bone on the line today. He's co-founder of Prime Home Buyers. They specialize in off-market single-family real estate in Jacksonville. They have over, well, it says nearly a decade of experience. Don't know when you wrote that, And he's also the co-founder of River XAI, which I think Jack and I are most interested in talking to you about today, Brad. ⁓ River XAI is an AI-powered platform specifically designed for real estate investors, and it helps you streamline lead management, accelerate the follow-up processes, and it's all used ⁓ with cutting-edge AI engagement. So Brad, welcome to the show. Really looking forward to hearing more about both companies. Why don't you just give us a brief history on sort of how you got into real estate and what you've been doing from a real estate investing side and then love to hear all about River X. Brad Bone (14:46) Yeah. So it's actually exactly 10 years, ⁓ just over 10 years that, that I started in real estate investing. ⁓ I live in California. So I started in California buying at the courthouse steps. And then if, if you're a California investor, only California investors would kind of understand this, but you're constantly trying to figure out how to get out of California and invest somewhere else just because rent laws, tenant laws, those kinds of things. So we were always looking at a state. We were actually buying houses out of state or way early on and doing some flips. Those worked out okay, not great, but eventually we landed on Jacksonville, Florida in 2018, started marketing there slowly doing some deals there. And then over time we got to where we actually liked that market better than our own market. Knew more people there, knew the city, almost actually know the city better than our own city. And so that's where we're at now. So we're still virtual. Still live in California, but our market is Jacksonville, Florida. Craig (15:45) So what are you focusing primarily on now? Are you doing fix and flips, buy and holds, wholesaling? What are you doing? Brad Bone (15:52) We're wholesaling and doing innovations now. We, we've done a lot of flips there. and we got a little bit squeezed when the, when the rates went up a years ago. that forced us to kind of shift our model. We wanted a model that was, ⁓ a faster cash conversion cycle. So that, that got us to, to focus more on those extra strategies that don't require you to close and you can go straight to, to, find a buyer. So that's what we've done for the past about two years. And that's been a better. Craig (16:17) That's, know, Jack and I, think both agree that of sort of all the tiers of investing that wholesaling, you know, lends itself to virtual. ⁓ I don't know that I would ever do virtual flips or ⁓ I guess you could do virtual landlording. ⁓ you know, how have you found, was, Jack, we were just at an event a couple of nights ago, Maryland Investor Network, where one of the largest wholesalers here in Maryland, ever market pro home buyers. He talking about going into very into different markets, Brad, and talk about like sort of the transition. You're literally all the way across the country. You decide to go into a new market. It's not there's no lack of competition in Jacksonville. So what's that been like? Brad Bone (17:03) ⁓ you know, I don't know that I'd it again, but when we, when we did it in 2018, we were pretty small and we're just like, sending direct mail and just starting, you know, and we had ever since we started, we had learned how to actually back then we were buying over the phone. We're buying everything over the phone and then COVID came and that became the thing to do for a while. And then sometime after COVID, we switched to in-person. So we were kind of set up to buy virtually anyway from the beginning. Even when we were in Bakersfield, we were buying over the phone. ⁓ we just transitioned into it by accident in a way. And then over time, you're just like, well, this seems to be a little easier than even doing deals in our own city. Craig (17:48) How many transactions are you doing a year in Jacksonville? Brad Bone (17:51) We'll probably do about 75 to 80 this year. Yeah. Yeah. Craig (17:55) That's That's a marketing machine. Jack BeVier (17:58) That's been a, I mean, Jack's from 2018 to 2022 was ton of fun, right? Like that was, that was a big, a big run-up market there with all the private equity groups and hedge funds buying their public research in there. Were you like, were you trying to serve that market or were you just buying anything and sometimes you would end up serving that market? Brad Bone (18:06) Yeah, yeah. Right. Yeah, we didn't go all in on the hedge funds. We were selling to them. feel like, you know, back then we were still not as dialed as we should have been. So I don't feel like we really capitalized on that great market that we had at that time. So, but yeah, we were selling to the hedge funds. We were, we were fixing flipping and, and so, so that worked out. We didn't, we didn't really focus on that necessarily. We were mostly focused on just fixing flipping that whole time. Jack BeVier (18:52) Yeah, gotcha. Cool. So wholesales and novations and Jack still that's the only place where you're investing currently. ⁓ Keep any rentals or just trends. Brad Bone (18:57) Yeah. Yeah. We have kept some rentals, we have some, ⁓ but yeah, mostly just at this point, we're just focused on cash conversion cycles, turning the money fast and building that machine. Yeah. Jack BeVier (19:12) Nice, that's awesome. Craig (19:13) What's your marketing stack look like right now? Brad Bone (19:15) So right now it's PPC and TV and I do want to add a third one. It's all inbound though. I like inbound a lot better than outbound. We used to do cold calling, texting, all that kind of stuff and that worked pretty well for a while. But you know, I think everyone and really every investor, most investors anyway, once they can get to inbound, it's just, you know, it's like you get addicted to inbound. So it's expensive, but leads are way better. Jack BeVier (19:39) Thank Yeah. The, so what got you into the, uh, the river X AI, like how, do you make that transition of a tech background? You're like, where'd that, where'd that come? Brad Bone (19:52) Good question. I don't have a tech background. I actually don't even feel like I'm necessarily an AI guy, but what happened is a year and a half ago, I came across wucender, which is an AI, ⁓ AI appointment setting platform and signed up with them. Cause you know, when AI became popular in the, you know, chat GPT came out a couple of years ago. My thought was the biggest impact it can have for our industry is, is re-engaging sellers. Like if we can have AI do that. That's going to, that's going to move the needle lot. So I was looking for solutions around that and it came across WooCenter. I signed up with them, spent the next probably five or six months, just getting the integration right with our CRM. Cause I didn't want to have to be managing two platforms because WooCenter is its own. It can actually be a CRM if you want. So it's got a pretty good backend and platform Salesforce left-made. Yeah. Yep. Craig (20:42) What are you using currently for your CRM run? Okay, got it. Jack BeVier (20:46) Nothing? Yeah, cool. Brad Bone (20:48) So ⁓ connected it with Left Main and got it integrated the right way. And then we started getting a lot of success with it. were, I think even in the first three or four months, we were, we had like 10 deals that came from appointment setting, from AI, re-engaging those leads and setting appointments. So, ⁓ so then I was talking to them, they had kind of a referral program. I was trying to negotiate them on a referral program. They're like, why don't you just resell this? I said, okay, so we did that for a little while and then eventually it turned into more of a partnership. So we're like WooSenders real estate investing partner. ⁓ If someone comes through us, you get a direct license through them. We're not like a middleman anymore, but we provide support and integration, custom templates for real estate investors. Jack BeVier (21:33) So you designed it all. Brad Bone (21:34) We, we designed the build out for real estate investors is probably the way to put it. Yeah. The actual platform itself is Wu center. They spent millions building that, the kind of the application for real estate investors, we built that up. Yeah. Jack BeVier (21:38) Yeah. Craig (21:48) When you say the application, tell us about that. What's the flow? What does a real estate investor get if they turn on your version of WooSender or what's known as RiverXAI? What do they see? What do get? Brad Bone (22:03) Yeah. So it's really not, ⁓ it's not like a separate application. It's still a woosender account, but what you get is our templates that we've created. We copy and paste those right into your account. So now you have out of the box, you know, ⁓ followup sequences that work for our industry, ⁓ for, know, whether it's drips, unresponsive, long-term followups, agent outreach, templates, things like that. And then we've also trained the AI to respond the way real estate investors would want us, would want the AI to respond. So we've got AI libraries trained for the industry that we copy and paste that into your account as well. So you come with some pre loaded sequences and training for the AI so that the learning curves a lot faster. And then beyond that, we will go ahead and integrate with your CRM. And that's always a custom integration because everyone has different CRMs and everyone has it built a little differently. That's usually done through Zapier or an API. But if you've got good Zapier or good integration capability, we'll make that work. Craig (23:07) Well, before we jump in honestly to all the sort of the technicalities of it, I just thought that like Jack is obviously and I have been playing a lot. several months Dominion has spent tremendous. I love it. And I have zero programming experience. mean, none. My father told me to get into computers in 1985, and that didn't work out so well. So when you sat down and you started to have this idea of how do we automate follow up, which is essential for every wholesaler, when you had that idea and then you had Chat GPT on your computer, how did you then say to yourself, wait a minute here, we've got WooSender. How did you train yourself to train the AI to make it good for you? Did you hire that out? Did you? Brad Bone (23:54) Yeah, I, so it was a lot of work with wucender and, and, ⁓ and my Salesforce team and, and then me trying to explain what I wanted it to look like. And so I I'm really not, I'm, I'm technical enough to be dangerous, but I'm not the guy you want like building a software necessarily. So it's really, it was, it was like talking to woosender and talking to our integration team and saying, is what, when, when, when I move a lead from this stage to this stage, it needs to move, create that contact, move it into the right sequence. And there's a lot to think about there. Like, you don't want to be accidentally, ⁓ reaching out to a seller after you're already under contract or, you know, cause like, Craig (24:38) Yeah. Brad Bone (24:40) the maybe the CRM and the center aren't talking to each other and one knows something the other doesn't. There's a lot that has to be coordinated between that. And it was all sort of in my mind and it was just, it took a long time just to kind of get it figured out, then get it tested and then document it, you know, so. Craig (24:54) Yeah, a perfect scenario to what you just said there was ⁓ you've got this lead in your system. And for whatever reason, and they're on an automated drip. And for whatever reason, they call you on day 17 and you speak with them, but you don't, you don't, but the automation doesn't turn off. And then the next day it says, Hey, we haven't heard from you in six weeks. And now you look like a total dork. And so, you know, I could see Jack, I know that That's something that you have to be thinking about as we integrate AI with our sales force as well. ⁓ So how did you overcome that? Brad Bone (25:28) It was, it was, you know, was a lot of testing. ⁓ so it was testing on our own, our own system and our own, you know, in our own business. So you, and, and I think it's thinking about, like I sat down and I was thinking about what are, what are stages in the CRM that, and really this is based on, this is triggered on stage change. That's probably the easiest way to do it. So what are stages that I would move a lead to where I want? sender to not send a text either permanently or temporarily and things like in negotiation, ⁓ you know, closed or one, obviously those are permanent stops. ⁓ And you go, that's maybe the first thing you try to think about. are, what are they going to be the pauses and what are going to be the triggers to push things in? And then once the communication coming back from sender. And there's really only three things. It's like, here's an appointment. This guy says he doesn't want to sell. or this guy wants a call right now. Anything else, just let the AI continue to work it until it gets one of those three responses. Yeah. Jack BeVier (26:29) What is the ⁓ AI ⁓ aspect of it? The ⁓ conversations are more dynamic. By the way, do you present the follow-up as an AI or as a human? ⁓ Is it say, this is Joe's, Joe the homebuyer's AI, we wanted to re-engage with you? Or do you say, hey, it's Tony? Brad Bone (26:54) Yeah, so it depends if it's an outbound voice call, it says this is a this is a this is AI assistant. This is an AI assistant. And honestly, it's probably best to do that because no AI voice is ⁓ AI voice can't compete with human beings. I don't know that it ever can. Even if it ever does. Jack BeVier (27:12) Yeah, they're getting better. Craig (27:13) It's going pretty good. Brad Bone (27:16) This is my opinion. If we get so good with AI voice that we're able to convince grandma to sell her house, I'm not sure we want that. Because now grandma, a year from now realizes the trust she built was with the machine, not with the human being. So we're running kind of inauthentic businesses. So I think we keep the trust building and the rapport building conversations to human beings and let AI do all the heavy lifting, like engaging sellers. Jack BeVier (27:31) Yeah. Brad Bone (27:41) That's what I think as we go forward, we gotta like walk that line correctly. Jack BeVier (27:47) So how does the AI like, I guess it's been ⁓ prompted to know what its role is, know what the context is, know what its goal is. The goal is re-engagement and handoff to a salesperson? Brad Bone (28:01) Yeah, yeah, for us it's re-engagement and handoff to lead management team is what we do. ⁓ But it uses text, email, outbound voice calls and ringless voicemails. Text and email is just presenting as the assignee. Hey, this is Tony, wondering if you're still interested in selling. The AI voice call is gonna say I'm an AI voice assistant. But all of these sequences are really built as like follow-up automations first. You know, it's, like it, it could replace your follow-up automation that you have in a CRM. And then once the seller responds, that's really when the AI starts to go back and forth and try to try to set an appointment. So with the, you know, I think this is something that WooCenter got right. They thought about, ⁓ if I'm going to set appointments, you first have to have a really good. Craig (28:41) interest. Brad Bone (28:50) engagement process, which is going to involve ⁓ sequences and drips and automations, just like any good CRM should have. But you got to get someone to be, you got to engage the prospect first before you can set the appointment. So they built that out really well. Craig (29:05) And it's all completely dynamic. it's, you know, it's, Jack, I understand we were talking about 1212 maple. It's, it's very dynamic. Brad Bone (29:13) Yeah, yeah, it's dynamic. I think the other thing that, you know, you don't think about is, is okay, it's four channels, it's text, email, voice calls and ringless voicemail, but they're not just independent channels. So if someone responds through text, the email part of woosender knows that. So the email is not going to be like firing emails at you. It's going to actually respond with a confirmation email. So, you know, Jack BeVier (29:30) Yeah. Brad Bone (29:38) There are some AI platforms that, yeah, we're multi-channel, but yeah, but it's not a collaborative process. So you gotta turn all four of them off if someone responds. Craig (29:47) That's brilliant. Jack BeVier (29:48) That's really interesting. what has been the, so you're, you started off with, say someone wants to, you know, wants to try it out. You dump in your pipeline, right? Like your historical pipeline from wherever you are. Right. Like is it, it doesn't matter. I guess your point is it doesn't matter where they're connecting from as long as they're connecting. like they can connect from Podio or another CRM, as long as there's an API out, as long as you're striking data. Brad Bone (30:13) Yeah, right. So if you have a CRM that has good integration capability, we can connect it through integration. ⁓ But you make a good point. So when you think about it, when you integrate with the CRM and it's going to be based on stage changes, that's only going to keep things organized from today forward. So what about all our leads that are sitting in the CRM from the past five years? What we do with those is we download those and dump them in as a list into WooCenter. So that's a manual upload, like a one-time upload. So now you've got all your leads in WooCenter, and then they start working. And you've got to be a little careful. If you have really old leads, you might want to work those as a cold list. So it's a slightly different campaign. Yeah. Jack BeVier (31:01) Yeah. Like, so the structured, like if you're working your leads in a way that are currently in an unstructured way, they pull them into here in an unstructured way, right? Like that's one of the problems about AI is that like, you have to be able to label the data for it to use the, you know, for it to have knowledge of that, like, Hey, we've reached out to this person six times in the past. And here was the situation from whenever, right? Like there's like notes in the system on a particular old lead. Brad Bone (31:14) Mm-hmm. Jack BeVier (31:28) you upload them via this list and it'll know that, right? The AI will be able to reference back. Brad Bone (31:32) ⁓ what we do, what we do is, is something like if you have leads that are maybe older than six months old, and there could be thousands going back, who knows how long we'll download those and, really treat those as a cold list, just like you would if you were to download a, any, any list that you buy. So we're going to treat, we're going to assume that they don't really know us unless you feel like, no, I've been, I've been reaching out. They should know who we are. All right, maybe we'll go back a year. or whatever the client thinks is appropriate. ⁓ So those are just like all lumped into one big group called cold list. That's run through an outbound campaign. The more newer stuff, we might break that out by what stage are they sitting in? Are they sitting in a long-term follow-up stage, regular follow-up, unresponsive? And those can go into different sequences based on kind of where they're sitting. And then as they respond, just get, you know, let's say an appointment gets set. Okay. It's going to update CRM and we're back to, you know, back to being up to date. Craig (32:30) ⁓ So, updating the CRM. So, let's say you get a list. ⁓ Does the AI go straight to the list or do you have a live person handle what would be like? Let's say you bought a list today. Is a live person making the first try or is the AI making the first try? Brad Bone (32:54) the AI is. So it's actually an upload into AI into the AI platform. And then you can, you can turn the dial. All right. How many, how many of these leads do I want to dump into the system on the first day? We might start at it like 20, just to, just to make sure deliverability works. And we're not getting a bunch of people that are upset. If you get a bunch of people that are saying, I don't know you. Okay. We might not have to, we might have to adjust some things. So yeah, it's always AI making that first. Craig (33:22) So AI makes a phone call and gets the person on the line. Is that call recorded? ⁓ Does that then get sort of transcribed into Salesforce for you? How does that lead then sort of get ⁓ tagged or categorized ⁓ for follow up and the type of follow up? Brad Bone (33:40) Yeah. So it does get recorded inside WooCenter. WooCenter actually has its own phone system that it uses. So it's not integrating with your phone system. So we've got Salesforce phone system and then we got WooCenter. We have had clients ask, and I was wondering the same thing, can we move all those conversations into WooCenter somehow? Can we integrate and move all that over? The answer is no, because it's a lot of data. And the other thing is, The truth is those conversations are really not relevant to the sales process. A lot of those conversations are just about when can we meet? Because that's what the AI is trying to figure out is when, when's a good time to have a, have a phone conversation. So they don't move over, but I feel like it's not going to be useful information anyway. And it's just a lot of data to try to move around for no reason. Jack BeVier (34:31) How ⁓ dynamic is the conversation? How colorful can the conversation get? If they're like, sorry, I can't talk today. I'm going to the game tonight to watch the Lions lose to the Ravens on Monday. ⁓ Will it engage on that? Or it'll be like, OK, cool, sounds good. I'll call you after. Brad Bone (34:51) Yeah, it doesn't, it's not like chat GPT. If you were to ever have a conversation with chat GPT where it can just go back and forth. And the reason is it's a selective language model. ⁓ it's not a large language model, so it's not built on one of those LLMs. ⁓ so it's really built to be pretty narrowly focused straight, just straight at getting to the appointment. ⁓ and you know, I think, I think the problem with using an LLM is unless it's really dialed in, but the problem is. It can go off on tangents. can start talking about random stuff. We just want this thing to set appointments. And, and so that's, that's, that's the business use for it. The army uses WooSender and I think there's, there's some, they've got 56,000 users across the world that use it. And I think that's the reason is just, we just want this thing to set appointments. It's not necessarily going to impress us with its knowledge of history, but it's going to set appointments. That's what we want. Craig (35:42) Yeah, it's interesting. My wife and I had to have a company come out to service our HVAC system a few weeks ago. And, ⁓ you know, I called the the number. They said somebody to get right back to you. And it was great. mean, you know, hello. It sounded like there were people in the background, like other people in cubicles talking. It took me at least five minutes to figure out I was talking to A.I. And when I really think about it, Brad, it was You're right. was very it was great. mean it was stunningly good. But it was very narrow as I think about it now. I could if I had gone off on a tangent I'm not if I wasn't following basically the prompts that it was giving me I wonder you know where things would have gone with it at that point. So that is interesting. Brad Bone (36:30) Yeah, if it was built on this. Craig (36:30) especially in the sense that they like the the ruse that they were pulling with the know the other the murmuring in the background of other AI agents talking as well. Brad Bone (36:36) Right. Yeah. Yeah. If, if they were built on an LLM, it might've been able to continue to engage with you. But, um, but again, that that's, you know, one thing I noticed about the, of the voice AIs that are built on LLMs, they can kind of have some really long conversations and sort of talk kind of too much because they don't have this, this ability with human beings to just sort of like, say it quick and get, get on with it. I don't know what, what it is about AI that kind of does that. So, yeah. Craig (37:08) Jack, ⁓ do ⁓ you think of how all of this would fit in with what you do at Dominion Properties and what we do at Dominion Financial, ultimately? What's your take? Jack BeVier (37:21) Yeah, no, think it's really interesting. We've been, I've been looking, I'm trying to figure out like, you know, what are the use cases that make sense for AI? Cause like, you know, not everything's worth the squeeze of getting your data in a structured format. And like, you know, depending on, you know, as Braav was saying, like how, you know, broad of a knowledge base it needs to come from, it may not be like the right, the right use case for this. So sorry, there's, I'm in Baltimore and there's an ambulance going by cause Baltimore. Craig (37:47) the Baltimore National Anthem. We're good. Jack BeVier (37:50) So anyway, I'm interested, I think it makes a ton of sense that it is a specialist, right? Like we've, a lesson I think that we're like taken to heart at this point is in the AI tools that we've tried to build for ourselves, is that trying to get it to do too much doesn't work. But if you just teach it how to do a thing, it'll, and if you narrow the focus or give it a specialized task, then it's really good, like really, really good at that. ⁓ And then if you want to like broaden the complexity of what you want it to accomplish, well, don't try to teach one model how to do more, just create more models that can do specialized things and then figure out how to orchestrate those models to work together. And that's been really effective. That's been... ⁓ Brad Bone (38:18) Yeah. Jack BeVier (38:41) Well, and that gets, that gets work done at human level of accuracy, right? Or human level of like, ⁓ don't know, quality. but if you create these like broader tasks, it just starts to be like, not that good. And people are then dismissive of it, right? Like if they can't do, you know, the thing as well as I can, well, then, you know, it's, it's easy to be dismissive of it, but it can do narrow things. Very, very. Craig (39:07) So do you see, ⁓ would you see something like this as beneficial with Dominion properties or is it something that's not even on your radar? Jack BeVier (39:16) Oh, no, very much. Yeah, very much in the radar. We we had worked with a company that was doing AI. I mean, I don't know if they're using a platform. Maybe they're just using your platform and just little man in the thing. But this was like maybe like six months ago, maybe eight months ago. And it was it was the same idea. was like, hey, upload a cold list and then just start working that list and. Craig (39:24) I was wondering the thing. Jack BeVier (39:38) You know, the AI will have conversations with that folks. And once they get the conversation to a place where there may be like something, you know, a business to talk about, tap into human. And, that was the, that was the approach. ⁓ so I, ⁓ and I'm interested to see like just more and more real estate and property management use cases. Like we've come up with a bunch of stuff for ourselves and the lending side of things, but I've just really started to scratch the surface of the real estate and property management side. Brad Bone (39:53) Thanks. Jack BeVier (40:05) some of the property management platforms, you know, because they were already in a structured database, they're adding their own AI on, right? They're not, they're like trying to not, you know, you say like, Hey, can I access your API? And they're like, you know, don't worry about it. We're going to build that, you know, we're building that feature, you know, like they want to keep you in the system, right? They don't want to have the data leave the systems that you can, so that we all like create our tools externally, even though I think that that, you know, it's going to I think we're going ⁓ to get there. the unstructured databases are where I think that plugging into custom-made AI tools are going to result in lot of productivity, a lot of unique custom tools. My opinion is that everybody just needs to lean into learning how to. Brad Bone (40:34) So. Bye. Jack BeVier (40:58) create their own AI tools for themselves. And if they do, you'll like have this huge productivity increase. But I think it's gonna take some years for everyone to learn. So this is a very cool use case. Craig (41:11) So, Brad, how has it changed your business? Brad Bone (41:13) It really what it's done is it's ⁓ well, first of all, we get about three extra deals a month ⁓ from AI appointment setting. So, you we can trace it back to AI, engage that person, set the appointment. We got on a call, went out to the house and got the contract. about three a month and it's been very consistent over past 18 months. Um, so that's just from a ROI perspective. I think what it really does for our team though, is it makes them a lot more efficient. So we still have lead managers. It's not going to completely replace your lead management team, but they, they can a lot more of their time is spent talking to people that have raised their hand, want to sell. So they're having more conversations and that heavy lifting of just trying to get people on the phone. We still have them cold call. I think that's always, you know, should be part of the culture, but, uh, they just spend more time. having good conversations. Craig (42:04) And so three a month, how long have you been using it? Brad Bone (42:07) About 18 months. Craig (42:08) Is there a reason why you think it's ⁓ three is not insignificant. That's thirty six extra deals a year. know. So is there reason why it's three and not five or 10. Brad Bone (42:20) Yeah, you know, we mostly have it working our follow ups. ⁓ going back into our database and working those. If we were to, and we haven't done a ton of this, but we've done a little bit. If we were to turn it loose on an outbound list, we have agent outreach campaign set up. If we were to turn it loose on agent outreach, mean, yeah, you can scale it from there. So we've had pretty good success with just the inbound campaigns working. work in follow ups, work in unresponsive, work over the weekend, leads that come in, that kind of stuff. Craig (42:53) any crazy hiccups along the way that were, that you wouldn't want to run into again. Brad Bone (42:58) ⁓ yeah, I mean, there's times when it, it says weird things, but we, don't think we've, I wouldn't say we've lost any deals from it. There's, there's, ⁓ times when it'll, it'll schedule an appointment and then say, Hey, ⁓ you know, we need a schedule appointment. like, okay, there's a little glitch there. So there's going to be a little bit of that, but, but at the same time we've had some huge wins from it, you know, like, We had a deal, probably the biggest deal we've ever done back in February is it was over six figures and a lead came in on a Sunday morning, like nine o'clock Sunday morning. AI, the AI worked that lead for the next three days, sent out, think five messages. Finally, the guy respond within seconds at setting the appointment. We went on the appointment, got the deal and it was the biggest deal we've ever done. So that paid for, I don't know, 10 years of wucender or whatever it is. Yeah. Jack BeVier (43:47) Thank Craig (43:48) What is the cost? you share that with us? Brad Bone (43:50) Yeah, so it's about it's it's 2,000 a month ⁓ It's a cost of a virtual assistant and you can you know, if you've got virtual assistance working ⁓ Follow-ups you you know, you can probably get rid of half the team doing that so it's 2,000 a month we often will have a you know setup fee ⁓ You know about two to three thousand dollars depending on the build that you need to do to integrate your you know, your CRM with the ⁓ with the platform. So, but yeah. Craig (44:20) Jack, did you have a chance to check that out? Jack BeVier (44:23) Yeah, I just showed it to me. It very cool. Craig (44:25) I guess I wasn't invited to that. Jack BeVier (44:27) I don't know. I don't know where you got man. ⁓ Craig (44:30) Yeah, so any other questions for you, Jack? Jack BeVier (44:33) No, man, it's really cool. I got the opportunity to check out the platform and exactly what Brad described. it seemed very cool. Loved chatting with somebody who's thinking about applying technology to make our business more efficient. Especially right now, man, every little bit helps right now. Nothing even. Brad Bone (44:50) It does. Craig (44:50) Right. Right. That's it. Brings me to maybe one last question here. You know, you've had this is by all accounts a crazy success. Three deals a month. You know, obviously, you know, making way more than it's costing you, I would assume, in wholesale fees. And so what's next? Like if it's this good now for you, if this one tool is just that great for you now. Like what else is on your radar? Where else are you thinking like, shit, man, I could really see us doing something like this in the future. Brad Bone (45:24) It's a good question. ⁓ Yeah, I think as far as AI, mean? You know, we've just been focused on the appointment setting side. So I think there's some really cool stuff coming out with even like, you know, I heard about some AIs that have to do with. Craig (45:28) Yeah, of course. Brad Bone (45:43) like training your, your, or, monitoring your sales team, even in person, like your in person. And I don't exactly know how that works. I want to look more into it, but like an AI that is able to give a feedback on an in-person sales appointment. Like, I'm not sure what, I don't know if you turn something on on your iPhone and it listens or what, but that seems, especially for me who's virtual, you know, and I can't like do ride-alongs all the time with my acquisition guys. So. That is interesting for sure. Yeah. ⁓ Craig (46:13) we are tasked. Jack BeVier (46:14) I've heard of a, I heard of a custom one that someone created for the, for their sales team where they, they had their call rail hooked up into feeding it, feeding the transcript into. so, you know, the call happened and then the recording of the call got sent to the AI, which, which ⁓ created a transcript and then ran the transcript through, ran the transcript through sales training materials, right? They basically like. you know, aggregated a bunch of sales training for like we buy houses ⁓ use case. And so then the AI analyzed this transcript against all the best practices that were enumerated within the study guides basically for the, for the sales training and then created a report of like, Hey, this was really good. This needs a little bit of missed work. You know, you have the opportunity to use this strategy here and you didn't, you know, consider that for the next time. And it was like one-on-one. So it's like, you know, as if like the person who wrote the sales training is like giving you, it's like reading your transcripts or listening to your calls and then giving your person one-on-one feedback immediately. Like, within like, you know, two minutes after the phone call ended. So, right. Like it's all still fresh in mind for them. And this guy was just like, I will put, I will put a normal person with work ethic with this, you know, with this side-by-side coaching up against the most up against any talented salesperson. And like, you he was just like, it's, it's with the coaching. It makes them the equivalent, if not better than, you know, someone who's just naturally charismatic. And I was like, this is going to be a really interesting AB test. But like, I hear what you're saying. It's a cool use case. Brad Bone (47:49) Yeah. Yeah. I think that stuff is huge. When you think about like, what are the pieces of our business that are the hardest, like most critical and the hardest to manage. it's, it's often managing your sales team. Like those guys are so critical and getting more appointments. Like we can have those two then man, 80 % of the business is solved. And then if we can get it to get us cheaper leads, that'd be great too. That are high quality. So. Craig (48:17) The leads, think, as you start to implement the better models, your conversions obviously increase exponent, I would think at least logarithmically and the cost of the lead should go down commensurably at that point. I will tell you that one of the things that I find exciting is the video and audio aspect of AI. If you could, like, if I send out a quote to somebody for a loan, DSCR loan, which is a very competitive market of lenders. What if it what if it was literally a ⁓ video with audio of me in an email and it just click this link and now I'm talking to the person as if I'm there, I'm using their name and it's got the quote right next to it. Maybe even a picture of the house because we pulled that from the Internet somewhere and we now now we're speaking to this person at a far more personal level than hey, I just sent you a quote just like That's interesting. And I don't think that that I think that that could be a year away. The video technology is very close now. ⁓ The voice with 11 Labs is very close. It doesn't even have to be my voice, frankly. But I think something like that would be really exciting, especially if you couple with them with the sort of follow up that your company is doing. Brad Bone (49:16) Thank Right, right. I heard about a company or a guy who it was actually at a mastermind I was at recently. This guy created an avatar of well, it's a real person. I mean, it looks like a person. It's not a real person, but it looks like it's a girl. It's a female and she's the CEO. I think she's the CEO and her name is like Alexis or something like that. And she's the face of the company now. And everyone thinks that she's the face. Craig (49:42) Right. it is. Ha Brad Bone (49:54) And so he's like, I got to talk to Alexis, you know, and see if I can buy the house that cheap. I mean, he's running with it. yeah, so I mean, there's stuff like that. It could be interesting. Jack BeVier (49:58) Yeah. Craig (50:02) Playing good cop, bad cop with AI. Jack BeVier (50:04) Yeah, Couldn't get you to bring her around. That's hilarious. That's hilarious. Craig (50:13) Yeah, man, it's a wild new frontier we're living in. Jack BeVier (50:16) Hey, so if folks want to get in touch with you, want to learn more about this, ⁓ how do they do that? Brad Bone (50:22) Yeah, they can either Facebook, um, Brad, but I think it's well, Brad bone. We're probably, we probably all have all the same friends. If you're a, if you're a real estate investor, um, or you can go to river X AI.com and you can book a book, demo through there and you know, can have a 15, 30 minute conversation. Jack BeVier (50:31) real estate investor here. Craig (50:41) Yeah, riverxai.com ⁓ for folks who want to reach out. Fascinating. Kent, thank you enough for taking this time on a Friday afternoon with us. Next time you see us, I'll dress up for you. seriously, man, it's been awesome. What a cool business, man. All right, well, go Ravens. Folks, we thank you for tuning in today. Jack and Craig with Real Investor Radio. see you on the next one. Brad Bone (50:56) All good. I liked your wardrobe.