Why listen to the past when you can listen to the future church? Welcome to the think future podcast broadcasting from deep in the heart of Silicon Valley, California. We focus on innovation startups in the future, not necessarily those and not necessarily in that order. Here's your host.
Yeah, yeah.
All right. Sounds good. Thank you so much for coming on the show. It's great to finally meet you. Because this is how we meet each other nowadays, right? So why don't you tell us a little bit about yourself and your organization and what you're passionate about?
Yeah, so my name is Nate Jones. I am the co founder of structurally, I was CEO, we brought in a new CEO about a year and a half ago. So I am now Product Marketing Manager, which is where kind of my love of product and marketing combines nicely. So it's where I'm, that's where I'm focused today, structurally, has been around for about five years now. We're really focused on qualifying sales leads mostly for real estate, mortgage, prop tech, also insurance in kind of those adjacent industries, I guess, if you will, that have a lot of leads, lots of them can be kind of tire kickers. So that means a lot of heavy lift for salespeople to sift through them. So that's where we focus and we do that with conversational AI.
And, well, okay, that sounds really rife with it with problems. Can you tell me how have you solved? How have you solved the AI? AI issue with, like, applying AI to, to sales? It just, I mean, sounds like a great idea. But how does it actually work?
Yeah, yeah, it is a tough problem to solve, I would say. I think probably everyone who would be listening to this has probably heard the phrase chatbot or something like that conversational AI, or just AI in generally, in general, it's kind of become a buzz, buzzword. So everything is AI this and AI that. And it's hard to decipher kind of what that actually means. And that's a good thing. And a bad thing. I think there's a lot of AI that does a lot of different things. There's computer vision, helps self driving cars, there's predictive analytics, that helps predict the price of a property, things like that, that's all AI, per se, in different ways. We focus on conversational AI, some people kind of call it chatbots. But there's lots of even aspects of conversational AI that don't necessarily focus exclusively on chatbots. But that's really, really where we focus in terms of like, at least the problem space. So the way that we work that's quite a bit different than traditional chat bots, I'll say is we listen to all of the leads that come into your CRM or lead gen, or marketing automation software, they typically have a phone or an email. Otherwise, they're pretty anonymous, and you have no way to talk to them. And we reach out to them immediately. If they're new, or, you know, they're, again, immediately if they're old to try and bring them back to life over text and email, we follow up with them for the long the long term, until they actually reply. Then at that point, we have a two way conversation to try and qualify them set an appointment, make sure they're actually worthwhile for your team to go after and off to the races from there.
So it sounds like you're not just the Chatbot section. Because right now, when you experience chatbots, it's like you go to a website, you call up a chat, you're chatting with a chatbot there, but it sounds like what this bot is doing is sort of beyond that, that that's only a small portion of the overall experience. Right?
Yeah, I think that problem, so unfortunately, and fortunately, because there's a lot of really strong companies in that space, drift, intercom, even, you know, tons of big companies that power all of those little live chats that you see in the bottom right corner of about every website these days, they're they're solving quite a bit different problem, I would say they're trying to solve anonymous web traffic going to your website that might not want to fill out a form or contact you because you know, that can be a tough experience, especially because historically, you'd fill out a form and hear back from nobody, which is essentially the same problem we are solving in a different way. They're trying to say we want to capture their intent on the site, turn them into a lead right there through the chat, and hopefully get a sales rep like on a call. Lots of things happening really fast. I think there's there's definitely a time and place for that. And I think that's actually a good combination to use with our product. Some people do want to fill out the form some People don't. Some people don't want to talk to a chatbot. Because they've had a lot of bad experiences like that with the live chat, chat bot that feels really robotic. And they come out and say, Hey, I'm a bot, you know, how can I help you on this chat? And then you say something, and it's like, I'm sorry, I don't understand. Let's start over. That's tough. That is not what we do our products. I
was like most of my recent conversations. Like, oh, sorry, I can't answer that question. I can't answer that questions. Let me connect you with somebody. It's like, what can you answer?
So that's where we're like, we are solving a similar problem in a different way. But we take a much different angle we set out with structurally to say our product is going to be indistinguishable from human. There's going to be no I'm sorry, I don't understand. I'm a robot, please start over. I can't answer that question. We've had millions and millions of conversations and 99.9% of our customer, our leads believe they are talking to a human. And that's down to a very granular level where we can understand when people are saying things like, Hey, are you a robot? What are you? This is a weird conversation. That happens point 1% of the time we've seen it. And we typically adjust for it. We say things like, Oh, I'm just the digital assistant here. Sorry for some of the weird questions, I'll get my agent to follow up with you so and other things like we put purposeful typos that we correct a message later in our messages. We have We empathize, especially in big purchase. Big purchase kind of items like real estate, insurance or mortgage. The last thing that someone wants to feel like is they're talking to a bot. And that you're typically moving for a very emotional experience in the industry, property tech industry, we like to say the three Ds of real estate death, divorce and disease. Those are big life events that a lot of people move, because of, if we just if someone said, because people will tell us our life stories, if they say, you know, hey, so and so is dying, we have to move because whatever reason we can't just say, got it. And how many beds Do you want? We want to say yeah,
yeah, Matt, if you want to sound like a human, right,
so we take a lot of pride in making our conversations. Extremely lifelike, extremely flexible, so that people can talk the way that they've always talked over messaging.
So how do you so I'm assuming you trained this bot, to be to act more human? How do you how do you train it to be more human deed? Do you get agents? Does it does it look at a corpus of agent communications? Or how do you make it? How do you get it to be like that? Yeah,
really early instructionally. We saw I'd say there's two parts of that. One, we developed a model conversational architecture, it's not necessarily the machine learning side itself. It's a lot of it's a lot more kind of algorithmic decisions, like, you know, in a certain point of a conversation, do we want to ask this question or not? Do we want to switch around in a conversation, it's the engine that makes it super flexible. The analogy we like to kind of use is most chatbots are like a train, or on the track, if you say something off script, there's going to be a train wreck. Ours is like an ATV, where most of the time you're going down the trail. But if someone says something off script, we'll go off trail, but we're going to try and bring them back to the end or back on on track here eventually. So that engine was kind of the start of the whole product for us. So that it's not like super unique. There's a whole concept in machine learning and conversational ai, ai of slots and intents and contexts. We just put our own spin on it to make it really flexible. So that laid the groundwork for us to actually be able to tag data in a really certain way. So early on, we kind of had those two things coming to fruition, but we didn't have a lot of data from conversations. So we went out to market and we had conversations, the same type of product that we have today with our customers, but we just did so with humans, we had a human team that took over the conversations had a number of 1000s of them, then we labeled them in a certain way that plugged into that model. And then we had a nice, really accurate data set to start with. That got us a long way. And then that just compounded and compounded and now we have more of that hand labelled data today than we know what to do with honestly. So we're, we're not short on on the the labeled data, but that's kind of how we got there. We were really short on it early early on, but we seated ourselves with kind Have a human have a very high emphasis on the human side of the human in the loop problem. We still have a human in the loop today where they're still tagging that data. But it's, it's become a lot more like they're tagging the really weird conversations that like I've we've never seen before, which happens all the time. Yeah, that's what lawyers. Yep. Yeah,
it's funny how human beings will always surprise you, they will always surprise an AI. Because it's like, Who knows whether you're talking about bubble gum or something like that? It's all over the map. Right? I think that that's the
most interesting stat that I've seen lately is like Google, I think last year, put out some staff that said, like, some some double digit number of their queries that they say that they see every year they've never seen before. That's Google, who gets billions of searches a day. So if you can imagine a little search string being so unique, every time someone puts something into Google, take that times, messages in a conversation, people, I know two messages that structurally, C's are ever the same, pretty much. So that makes the challenge really tough.
Wow, that's, that's amazing. So what makes you different from other? I mean, sounds like the sort of expanded scope where you're not just doing the Chatbot piece is makes you different as well. So I mean, how do you how do you decide what to go for it? Like, does a human come in and go, Okay, here's, here's 300 people that we need to communicate with. So let's start sending me emails, or let's start calling them, let's have a conversation and then draw them back in? How do you determine does the system actually determine who to go after?
No, that's entirely up to our our customers, typically, inside sales agents, or SDRs. Or typically, their managers, they're saying, Okay, we get all these leads, some of them are from Facebook, some of them are from Google, we'll just keep it simple. I want my human team, maybe speaking on behalf of the manager, maybe I want my human team to focus on these really expensive high value Google leads, and they'll just call them. And if they can't get ahold of them, then maybe sending them structurally which we see all the time. In these Facebook leads I want structurally to go after right away, they can actually set up that routing, again, typically through their CRM, because we have tons of integrations with Salesforce, HubSpot, the big CRMs, some more niche, CRMs, and kind of our prop tech vertical, where they can say, you know, if it's from this source, and infrastructurally, if it's from this source, don't, after a certain day that I can't get a hold of them do send it to structurally so they can slice and dice however they want to send the leads to us. And that's on the new lead front. There's a huge opportunity that I think all listeners, if you're in sales, marketing, or product anywhere, you probably are doing business, and you've probably forgotten about your old leads, everyone wants to chase the next new shiny lead, because they're, they're hot, they just filled out a form there on your website, of course, they're probably going to convert higher, and you need to get on top of them and qualify him. But you've probably given up on a lot of leads that shouldn't have been given up on. So we also have the ability for people to send kind of in bulk, big chunks of leads over to us that we kind of bring back to life, we say, Hey, I know you've been on our site in the past, we didn't get a hold of you wanted to see if you're still interested in doing something. And obviously, we see a lot lower conversion rate on those because they're older. But a lot of those leads teams have just given up on they said, Yeah, we tried didn't get a hold of them. It's over. But we bring a lot back to life that they've given up on from that campaign.
No, I totally believe you because it's all about timing. Yeah, right. I mean, those leads, I've had situations where I followed up with someone over years, years and years and years, and then eventually they turned into an actual sale. So it's all about the timing. If you hit that person at the right moment, you should never you know, then you're good. But if you give up on it at any point in time, then that lead could go nowhere. And it's I find it unbelievable, because I hear salespeople who are like, oh, you know, hit him nine times. And then you know, forget about him. And it's like, well, no, no, keep hitting them forever until they tell you I'm Shut up. I don't want to hear from you ever again. You want to get to them. But it's all about timing. Yeah, yeah. And there's probably more gold there too. Because these are these are people who know you. So they have a little bit. They have a little bit of knowledge of who you are. I mean, they might have to sort of search their memory a bit. But you knew at some point they were interested enough to fill out a form or to click a lead or whatever. So those are probably the warmest ones that you can probably go to and then get Can you tell me a little bit about how that outreach works. So say I give you 300 leads, like what do you what do you do with them? Do you start calling them right away? Or do you guys have like a script that's Guaranteed to Work or do you work with the companies to determine what the outreach is gonna be? Yeah, good
question. We don't do any phone calls. So this is all text and email right now. So, in that case, new or old leads, we have a, we have a library of scripts that our customers can pick from. There's tons of them. So they typically fit a lot of use cases. But a lot of our bigger customers, you know, we have some who are fortune 500. Level companies who care a lot about their branding. So they need the exact right thing to be said at the exact right time, because they're a public company, and they're under a spotlight type of thing. So they, all our customers have the ability to completely customize their scripts through our product as well. What we do typically see is they trust us to be the experts, because we can see millions of conversations and say this is working, this isn't. So they'll typically like copy a lot of our scripts, and then just tweak certain things. And that works really well. But there are some others that go completely into a new avenue that we would have never expected. And that's okay, too. They can build to their heart's desire. Really, really flexible. So it is, in our product, it's not a it's not a decision tree, we kind of get in, we kind of get down this debate, technically, with a lot of a lot of folks are like, Oh, it's just a decision tree. If this you go down this path, if this you go down this path if that? No, that's the problem with most chat bots, because if you say something that's not on script all the way over here, you say something over here, it can't make that jump. And it's like, no, wait, hold on, I have to go down this tree, right. So it doesn't work like that conversations don't work like decision trees, although you might want to think they do. Our script builder is a lot more flexible in linear. But that does give a lot more possible. power and flexibility to our users to build because you have to think through every turn of a conversation that could happen. And that's really hard to do. You need to write scripts, to make it really tailored one, a few really tough examples of that are, you know, our product works in Spanish, too. Out of the box, all of our scripts are written in Spanish. So at any point a lead can say, Hey, do you speak Spanish, we'll flip to Spanish in the script. And write that when we've written all our scripts in Spanish to support that. But they can also we typically see that they're, they're buying on behalf of a third party. So in real estate or something, they're like, Oh, I was just looking for my brother in law. Well, if we keep the conversation going, saying, Okay, tell me about how many beds you're looking you're looking for, that doesn't work, because they just said, I'm looking for my brother in law. So we have to change the pronoun to how many beds are they looking for. So that's another version of the script. So it's like, you can see how it can expand pretty quickly in terms of what people can write and what they need to consider to make a really human like conversation. But that's the type of flexibility and care that we have that I think most chat bots have kind of said, That's too hard to figure out. I would rather just come out and say I'm a bot and do this decision tree. And if I screw up a well, it was just the bot and maybe the agent or maybe the lead can say human and talk to human. That's, we don't want that. That kind of defeats the whole purpose.
Right. Exactly, exactly. So we can you tell me a little bit about how that conversation works? Because it sounds like I'm gonna send out an email. And then are they going to respond to my email? And then does the conversation actually happen over email? Or do you drive people to the bot? I mean, how does that conversation actually work?
Yeah, so we, we do send outbound. So as soon as you hit, you know, send lead to structurally from your CRM, or it happens automatically, we will reach out to that lead over text and email. And we reach out to them outbound. And we keep hitting them until they respond. So we do the long term follow up as well, pretty aggressively, because like we talked about, people give up quick, too quickly. And then as soon as they respond whatever channel email or text or both, because they can switch. We have that accounted for as well. We have the conversation with them. And a text conversation is actually not the same as an email conversation and we've scripted accordingly. It doesn't email conversations, you can lump a lot more you can say, hey, tell me what's your timeframe to move? Are you working with an agent? You know, what's your price range, things like that all in one message. So they can just reply back in one Email. Text it's a lot more back and forth. Shorter messages, but in either or channel, they can say all their all of their information at one time. So if they're really ready to go, which we see, they say, Hey, I want to move in the next 30 days, need to put an offer and not working with an agent, I'm approved, let's do this, the conversation is going to not just slow down, it's going to speed up and say thanks for all that. Let's get some time on a calendar. Here when when works for you. So we take a lot of things into account based on the channel in the kind of way that people talk in the conversations as well.
Fantastic. Well, I love the concept. This is great. How did you come up with it? I mean, what was the driver or the trigger? That made you think, think of think of this? Were you having the same same issues with others?
Or? Yeah, so myself and my co founder actually started this right out of college, we decided we wanted to be real estate investors in college, but quickly found, you have to have some money to do that. And in college, you typically don't have any money. So that didn't work. But we ended up this is true. Yeah, so we ended up talking to a lot of real estate agents and lenders and leasing agents, kind of the whole prop tech industry, which is where we started. And basically, follow them, shadow them, ask them what they hated doing what they love doing. And they kept saying they hate lead follow up, they hate it, they're not very good at it, because they give up quickly. I don't know if you look at the DiSC Profile at all, but they're they're kind of high, high D, they're they love to be in person, they're very direct, but the systems behind the scene work is hard. And that's probably true for a lot of salespeople, they want to be face to face, they want to close the deal. They want a qualified lead that just shows up to an appointment to them. So they can so they can sell. They don't want to chase 500 leads and half of them say no, or Screw you and get lost. Because they'll give up really quickly. So we set out to solve that problem. We just didn't want to solve it with a call center. And at the time, this was 2001. Did I graduate 2013 ish? No, not 2013 2016 2017. I would say AI was kind of approaching its hype curve. If you've if whatever that that chart is where
the Gartner Gartner Hype curve? Yeah, it was probably right
at the top, which was awesome. And I think that was warranted because technology had caught up enough to make it useful enough. And we actually found some really, really strong machine learning engineers who actually went to our same school, Iowa State University has one of the top graduate statistic and mathematics department in the country breeds Great Machine Learning engineers. So we picked some up there, convince them to join us. And they built the kind of first iteration of the product. And we actually did a lot of our discovery and demos at conferences. Another great way to kind of get your product out there, prove it out without having a lot of product, because we just had an unworkable demo that we went to some of the biggest real estate conferences in said, like, Would you buy this Do you like it amassed a very large kind of waitlist and launch from there?
Well, I can imagine it being a huge pain point. Because I think that part sucks, too. But I guess that's what you have inside sales and outside sales, right? You've got the, the, you know, dial and dial 100 people a day, you know, folks, and you've got everybody else, right?
Yeah, and that's a lot of the reason why we didn't want to build a call center, we actually had a lot of competitors and still do that are solving the same problem with a in house call center. And basically people can just subscribe to their product and get a get a call center for for them. But what we believe and I think that this podcast is really relevant is because thinking about the future of work and just the future of of technology and how it impacts work is a lot of people are scared of AI scared or they will replace inside sales agents that will replace STRS. It'll replace whatever whoever your job. I think our belief in what we've seen is our job is to augment the role of humans in whatever way we actually started out structurally by saying we are going to replace your inside sales agent. And we got a lot of like, I don't want you to do that. I would rather still have my ISA and you And fast forward to today. Our best customers our best users are actually isa teams themselves. We don't replace them. We make their lives easier because they can take on more leads. Their lives are more enjoyable because they don't have to cold call 500 leads A day and get told, screw you leave me alone, a million times the AI does that and then qualify some and then they can call those leads that are expecting their call. And that just makes their life so much more enjoyable. So I think that our philosophical idea on AI is at least an art. In our case, we want to augment and make people's lives better. And let them focus on what they're really, really good at the more creative, irreplaceable human work, and let the AI take the brunt of the difficult kind of mundane tasks. Phenomenal. I
love it. So every startup has like a moment where they knew they were going to be successful. When did that come for you? When did it come when you went to those conferences, and you realized you had 100,000? Leads? Or what? What lay? At what point? Did you say, oh, yeah, this is definitely gonna fly.
I think that was the first one. The second Yeah, the first one was kind of that building that waitlist and saying, Okay, there's people that want this, they're at least signing up, ready to wait. Then I would say the next one was, you know, we always startups are competitive, some go and some stay. We actually kind of saw one of our biggest competitors, unfortunately, not make it. They were one of those companies that tried to do what we did with a human call center. And they found out that humans are really hard to scale. They're hard to train. Tell me about it. Yeah, they're hard to train, they're inconsistent. And they just don't give you as similar of experience. And that was really hard for them to figure out. And they just, unfortunately, turned and burned through all of their customers, because they couldn't deliver the same experience by staffing up humans across the world, basically, to do call center ask tasks. So that was both sad for them. But kind of like, told you, we can do it with AI moment for us. And it's better. And I would say the last was, we actually uncovered a really unique way to sell our product kind of two or three years ago, where we actually believe that all CRMs and lead gen, or marketing automation companies will have conversational AI a lot already do. The problem that they face is actually the exact same that we solve for our, we'll call it direct customers, our SMB customers, they, these other tech companies, you know, generate leads for you or manage your leads for you. And then what happens is, their customers who could also be our customers turn around and say, like, your leads are bad, you gave me a bad lead, your software didn't convert the lead? Well, you probably just gave up on the lead, you didn't do your job you gave up after the third attempt. And they feel that these other tech companies feel that at a really high level, like tons, they're generating hundreds of 1000s of leads a month. And all of their customers like these are all bad. So we've actually, we actually found that we could white label our product, and sell it into these other CRM and lead gen companies. And then they would act as a reseller of the product directly in theirs to what would end up being our customers anyway, just a lot quicker. So I think that model has worked really, really well for us. And that was kind of like, we could see this growing really quickly through this one to many kind of distribution sales approach.
Well, I can see that the other company that you were talking about that was using humans, I mean, they're probably look at it is, well, you know, it's cheap, we'll just get some offshore folks and train them up, and then they'll do it. And but yours your model sounds like it would be really expensive. How do you how do you keep the costs down on on like, comparatively?
Yeah, it's, I mean, you can you can look to offshore in compared to ours, but I mean, ours is just some code running in the cloud. So how it can't be that expensive.
But I mean, you got to get to that point running. Yeah, operationally, yes. But I mean, to get to the point where you have like, systematic, amazing AI that can actually turn cold leads back to warm. I mean, that's, that's, that's a feat that needs to be continually managed. Right.
Yeah, I think it kind of comes back to that. That kind of founding story of like, we didn't have any data to start with. So we had to create data and tag it really aggressively. And now we have more data than we know what to do with it at this point. That is kind of where the economies of scale I guess if that's the right term for that help a lot because we have the data. We know that we know exactly what it kind of costs to support all this and we've actually seen our our operating costs stay for Flat as we grew our leads processed, which is exactly what we want to see early on, it made no sense at all the economics were very bad. And our board, called us out on it aggressively early. And we said, we know we'll fix it, we just need the data. So that there came a point where it kind of switched. And now we're, now we're seeing the benefits of have a lot of data, we can we can scale really nicely, and we have the means to do so. You know, I'd say we're still a lot cheaper than a lot of our customer, our competitors, who are entirely call based, or I mean, human based, partially because they also do phone calls. And that's the saving grace of humans, in a call center that they can offer is, we won't do calls that structurally, probably ever, because of a couple of reasons, be impossible to do it with AI, at least now maybe in the future future. That could be a thing. But
also Google have that didn't do the demo. We're
Yeah, that was crazy. And then that piqued a lot of our interest and structure. And we're like, if that's going to become a thing, there's a chance, but they've been really quiet on that ever since. So who knows what they're doing with it?
Well, I think it's a really hard problem that you probably don't need to solve. Right?
Yeah, right. Phone calls would be a whole different ballgame. Yeah,
because I think I think as as time goes by more and more people, I mean, how the only thing I ever get to spam calls, right? So whoever, who actually calls somebody or uses the phone anymore, I mean, I think the modality is kind of switched to text and email and right. I mean, more, more people. More people use that to communicate than than anything else. Is that a synchronous peace out? I think
we're a little biased, but we've done a whole series on dials are dying. They really are people are answering the phone less and less and less, because you have no idea which spam call this is going to be today. So yeah, I just didn't want to deal with it.
Exactly. I don't remember the last time I picked up a call from somebody I didn't know I didn't think I don't know. It could have been could be years ago, I don't know.
Exactly. And what we found is phone calls have the time and place still. If you if you set it up so that you are texting, for example, and say you establish a conversation. So now that number is like saved in your phone or could be or at least it has a text thread. And then you say like, Hey, can you know someone on my team? Give me a call in 10 minutes? They're expecting that call? They know who it's from? There's a text thread there, they're a lot more likely to answer they might not but a lot more likely than just making a random outbound phone call off the cuff. Because yeah, no one answers those anymore.
Exactly. So do you guys don't do any cold, then it's always warm.
In terms of our outreach? Yeah. We don't like if it's a if you're just like pulling down a list. who's never heard of you that's never filled out a form or anything we do not because there's some compliance around opt in and TCPA texting laws where they have to have consent. And they are the the carriers are getting a little more aggressive with that. They're even requiring a lot of messaging, in our messages themselves that say, like, reply stopped, stop to opt out, which makes us sad, because it makes it feel a lot less human. But it's the law. So when it comes down to cold calls, or like outreach to lists that haven't opted in, we always say, you know, they could definitely sneak through infrastructurally. But, you know, that's that is ultimately on the customer to make sure that the lists have been opted in, and that they're in compliance.
Yeah. And I mean, you don't really want if that's the lower hanging fruit anyway. I mean, if they've already had communications, it's it seems like a stretch to apply this to cold calling, but maybe at some point in the future cold emails,
we yeah, we have seen some use cases with cold email, because that's a little more flexible in terms of the laws and often it doesn't work to it doesn't work as well as the other use cases, you get, you get low response rates, low conversion. So if that's kind of like, you're desperate for it, then that's a deal. It's an okay, use case that can be supported a little more lenient ly through email was structurally, but not a best practice for us.
Right. So you mentioned before that you were in the property tech, is that what it's called property space. So you're doing mostly real estate and things like that. Do you think the model is actually extendable to other other areas or is it pretty focused into that space?
Yeah, so we're really focused on expanding into insurance right now as well. It's pretty similar to the market that print In prop tech, they'll kind of act like there's brokerages with salespeople who get leads that they need to convert into appointments. They're just selling policies, not houses. So we think that that type of customer is kind of where we see our sweet spot. We see our ideal customer profile is the businesses that have a lot of leads that are generated or managed digitally through like a modern CRM. And that take a lot of back and forth to convert. Kind of an optional one is like there's a lot of tire kickers, which is really common in real estate, and insurance in other industries. But yeah, our our conversations themselves are instantly extend and extensible into new markets. So we can go work in insurance. Tomorrow, we can go work in name your industry, we've even seen seen a lot of our current customers, extend our scripts into recruiting, like HR recruiting, because agents also have a means to recruit a lot of other agents to their brokerage because they make more money that way. Totally different conversation than buying and selling a house there, you know, questions like, tell me about your professional work history. Do you want to work full time or part time? What do you do for fun, things like that, totally different conversations that our product has had. Without, you know, lifting a finger. So we were, we see a lot of blue ocean in terms of our product, but we're trying to narrow in kind of where we would make the most immediate impact. And it's kind of those B to C type markets, you know, the big b2b enterprise sales don't make a lot of sense for us, because they're really long sales cycles, they require a ton of personal touch, you probably gotta get on a plane, you know, the guy or gal, buying them a drink, like, you know, um, you don't want to the last thing you'd want to do is throw another random AI to try and qualify them in that conversation. So it's a lot
more touch anyway, in to take a really long time. Yeah.
So we're a lot more on the transactional, I would say side than the big kind of bulky sales.
Sounds good. Okay. So it's time to think like a futurist to see your 2030 to 10 years from now. Where will you be 10 years from now? Where you going to be making cold calls?
That would be amazing. Because I do think so in 2030? Yeah, I think. So there's a here's my little secret sauce that I've been watching on that Google project. There's a there's a I don't know if it still exists, but I think it does. There's a team within Google that I've been able to find who's mostly responsible for a lot of that project. They call themselves taco Tron. So if you Google taco Tron, you can, you can keep it. I don't know why they picked that name. But you can keep up to date with a lot of that work. It's very research oriented, they're just kind of messing with things. But I'm sure that's the team that created a lot of that demo from 20, whatever, 2018. So if you want to stay on top of them, that's a really interesting aspect, they will define kind of 2032, I would say, in terms of cold calls with AI type of thing. But otherwise, I think, in terms of structurally, I think that we have a vision that we want to make salespeople more productive. Right now we're trying to take a big chunk out of their day by qualifying leads and setting appointments for them, basically, we think there's more that we can do, especially in terms of like writing content. If you kind of take a look at all the things that salespeople do, it's set appointments, qualify leads follow up with leads, these are things that are time spent not selling in, that's really annoying to us, we want to we want to make them basically spend our whole day closing deals. And kind of the last one of the latter frontiers is writing emails, writing, copy writing, content, writing something seems like salespeople a lot of people are just writing all the time. So we want to make it we want to we want to take a bite into that and say, Can we can we actually write emails for you with AI? Can you click a button? And and that that is kind of here and now. So I would say it's it's probably even before 2032 That something like that could happen. But I think that there's a lot more to be done with kind of generative AI for salespeople to make their lives even easier. Today, do you
think do you think we're moving more towards an inbound model than an outbound model? Because more and more people are just sort of shutting off sales messages and going okay, you know, I'll contact you when I'm interested. Yeah, so that would require writing, writing content, writing emails, just putting stuff out there so that we You're seeing and then there's an inbound you see us moving more towards?
Absolutely or Yeah, absolutely, I think there's always a time in place for outbound to make an attempt to be really personal. I would say that in terms of inbound, we're, so we've always kind of known, especially for the last 10 years that buyers are much more in control than they ever have been, they dictate kind of the sales cycle. But I think they've become more like quiet, they figured out when they're in a sales cycle now, so they won't fill out a form, they won't reach out to you, they're going to try and stay in the shadows until they're ready to actually really pull the trigger. So I think that there's a lot of new software out there that actually open up opens up buyers intent without them actually knowing. So like Clearbit, if you're familiar with Clearbit, they can like D anonymize IP addresses, and say, like, I can see that structurally is on my website right now based on the IP of that person visiting. And then they can kind of keep a profile of like, oh, they keep coming back to the site every day. They're getting warmer, they're they're hotter, they're hotter, and then making a really personal touch, like, Hey, see? Or was structurally and you've been on the site for few days in a row now, wanted to see if there was anything I could answer. So that is like a really not creepy at all. It's a little creepy. But it's kinda like, that's, I think that's where we're going a little bit.
I think we got to go there, because that's where, like you said, buyers are really shy. Like, that's the kindest term I can use. Because we've been bombarded with so many Miss sales messages all the time, you know, outbound and ads, and everything. People are like, okay, it's kind of like playing poker, right? I'm going to keep my cards really close. And you're not going to know what I want to do when I want to do it. But I think we've already beaten that, because we've got all these devices all over us, protecting what we're doing, and I need systems are going I think this is what he wants to do. So you know, a sales call is eminent. Yeah.
Yeah, exactly. You can't ever get away from it.
Well, I can see the point where at some point, there will actually be a an intermediary on our side. So it won't just be our AI talking to us, it'll be AR AI talking to their AI, and then these two will have to duke it out.
We have seen our AI somehow get in a loop where it was talking to itself that a little bit of a disaster. We've put some Oh, no, we put some some safety mechanisms in really early on when we caught it like the first time. But yeah, I just didn't make it anywhere. It's just back and forth.
Ya know, and it's, it's, it's interesting that this is happening everywhere. It's like the poor consumer is being inundated with these super high powered AI is everywhere, trying to sell stuff. But the poor consumer doesn't really have any kind of hedge against these super powerful AI is because there's no super powerful AI is working on behalf of the consumer. So they're like,
and I think that's actually where Google is, has moved that kind of project to where I can't remember if it's Google or not, but they they put in like screening. So they'll answer your phone call to start. And if it's spam, they'll say it's spam. But if it's not, they'll transfer it to you. So it's like, I can see the general AI players like Google getting us in a spot where it's like, you know, hey, we'll take care of dealing with that sales call in that sales email, and you just, you just will, will qualify who you need to talk to and not for your personal life.
But ideally, you should be able to tell this AI Oh, this is what I'm interested in, let this through, don't let this through whatever and it should be able to figure that out and then sort of protect us. So at some point in I know we're going to be there that you know, it'll be my AI talking to your AI just like Mike, my secretary talking to your secretary my people talking to your people. Yeah, it's gonna people it'll be AI is talking to each other. And then it'll be tough because it's kind of like, how do you sell an AI house? Right? Because you have this they get they get to sell it on behalf of, of the buyer. And it's like, oh, you know, do I want this room? Do I don't want this room? Yeah, so you just gonna have AI's talking to each other?
Yeah, I guess.
So. So the future plans for you guys is it sounds like you're gonna go you're gonna get into voice if you can, and then what expand beyond the space you're in or like, where do you see the company in 10 years? Where do you see them?
Yeah, I always see us looking into new markets for sure. We believe that again, every CRM will have Converse every CRM lead Then marketing automation company will have conversational AI. So we want to be in as many of those products as we can, whatever industry that are in, but obviously to get there, we can't just go, you know, sell to the the biggest tech company in whatever industry from day one, what we've seen is we have to kind of prove ourselves in the market first and then go top down, I guess. So, yeah, I think it's about knocking down our new industries. And we're being really deliberate with them going after Prop tech, kind of collectively, because mortgage and leasing are in that they're just smaller, but they are wildly different conversations. Kind of insurance is next. But then there's also travel auto. A lot of these industries that have a lot of leads, again, that need to be qualified with a lot of looky loos or tire kickers that talk over email and text. So kind of about knocking down those industries. And yeah, over time,
like thank you staying, staying in a narrow focus, because I think that's where you have to go instead of like, going, Oh, we're gonna provide everything to everybody. Yeah, exactly. That's the way to go. Right? Yeah. Very cool. What's this? This is very interesting. I definitely got to check it out after I get off the phone. So if somebody wants to get in touch with you, what's the best way?
Yeah, you can actually try it yourself. So if you go to test your ai assistant.com, I always have to take this test your ai assistant.com. You can fill out the form.
I'm just making sure that that's right. I might need a refresh. Yeah, so it is test your ai assistant.com. You can fill out the form there and act like a lead yourself. But that is we always like to say, you know, have a conversation with it as if you were a real lead. And this is going to be in a buyer, buyer. Real Estate sense. We should see a lot of people fill out that form and actually, like say, you know, when? When are we going to discover aliens? And what language are they going to speak something super weird, which we'll have a conversation about. It's just, if you keep asking it that over and over like, you're clearly not, you're clearly not there to actually talk. That's not how leads talk. We see weird conversations like that. But you know, give it a little bit of leeway to actually have a conversation
when here's a suggestion for you. Because all it's almost real, all real estate is really geographically focused, right? So you say I want to live in LA, I want to live here. I want to live in here, here. But now with remote work. You can kind of live everywhere or anywhere, right? So can I can I go in there and say, hey, I want this is the house I'm looking for and I don't really care where it is. As long as these criteria are met. Can you do that yet?
We can't that is a whole can of worms because we would have to plug into home searches. And yeah, we we like to say leave that to the agents will just take their leads and qualify them as they get them regionally and locally. But yeah. Someone else can solve that big hairy problem.
See, there's the thing is like, I would have to go to an agent in every territory, right? I mean, I want what I want to just say I want a three bedroom house with two baths, anywhere in the country. Like nobody can do that yet. So right maybe in the future. Right. Right. All right. Well, I'll put all the information in the show notes and people can get get in touch with you that way. So it's been it's been great talking to you. Thank you so much.
Yeah, absolutely. Thanks for having me on. So it's great to talk to you soon. All right.
