He Built 80 Dental Practices Without Taking Away Autonomy | with Lane Theriault, BOAC #56

0:00We started coming off of COVID, thereabouts, eight groups, now 80, that's 10x in the last, say, five years. Practice telling, we keep losing people's to Buc-ee's down the street. This is what we call a leaky bucket, or sometimes the analogy we use is the toilet versus the bathtub, right? The practice is a toilet. You can add water to it and the water level stays the same, right? Whereas the bathtub with the plug in it will actually fill. And so you're absolutely right. There is no sense in adding market expense to a leaky bucket, right? I caution all of our folks. Do you know how much it costs to get that phone to ring?

0:29Imagine you got a couple hundred dollar bills in your hand, let patients on the phone, and if you convert them, that's going right to the garbage. A lot of times people call and get the voice AI, and they know it's AI, and they hang up and they call back again to try and make sure they can get around it. Right. What we also found was that actually the most effective way to use the AI was to be very explicit and upfront about the fact that it was AI. He said the first thing it's going to say is, hey, I'm the voice AI receptionist. I'm here to help you, right? And people go, oh, right, right off the hop.

0:54I think, first of all, like a lot of people don't really know how to use AI. They think they do, but they don't. I think that there's automations that are happening that we can't even fathom right now, right, that in 10 years from now are going to be kind of the norm.

1:12Hey, everyone. This is Jared with Coherent Healthcare. We're back for another episode of The Business of a Clinic. Today, we're with Lane, who's got a really interesting set of experiences wearing many different hats in executive leadership roles in healthcare. We're going to dig into that experience in a minute, but just before we get started, Lane, I have the benefit of knowing a bit about your background. Others don't. So I'm wondering if you could give us the, if you could compress the last 15 years into two minutes, how would you do that? What's the grand introduction just to get us going?

1:46Well, first of all, I'm not a clinician. I'm an investment banker by training. I spend most of my career at Citicroup working in the M &A department. and when I left the bank, I wanted to strike out on my own and do a roll-up business and eventually decided on dental. It's kind of the space I thought was most interesting. At the time, this was 2016-ish. I was head to Canada and the only kind of DSO out there was Dental Corp, so it seemed like a ripe environment for a new player. So myself and some partners launched 123 Dentist in kind of mid-2017. I stuck with them until the first kind of private equity transaction in 2019 and had an opportunity to leave the penalty free and had had some success in the space.

2:30So I wanted to stay with Dental, but I didn't want to compete against my alma mater. So that's what kind of brought me down to the US to launch Independence Dental and sort of started with the same playbook. Launched with paid practices at the end of 2020 and I've grown to about 80 practices now. So that's the super short version. Got it. And there's a different conversation we can have about the fact that I interned at Citigroup back when I was at university on sales and trading desk. So if we can go down that road, it was only for about three months. So I'm going to do a whole career there, but we can probably talk about that separately.

3:05Okay, interesting. So investment banking, dental consolidation in Canada, and ultimately now building possibly a larger scale version of that in the US. That was the two-minute version. Maybe I can ask for the five-minute version, possibly double-clicking on the Independence Dental story. So just help us understand a bit more the context. You started sort of coming off of COVID, thereabouts, eight groups, now 80. That's 10x in the last, say, five years. How do we understand the size of this aside from 80 sites, patient numbers? How would you sort of help us understand the scale of the operation right now?

3:40Tone man. Yeah, I mean, it's so fluid, right? Like we're constantly finding new partnerships, right? As you know, when people ask me, you know, how many patients you have, I can barely remember how much revenue we have because it's constantly changing. But I guess our model really kind of derived from the Canadian experience. And, you know, when we first launched 1, 2, 3, where we had Dental Corps was our competitor, we knew we couldn't copy them and beat them at their own game, right? So we had to differentiate to come up with a new strategy. and so we went through the same exercise when we were launching in the U.S.

4:17with independence and we basically found that there were a lot of DSOs out there that had a very rigid playbook like when you join us, you're going to do things our way and what worked well at Canada was a supported autonomy model and so we saw not a lot of folks doing this down in the U.S. and so we saw a clear opportunity to differentiate ourselves there hence the name independence dental right you can preserve your autonomy is essentially what the name means right and that's that's really resonated with doctors right because if you think about kind of the traditional dso sales pitch they almost tell you that your baby's ugly right if only you did things our way this thing would be so much better um and so we've chosen them to uh partner with you already successful practices and support them to be a little bit better rather than kind of doing the the big overhaul um so that's kind of how independence came to be.

5:10And then, you know, over the years, I kind of, we started, we had, you know, a shoestring of staff compared to what we have today and, and have now kind of grown into kind of like a full support model, but it's kind of take what you want, leave what you don't. We're more of like a menu of services to pick and choose from. And we can still have, you know, two practices in the same market with totally different strategies and both be successful. That's totally okay in our model. So if I'm not, it's not me trying to get you to do the pitch, but if I'm a dental practice in a prime market for you, a place that you're interested in growing and requiring, I understand the independence and support model.

5:52And I think they're at least in the UK, I'm based in the UK from the US originally, but at least in the UK, there are quite a few players who are, I think, pursuing that strategy outside of dentistry, medical aesthetics, MSK, et cetera. So I think I can see why it's appealing. I understand it reasonably well, not on the investor side, but what's the pitch to me if I'm a dental clinic owner? So what do I get to pick and choose from the salad bar? What is my journey with you post-acquisition integration? What does that look like moving forward? Yeah, I mean, it's all the usual stuff that a regular deal would come from, right?

6:26So the biggest one we get asked about is recruiting, right? How can you support me, right? Staffing these things has become increasingly difficult. Even in the last five years, independence has seen a big change in how that works. That's where we spent a lot of our corporate resources in developing that recruiting function. Probably the second biggest one is marketing. I think everybody has that marketing person that they rely on. What we found is they're generally set it and forget it. right they once they get you as a client you know that's the job's done right and so we spend a lot of time not necessarily changing their marketing person but holding those people accountable um so those become the two biggest examples and we do have like some mandatory things that you have switch over but it's you know practical amount of stuff right you got to do the payroll we got to do the bookkeeping the accounting all those types of things have to fit under our umbrella right free normal right and that list has gotten a little longer over the years right procure rent has to be kind of all under one umbrella so we can make sure we get the best deal for everybody right those types of things but um yeah generally speaking um you know any kind of vertical you can think of revenue cycle management help there you know posting claims seems to be a popular one that you know people want help with those types of things working claims um you know big space in the u.s big difference from u.s and canada much more difficult than the U.S.

7:50So all those things we support with, but I've kind of given you the rank order priority ones there. Got it. That makes sense. It's interesting, the two that jump off the page, and I guess you've listed out first, recruiting and marketing. I'll call it marketing effectiveness or marketing accountability. Maybe just looking at recruiting for a minute, I'm interested, is that on the clinical recruiting side at the administrative recruiting side or both? Is this a sort of recruiting end to end full stop support? Is that something that you find or is there a particular pull on one of those two? No, it's definitely end to end, but the hard one is the clinical side, right?

8:36And not just doctors, right? Like we're seeing like a lot of folks can't find good DAs and, you know, the market has been really pressured by inflation. And, you know, I have one story where practice tells me we keep losing people to buckies down the street, right? You pay very well and you can kind of work any kind of hours of the day you want. So those are the kind of challenges we face. But I'd say, yeah, generally speaking, the clinical side is much more challenging than the administrative side. Okay, got it. And then on the marketing accountability effect of this side, I'm sort of wondering when we work a lot of practices who I think either have been acquired or are assumed to be acquired or were just acquired, the marketing transformation is usually quite dramatic again like you're saying not because they weren't doing anything before but because the extent to which that spend is is really rigorously assessed and optimized is usually it's usually not just not there so when you go in and you're thinking okay we're going to start helping the practice mature their marketing stack what sorts of levers are you usually pulling first or second i mean there's probably a hundred things you would do but what are the sort of big the big whale moments we're going to look here and here first i mean the number one thing is the you know the essentially free stuff like we call it the guerrilla marketing right like do you have boots on the ground are you out talking with community right um and there's a whole kind of laundry list of uh kind of tactics that that we use to get um you know community engagement right as usually very cheap like you know you might spend like a thousand bucks having like um you know some kind of tent outside the practice or you know visiting people around you, businesses around you.

10:16Those are definitely number one and by far the most effective, right? The trouble is, is it's just, there's only so much you can do in that realm. There's also much reach you can get, right? And then we have not really found that, you know, the Facebook ads, like the social media stuff, kind of low intent, right? Like if I'm cruising through Instagram late at night and I see an ad for a dental practice, it's unlikely that I, really need something. Whereas like in a Google ad, you're seeking it out, right? And so we found much more success with those where there's an active intent behind it. And then I would say, you know, it's somewhat counterintuitive.

10:54Like some folks are managing their marketing budget and trying to spend less. And when we do a lot of, you know, the analysis, like we find a lot of our practices, you know, if you could keep, if you could spend twice as much and get the same kind of cost per patient, and the patients are valuable, right? They need an comprehensive exam, right? Usually they come in because they've got some kind of issue, right? Generally speaking, they're the bigger ticket is this. And a lot of the time we're coaching our partners to say, hey, you actually need to spend more on this, right? You've got an opportunity that you're missing.

11:26It seems to use the language coaching your partner. So does that decision, I guess at the practice level, does it, okay, I'm going to ramp 2X on my marketing. Is that a decision that's ultimately taken at the practice level? Or is that a decision that you sort of take at head office, but then coach them through making the same, where are those, how does that decision get made on a practice by practice basis? Yeah, I mean, so in our model, right, we sort of made this promise to our partners that they're going to have autonomy on both the clinical and the business side, right? So everything we do really has to be consensus driven, right?

12:00And I liken this to like this old parable I love, which is the sun and the wind are having a contest to see who could get the traveler to take his coat off, right? And the wind puffs and puffs and the traveler only grips his coat more tightly, right? But the sun just needs to shine gently upon him. And he chooses to take his coat off, right? And that, I think, is the perfect analogy for how we build consensus in our organizations. Because the reality is, like, you can mandate it, you know, it doesn't mean it's going to be effective, right? You're going to get people resisting. and so you might as well go through the exercise of building that consensus because you need to do it anyway for it to be effective.

12:38Got it. So part of me wants to go down the road of marketing efficiency because I think it's a particularly interesting topic right now because cost of acquisition is generally going up in competitive elective care services, at least from what we see in dental MSK medical aesthetics. On the other hand, that's probably a a personal interest more than an interest that may reach the full audience that we have. So maybe a quick dive into that hole and then move on from there. The dive in that I'd love to do is understanding from you how you think about what comes out the other side of that marketing spend.

13:14So you said, if you can ramp 2x on the marketing, get the same economics coming through, do it. Why not? What we often find is that sometimes that marketing spend and the efficiency that marketing spend gets lost at the point of patient handling. So they sort of drive the intent, but then no one's there to, you know, who's there to catch it. And sales always makes everyone comfortable healthcare, but ultimately there's a sales process that sits on the other side of that marketing activity. And oftentimes we find that clinics are not equipped to optimize the conversion muscles and the retention muscles.

13:48So you're doing all this marketing work, but you're, you know, you could have 20 % organic site level revenue uplift if you tweak one or two, three things in the handling process of that inquiry. So I'm curious, when you look at that marketing transformation, are you doing it at the spend and advertising level? It sounds like the answer is yes. Are you also doing it at the desk level in terms of the actual conversion mechanics and the retention? Yes, you've hit the nail on the head here, right? This is what we call like the leaky bucket or sometimes the analogy we use is the toilet versus the bathtub, right?

14:23Like if you have a, you know, if the practice is a toilet, you can add water to it and the water level stays the same, right? Bad analogy, but that's what we use, right? Whereas, you know, the bathtub with the plug in it will actually fill. And so you're absolutely right. Like there is no sense in, you know, adding a market expense to a leaky bucket, right? And you'll see that, you know, we're very kind of metrics focused. We've got a ton of KPIs we use to analyze the business, you know, to thank again with the Avid AI. We're analyzing even more. And absolutely, like just that one step from, you know, who picks up the phone?

15:01Like, you know, I caution all of our folks. Like, you know, do you know how much it costs to get that phone to ring, right? Imagine, you know, you've got a couple hundred dollar bills in your hand when that patient's on the phone. month and if you don't convert them you know that's going you know right to the garbage so you've got to fix that stuff first right and that's not the only one it's not just called conversion right you actually have space in the calendars there's space and efficiency right where um you know a lot of our uh training is around kind of the rock pebble sand or rock sand water or maybe for these type of analogies right it's like saving space for that new patient right so you've got a uh not just a full schedule but a productive schedule so these are all the things you need to get sorted out before you really pour gasoline on the fire for marketing.

15:43We can move on for you. I'm going to steal. I will credit, but I'm going to steal. Most of our focus is on leak stoppings. It's the entirety of what we do as a business. I've never heard the toilet versus bathtub analogy. We do the leaky bucket a lot. We've got a lot of visuals in the leaky bucket. I'm going to steal the toilet bathtub. That will be one thing that I will steal, but we will credit. We will steal and credit to Lane for that because I've not heard that metaphor. That's really helpful. Thank you for highlighting that. I'm curious. So thinking about the journey, going back to what you said earlier, 8 to 80, there's now a scale.

16:26You're in the scaling phase. What does the future hold? Are we going to 800? Are we going to 160? Do we not know? Is the future however many we need to go to until this becomes attractive to a bigger fish higher up the chain? What's the sort of vision for the group, given where you are right now? And has that always been the vision? Or is it sort of a quasi-murky North Star where you're sort of working toward it? But hey, there's an opportunity. Things have changed. AI has hit the table now. so actually we can do things differently and now we're going to get even bigger. How have you been thinking about the end state and how has that changed?

17:06I mean, it's definitely changed since the beginning. We always wanted to be kind of one of the top 10 DSOs in the country. But the current market is just dramatically different from when we started five years ago, right? And in my opinion, like right now is not really a great time to be out acquiring, right? If you think about, quote, unquote, like the spread, you know, high quality stuff doesn't trade for much less than the whole platform does. Like you've got costs, a good cost of transacting, plus you've got additional managerial burdens for having more practices. So for us, we've really slowed down and become a lot more picky about who we partner with, where we partner, and are kind of saving our dry powder for, you know, greener pastures in the industry.

17:55So yeah, I think that's a huge thing is that you need to respond to kind of the market conditions. And just because the playbook worked last year, the year before that, doesn't necessarily mean you should be continuing to do that, right? And in the old kind of model, organic growth was not really prioritized. That is very much the contrary today. It is an absolute must-have. And I would say there's almost no point in acquiring if you can't then grow that practice, the original practices. And so for us, 2026 has really been a big year for us in terms of technology and operational excellence. And we've had some interesting experiences with kind of vendors in the technology space.

18:43and have found that, you know, with the advent of kind of these coding agents where you can essentially build a lot of this stuff yourself, get better customization, right? Create something that's perfect for you as opposed to something that's out of the box. It's kind of works okay for everybody. That's really where we've been focusing and trying to develop, you know, what I call a culture of AI, right? where it's not just, you know, somebody in the IT department who's developing all this stuff, right? We're kind of giving everybody at the corporate office the tools to go and make solutions that fit their specific needs, right?

19:22So I would love to give the example where my director of HR has done a wonderful job here, right? Kind of gave her the tutorial on how to create her own apps. And she's essentially digitized what was before, you know, a number of like Excel spreadsheets and created a whole process that, you know, automates everything. You can't get past this gate until you submit all your licenses and, you know, your production reports, right? It can scan the, are we have a knowledge base, scan knowledge base, make sure that interview was scheduled, reporting on all the KPIs, and it's all very tight and automated, right?

19:54But, you know, I couldn't have built that. She had the specific knowledge about her problems and her vertical to go and build that. Right. And so for us, it's, it's kind of flipping the script a little from that traditional, like, okay, here's our like development roadmap. Um, we're going to build this, then we're going to build that, then we're going to build this, um, to being kind of a, everything in parallel with the people who are closest to the action, building their own solutions. That's been a huge game changer for us. Yeah. I think there's a lot to pick apart there. I would imagine that at 80, resourced the way I assume the group is resourced and with the talent density that you probably have, particularly at head office, I can see why the sort of in-housing of some of that custom application development that lets you save hours on the day and it's sort of victory by a thousand hours sort of logic.

20:51Some of the smaller groups that we speak with, when I say smaller, I mean, you know, 40 sites, 35 sites, there is a very real challenge to produce safe, reliable production grade capability. And so you end up with this, you know, collage of 80 % good enough tools, which actually in some ways you're undertaking operational risk because you're trading, you know, patient data when you shouldn't be. So one of the things I guess is on the HR side, it feels like a lot of that optimization is in low risk territory. Maybe not some of it with Canada details, but a lot of it probably, you know, knowledge-based scanning, et cetera, feels like it's in low risk territory.

21:32It's like automation of low risk processes. I'm curious when you, when you think about your two by two of where you're going to direct

21:43that efficiency gun, you know, is it repeatability and risk? Is it like, what are the axes against which your team is, or you're maybe challenging your team to make decisions about how to, where do we, where do we buy in tech and capability? Where are we comfortable to build ourselves? Like what's the decision framework for how to, how to buy or build? because the buy or build decision today is a completely different animal than it was three, four years ago. So I'm just curious how – this is now your second group. So obviously, I guess in your previous group, that framework didn't need to exist in the same way.

22:18So now you've had to sort of build that from scratch. I'm just curious how you're thinking about that at the sort of top down. yeah so um first and foremost right on the buy versus build equation you you typically don't want to build something that has you know uses proprietary information like you know the um like these um ai diagnostic tools is a perfect example right like we don't have the same suite of you know x-rays and all this we're never going to be able to replicate that right we're probably and now we're going to do it as good. It's like, that's one that we bought. Right. But some of these other ones are, um, you know, really just an application layer.

23:00It's moving your own data from here to there. Right. Like that's something that you should be building. Right. So I think generally that's kind of the distinction we make. And then in terms of like, okay, well, now that you've decided this is a buildable opportunity. Yeah. I think the matrix for us is like impact relative to risk, right? And so when we first, you know, started playing around with this stuff, I really wanted a tool. There was this group called Glean that wanted to sell us like a tool that basically like aggregated all of the digital information in our ecosystem into one place and you could ask it questions.

23:32And it was really expensive, right? And I thought, geez, like, I got to believe we could do this ourselves. You know, we know where all our information is and, you know, we've got everything on our server. Like this should be achievable. So that was kind of the foundation for everything that we built. And then we give people access to that on a read-only basis, right? So like you can't change what's in the knowledge base where you're not really editing anything in the PMS, but you can create your own application that writes its own information, but outside of that kind of secured box, right? And then we put all the controls in place to manage, you know, PHI and, you know, all of the confidential information, things like, you know, things that are attorney kind, privilege, those nature where there's all these safeguards around that but once we had that set up right we created all these like what we'll call read-only applications right and so a lot of times it's like sending a notification right or building a dashboard these types of things that we can do right so pretty low risk right you know worst case you're gonna you know make somebody upset because they sent them an email that they shouldn't have received right not a big deal and as we've gotten more confident with the technology, we've moved into kind of more write-enabled stuff, right?

24:43And so RCM is kind of the big kahuna of automation. I think everybody's trying to figure it out. I have yet to find an organization that has got it kind of dialed in end-to-end, right? There's always kind of, and certainly not one solution, right? If you're usually patching in a couple of different things to make it all work, because nothing really does everything well. and so we started tackling verification first right but you know again like pretty low impact if you get the verification wrong well you know patient's information really isn't getting compromised right you might have a write-off or two not the end of the world right um and then you know kind of moving up uh from there but we yeah now we've gotten we've gotten pretty slick like we we've got direct write access to most of our pmss and verification kind of working and you know now I'm going down the list of like okay where is our big technology spend right we've been spending a lot on these like voice AI receptionists and we found that there's really not a lot special about them it's kind of a repackaging of other people's technology and you know the ones that we've used like you know they they work but I would say they they don't were great.

25:57And if you could take some time to, you know, find the best voice for it, right? We actually used, we scanned a lot of our own call data to come up with, you know, what is the most successful pattern, right? How, what information should you ask for at which times? And, you know, a lot of the tonality and things like that to get, you know, folks to actually use it. Because A lot of times people call it, get the voice AI and they know it's AI and they hang up and they call back again to try and make sure they can get around it. Right. And that to me is like the big test is like, okay, do they, do they really want to use it?

26:32No, they don't. Right. They use it begrudgingly. So that's another example of kind of things that we've been, you know, working on in housing. But like, if I go through the list, like there's probably 20 or 30 applications that we've piloted, tried, where there's just really no need for the third party service provider. like we can do it ourselves we you know they're not complicated it's really interesting the um it is i think there's a lot of you know outside of health care outside dental uh noise sass is dead sass is dead sass is dead your software is dying um and it's hard to understand how that actually shows up and then you give an example like that where it's not so much that software applications are dead it's that maybe you no longer need the vendor to to provide it now you can sort of uh you know create and wield the power yourselves with you know within reason it's interesting the voice ai the reception piece is in some ways i feel hugely validated based on what you just said because i have been banging the the drum when i was running practice we we were early to testing some of the very very early voice models and trying to do this ourselves and their you know their issues are quite you know buggy and and you know tone wasn't right and Hours, right?

27:46We were losing marketing spend dollars when new leads were called. What we came to find was that it was really good at covering the gaps, right? Out of hours handling. What we also found was that actually the most effective way to use the AI was to be very explicit and upfront about the fact that it was AI. And so we found that it was really helpful if the AI agent answered the phone and said, hi, I'm an AI agent. The reason I answered is because it's out of hours. So you've got an option. You can either talk to me and I can try and get someone to help you quickly. Or if you want to call back tomorrow morning at 8 a.m., someone will answer the phone.

28:20And when it was very sort of like owning your plague, patients were actually quite happy to use that effectively as a very enriched answering service. But when it tried to, you know, trick the people, they got quite frustrated. And that like mismanagement expectation, it sounds like you've found versions of that in your own experience with AI voice agents. Is that right? Yeah. Okay. Yeah. Interesting. I feel like we've had the exact same experience, right? Like that's exactly what happened when we rolled it out, right? And confusion and, you know, people getting upset, right? I don't want to, you know, they're almost like kind of upset that they feel like lied to almost, right?

29:02Like you're trying to conceal the fact that it's an AI, right? And that's exactly what we did. We said the first thing it's going to say is, hey, you know, I'm the voice of the AI receptionist. You know, I'm here to help you, right? And people go, oh, right, right off the hop. you know, they're much more, you know, accommodating or, you know, they're willing to sort of entertain and go through the motions with the AI once they're up front about it. Completely. And even I think for, you know, for certain audiences, they might even change the way they speak. They might even recognize, you know, I'm familiar with interacting.

29:35I know how to interact with AI agents or, you know, models or, you know, chat interfaces. And so now that I know this is AI, I'm going to, I'm actually going to change my own tone and the clarity with which sort of works in both directions and you get that, you know, the reciprocal exchange of I'm going to be up front with you and you're going to help me serve you as best I can as an AI agent. Interesting. Yeah, very interesting. Have you found, so when you've tried to implement, you know, this dark animal of AI across the practice, where have you found it really gives you the sort of, where the promise and the reality line up?

30:13Because I think oftentimes, and we will say this to our customers today, they'll ask us, oh, are you an AI provider? We say, no, we use AI to provide the thing that we do, but we are not an AI provider. It's part of how we deliver the thing. It's not the thing itself. Where have you found that actually what's written on the label matches how it stacks up in the practice, in the operating environment? Have you found places where actually, you know, they said it had seven grams of sugar and it actually had seven grams of sugar? Has that been your experience anywhere? I mean, I guess, I think first of all, like a lot of people don't really know how to use AI.

30:55They think they do, but they don't, right? They think it means like opening up, you know, the chat agent and having a conversation with it, right? And so the first thing we did was that knowledge base to really kind of bring in the context for the AI to understand our organization. It's like, that's got to be stuff. If I was talking to another DSO CEO, I'd tell them they're interested in this. I'd say, number one, get all that data in there so you can actually use it. And then number two is you need to learn how to code. like even the most simple AI, like if you're going to use it in a production capacity, like you're not going to sit at your computer and type in commands all day.

31:36Like that's not automation, right? You need a program that's like accessing the AI as part of the tool, right? So, you know, using the Glees Claude, Claude code, like that's table sticks. You're not going to be able to get any kind of real use out of AI unless you're able to program it in a way that it's actually useful, right and then i would say um generally speaking like ai is pretty good at doing um you know simple tasks right classification reading documents you know all those types of things but like a human has to interact with this at some point but like we it's a people business that we're in here and so communication becomes really the main output right and like most of the things that we're doing, like what the AI is doing is very simple, right?

32:30It's simple math, right? Take this information, add it to that, divide by this. But the communication and the workflow is really where the, you know, the value is added, right? So for us, like, you know, we generate a lot of emails, right? Out of our, our, our applications. And, you know, I was just on a call with all of our practice leaders just, you know, last week, you know, the email traffic's getting pretty high. And we asked him, you know, is this too much? And I was shocked to hear the answer. They said, absolutely no, we love it. I love getting a personalized email that is specific to this subject.

33:04I can forward it on to the right person, print it out and post it on the wall and show, you know, my other staff what exactly we're doing, right? And in fact, what they wanted was not less emails. They're like, could you aggregate this all in a dashboard for us? Can you show it to us, right? So it's really about, yeah, The function is just that one layer. You've got to communicate and get it into kind of the everyday workflow for it to be successful. It's interesting. One of the very big part of our business, obviously this hits dental as well, is in patient recall and reactivation of the legacy.

33:40We're not even legacy. Patients haven't been in for a while, need to come back. And what we often find is that, I'm thinking about this because of workflow as you're describing it and process. But we often find that we start working with practices in the first conversation. They're interested in patient recall. We'll say, do you mind if we ask you a few questions? Tell us how patient recall works today. Oh, well, if a patient hasn't been in for three months, then they get an SMS. Oh, OK, great. Like, then what happens? Oh, well, and if they haven't rebooked by six months, then they get an email.

34:08So great. Then what happens? And then two, three more questions. And the answer is, I don't know. I said, well, it's hard to imagine augmenting or making that process much more impactful and efficient if you as a teen can't articulate what happens next. And I think it's exactly as you're saying that there's this great visual of, you know, stuff going in, but the stuff is garbage. It's trash. And like suddenly you run into AI, like magical unicorn comes out the other side. Like life's not that good. That's just not how this plays out. So it's interesting to hear you say that actually you've sort of come to terms with the fact that the AI is not all singing, all dancing.

34:49It's really high impact when pointed at a very specific thing in a very specific way with very specific guardrails, which we agree with that it's exactly why when a clinic leader tells us. So, yeah, we built our own AI receptionist like over the last month. My head explodes and says, oh, my goodness, I don't even want to know what bag of spaghetti is underneath the thing. It's like, that's a totally different problem. Okay, interesting. So the tech transformation story is happening right now at Independence Central. You're on that journey. I am curious for CEOs of other groups and or clinic leaders who are perhaps smaller in scope and size.

35:25This is an animal. It's a whole, it is a discipline into itself. What is step one? if I'm running a practice, I'm running a group, and I want to start this journey of enabling my team to be more effective, to be more technologically literate, to be more technologically empowered, what is step one for you as the decision maker? How do we begin that journey that whether it's cultural transformation, technology transformation, what is step one? And forget step two, just what's step one and how do you know when you've done it correctly? sure and listen this is this is the way i did it i'm not saying it's the only way to do it and and like i'm certainly not the greatest ceo to ever walk the face of the earth like i um you know i'm really just an entrepreneur who kind of got thrust into this role so like there's lots of better leaders out there than me but what what i did was you know i'll go take a step back you know first i wanted everybody to start using this we got everybody you know chat gpt subscriptions and we were asking our monthly call like how did you use ai this week and i'd get answers like oh i uh i created a job description with ai right and which i was you know great but you know not really impressed um and people were just weren't kind of taking the bull by the horns here and learning you know how to how to use it so i started doing it myself right and showing them hey you know here's what I built, right?

36:52You know, we had, uh, the first issue that I wanted to deal with in my AI journey was, um, you know, we have all these online schedulers and I wanted to see that I could actually book an appointment within seven days. So I created a little app that just, you know, went and tested all the websites that we had and see if we could book an appointment within the next seven days. And I couldn't, you know, you're going to get an email and it's going to call me and your manager and so on and so forth. Right. And, you know, showed the team this and And they, you know, you see a few kind of heads perk up and go, oh, how did you do that?

37:22Right. And, and right. And then I just kept going, you know, down, you know, built the KPI dashboard, built this. Right. And eventually, you know, I got people ask me like, hey, I really want to know how you did that. Show me what you did. Right. And so I took the time to start with kind of the most senior people in the organization and set them up just the way I did. Right. And then from like once we got going, we pretty quickly realized, thank goodness for our like general counsel keeping us honest on, you know, PHI and all these compliance issues. She said, Laney, you're going to do this. You really got to button this up first.

38:01And so I remember being like a little annoyed and thinking, oh, this is just slowing me down. But I'm very grateful that we did it. Because it turned out not to be nearly as difficult as we thought. Right. a lot of these things um like you essentially need a baa with your service providers we were on google cloud and google cloud that we had a ba with and it covered a lot like you can actually access these ai models using google cloud um so right there we got you know a lot of our kind of problems solved kind of quite quickly and then you know because we're using claud there's a um there's a called the claud md file which is you could think of as like a file that gets read before you, you know, every prompt you put in, right?

38:44It's got all the rules of the road. And so we created like our own kind of custom Claude MD file that says, okay, you know, here's where the PHI is, right? We don't have a BA with Claude because they want to charge us a hundred grand for that, which I said is crazy. So here's the redacted one where we can strip the PHI, right? We have the approved AI bot do that. And you can read this file and that file. And then it also had like the minimum security requirements in there. We were really concerned about everything's web-based. How do you log into this, right? You've got to have an advanced email address, right?

39:18You can't access things from, you know, Kuwait or, you know, Russia. You've got to be in the U.S., things like that, where the data is housed has to be housed in the U.S. So figuring all that out up front is important because if you do have a problem and you've already developed, like it can get out of hand very, very quickly. So that's something that we kept an eye on. And then, yeah, I'll tell you kind of more in detail about what the knowledge base is, because this was really the first kind of online for us was basically we got a very simple AI to read every email, every email attachment, every document in our drive, right?

39:57PMS data, Sage accounting data, you know, EOBs are in there, bill.com invoices, right? And ingesting all of that. And then what it does is it will kind of create a summary. And that could be, you know, pages long. It could be not a single line, depending on what the document is, classifying that. And then we put on top of that, a super sophisticated AI model to go and do what's called like a vector search, right? It's better than just a keyword. It's just for concepts and things like that in those size and through those classifications. So from that, like we can find information very quickly, right?

40:35It takes a couple minutes with the AI prompt to find something very specific. Oh, you know, Amanda was looking at some spreadsheet on a call the other day. I really want to find that. You know, here's what it was sort of about, right? Boom. It can pull that thing up super quickly, right? So once you have that understanding of kind of how the business works and like now we record every meeting transcript, right? All that goes into the knowledge base. And that context really allows the AI to understand you and your organization and what you're trying to do, right? So it's not fishing in the dark anymore, right?

41:07So when you're trying to build something, really understand like, okay, well, here's what they were doing before. And like, here's really what my goal is to automate that, right? And so that to us was kind of the foundation for everything it did. So I'd say go through those kinds of steps. So you got to lead by example, show people what's possible. Somebody in the organization needs to show people what's possible, right? Spend some time developing your team, right? So they can teach others. Right now, I don't do a lot of the teaching anymore. We've got other people who can do it. And then make sure you've got the compliance layer.

41:37Really dialed in. Have that under your control very, very early. And then make sure you've got all the connectivity with the data. Those are kind of the four kind of key steps to me. Yeah, first of all, thank you for going through that. I think that's very clear and it's interesting Also, it sounds like in some ways the thing, if you were making purchase decisions rather than build internally decisions, it sounds like almost the stuff you're deciding to outsource is actually those four steps more than the application. You're purchasing the compliance, the security, all that plumbing is actually what you're buying from the vendor.

42:13You're saying, I don't want to go do all that stuff. the capability layer is maybe easier to spin up, but actually getting there is more complicated. That is really helpful. Thank you for taking us on that journey. I want to be respectful of your time. You've been very generous already. One of the questions I love to ask usually towards the end is, and I think it's probably particularly relevant for you because this is now round two of your DSO journey. You've done round one already. Now you're in round two. If we had this conversation in five years, 10 years, sometime in the future, and you were still building scaling DSOs, what do you think would be the number one thing that has changed?

42:57And then similarly, what do you think would be the number one thing that will not change in the future of providing high quality scale dentistry, in your case, in the US? Well, I mean, I think the thing that's not going to change is the importance of kind of customer service and relationships and how you deal with your patients, right? Like it's a human to human interaction. Trust is extremely important. I don't see that going away anytime soon, right? What I do see changing is we have a real kind of labor crunch in our industry. And it doesn't look like it's getting better. In fact, it looks like it's getting worse.

43:44And I think that there's automations that are happening that we can't even fathom right now. That in 10 years from now are going to be kind of the norm. And the robot completing the crown is the one I'm thinking of right now. I think this is really going to happen, right? And people are kind of got their hands in the sand saying, there's no way a robot can do this. I'm not quite convinced. I think it can, right? And you see this happen in lots of other industries where it used to be a person with their tools doing it. Now it's a person operating a machine that does the same thing. And I see something similar happening in healthcare, not necessarily exclusively because the technology exists, but also because there's a need for it in the industry, right?

44:31The conditions are such that these become viable solutions to the real problems that we're chasing in the industry. Thank you so much for that, Lane. I think we'll maybe wrap there. Again, for all listening in, this is Lane, CEO at Independence Dental. Second time through the DSO. He didn't get enough of it the first time through, so now he's gone ahead and done the second one. And also, at least in the conversation, It sounds like from the folks that we've spoken to here on Business of a Clinic, someone who is quite far along in the AI transformation, cultural change journey that I know is of interest to a lot of group and executive operators.

45:08So if you're interested, I'm sure Lane would would very happily brainstorm or do coffee to talk about how to how to bring that to life elsewhere. Sounds like something he's quite passionate about. Lane, thank you so much for joining and sharing your story. My pleasure, Joe. Really appreciate you having me.