220: Book 30% More Appointments Without More Staff in Your Dental Practice

If you're missing these 20 to 30% of inbound calls and then you have a service that starts answering these calls and generating extra revenue for your practice, then it's a no brainer.

Hi, I'm Emmett Scott and just like you, I've had my ups and downs in the dental industry, but times are changing. The DEO has built a map of how to succeed in one of the most challenging industries to navigate. On this show, we talk to dental industry leaders, solve problems, and get you the answers you need to face these challenging times and thrive. Welcome to DEO's Growth Saf Secrets podcast for dental organizations. Now let's get started.

Hello everyone, and welcome to another episode of DEO's Growth Secrets podcast. I'm Jake Poole, the CEO of the Dentist Entrepreneur, the DEO. And with me I have today the CEO and founder of Arena, Abdul. So, Abdul, do you want to give yourself a little introduction? And I think arena is one of the most exciting softwares and companies inside of dentistry right now. So we're going to go deep into what you guys do, so let the listeners know.

Yeah, thanks so much for having me. Very excited to be here. Yeah. I'll give you a little background about myself. So I'm currently the co founder and CEO of irini. My background really is in the computer science at Harvard. I did quite a bit of research in natural language processing during my time there. And then towards the end of that, that's when I really got into this idea of applying artificial intelligence in real world use cases. So really after that time I went and worked at a company called Threads, led the search team and applied AI team and started applying these foundational models that nowadays everyone is familiar with. You know, everyone is using ChatGPT and these models and applying them for real world cases. And I kind of just got hooked on that, just taking these models and putting them where I truly believe they're needed the most. And that's where the journey for ARENA really started.

Yeah, so arena is an AI powered phone answering phone assistant. So this is kind of where people thought AI was going to go. They're like, man, it'd be great if AI could answer my phones or at least help me answer phones. And so that's what IRINI is. But before we get into that, I'm actually curious because you said you're Harvard and you've done a lot in some, you know, some deep areas of AI, and we're seeing this in dentistry, like talented people, super smart people like you are coming to dentistry of all Places and we're all kind of like, oh, this is amazing. So what drew you to dentistry? How did you get in this industry?

Yeah, absolutely. So around over a year and a half ago. So my co founder Rami and I were really hooked on this idea. What vertical should we focus on? We have this incredible technology at hand. Where do the problems exist and where can we fix them? So what we saw is that we started researching healthcare and we found that the front desk is absolutely chaotic. We're talking literally 20 to 30% of all inbound calls in healthcare in general, including dental, are completely missed. Literally no one picks them up. There's not enough staff picking up the phone. If you hire staff, they turn. Training them is completely like very difficult. So it was a very clear problem for us, for us to fix. Now why dental in particular? Dental is really interesting. There's a lot of consolidation happening. So we even see with do, we're talking two locations, they want to become five locations, the 10 locations want to become 20. And this market segment is growing extremely fast. In some years, even 17% year over year growth in just this mid market dental group target. So that was like a phenomenal market to focus on and do. Dentists are entrepreneurs by heart. You know, they want to try new technology, they're willing to try things out. So when we started over a year ago, we had a few clients that we were testing things with and the feedback that we got from them is what got us to the stage now where we're operating with hundreds of practices in the US and Canada, taking thousands of calls every single day. Yeah, yeah.

It's pretty wild, this 20 to 30% of missed calls saying. And it's been something that folks who have been in the industry for a while, like us, have been kind of harping on for a long time that if you need new patients. Right. A lot of dentists will go and spend a lot of money on a marketing company. Right. Or yeah, but, but if you're missing 20 to 30% of your calls, if you want a 20 to 30% increase in patients, that's the best place to start. However, like as you said, it's, it's harder in practice than, than theory. Right. Like, it's actually hard to, to accomplish that.

Yeah. Like, even if you have enough staff in the office, there are after hours that there is lunch hours, there's the hour before the practice open. You know, what's a metric that's really interesting for us to see is when is irini the most helpful? It's the first hour before the practice opens, during lunch and the hour after they close because patients don't know whether they're open or closed or not, people are still calling at that time, the practice is completely overwhelmed. No one is picking up the phone. So what we do, we go with our dental groups and we tell them give us all your would have been missed calls rather than going to voicemail, just send us these calls. Let's show you how many appointments we can book for you. And literally in the first month we see with some high volume practices, we book hundreds of appointments. We have groups that we're literally booking 7 over 750 appointments per location per month for them. We have other that we're booking a lot less appointments for them because the volume is different. So it really depends on practice. But at the end of the day, if you're missing these 20 to 30% of inbound calls and then you have a service that starts answering these calls and generating extra revenue for your practice, then it's a no brainer. That's how we view it. View it right.

It's a total no brainer. So let's talk about some, some metrics that would help me understand if this would work for me. Which by the way it will because it's a no brainer. Like it's only, it's only gravy. Right. You're going to start with calls that aren't getting answered for sure. So let's say I'm an office, I'm open from like 8 to 5, Monday through Thursday and then half day. I'm trying to get like the most average dental office half day Friday and, and I'm getting about 30 new patients a month. So like how would you like, what's the opportunity there? What have you seen in the past?

Yeah, absolutely. So what we see, it's really, it depends on the number of missed calls. Right. Because that's what we're starting with. So I can give a few examples, I'm thinking actually of a few do groups now. So we'll see anything around like 20 to 30 appointments booked every month per location. If we're thinking of only putting Irini after three rings. So during business hours your phones ring three times and then during after hours and then lunch and then lunch. So really what happens in reality the after hour bookings, that's not what the problem where the problem is you'll get very few calls every now and then maybe at midnight or. But the most calls are actually missed during business hours. So during these three hours at the start of the day, during lunch and during the end of the day. So if you're getting a service, you're training it. Once you're training this AI to figure out how to schedule appointments optimally for you, and then you just get extra 30 appointments a month, then why not, you know, try that out, see how that works you and see the extra revenue.

Now is that what you think? I mean, just take an average. You think an office that's got 30 patients coming in, they're open four and a half days a week, you and they just got a, you know, your normal front desk is operating normally pretty good. Everything's going great, fine. You know, you think they're missing out on what, 10 to 30 appointments a month, you think?

I think for new patients, yes. But there's also a lot of existing patients that are trying to call and book appointments.

Yeah, right.

So practice is getting 30 new patients a month. Then like from what we see, like we're actually booking them another 30 appointments from these missed calls. Like, sure, maybe the front desk, you know, they're getting voicemail. If arena doesn't exist, they're getting voicemail calling the patients back. But there's so much just left on the table because you're not answering that initial call. And if a patient, if a new patient calls you and you don't answer, they're going to call the next dental practice 100%. We're going to book that practice. So really the biggest ROI is on new patient appointments. And that's what we care about the most. Get those bookings as soon as possible. Now what's really fascinating here is we try to get started with overflow calls, which is, that's really the no brainer, you know, give us the call that you would have missed. Let's help you with that. But eventually we do want to take 100% of calls and we have been doing that for quite a few DSOs. Like we're talking 9, 10, 15 location DSOs that literally give Irini 100% of call. And that's when you start seeing insane, insane returns. You can read some of the testimonials we have for a DSO, such as Unified Dental Care, literally reducing staff by 17%, increasing revenue by 12%. These are numbers that like, this is really what matters. You're building the next phase of DSOs or dental groups.

So, so let's talk about the nuts and the bolts of like what it would look like. But first, so you start with the stuff that's just no brainer. You know you wouldn't have answered anyway. Maybe when you're on hold. Yes. You know, your office is. Your front office. Person's already talking to somebody. So that's like a busy, whatever. After hours lunch, three after three rings. But then you're saying the next level is taking all the calls. So why would, why would an AI software taking a call perform? Was it perform better at booking appointment the same than a, than a human?

Yeah. So here's the thing. If you are training, if you have staff that you've trained for years and they're answering every single call, AI will not perform better than that today. And that's a fact. But this is not to say that the AI is not going to eventually become better. It's a fact. If you have these incredible front desk people that are answering every call, they're trained so well. But do those actually exist? Some practices have these incredible receptionists that are really well trained. But in practice that's not what happens. We literally listen to calls where sometimes we can't even hear what honestly the front desk are saying. They're not asking for all the details. You train the AI and it will get you that first name, last name, date of birth, insurance, subscriber id, who the subscriber is. You kind of just like encode it once and it goes and does it for you. The good, the cool thing with AI is once you get to a high level of success rate, we're talking 85, 90 success rate, it's never going to go down. You don't have to retrain it, you just need to train it once. At the beginning in the training phase, you do have to spend some time tuning it, just like when you hire a person, you know, so the expectation shouldn't be that I'm going to turn on this AI service and instantly I'm going to have so many new clients come in. No, we have to tune it together. So it's a very white glove experience in terms of the onboarding, listening to calls, making updates, making sure that if someone says, hey, I have Medicaid. If you don't have, if you don't accept Medicaid, for example, what should the response be? Should it be, oh, I'm sorry, we don't accept that, or should do you offer a package or a promotion? These are the things that really matter and we give you full control to tune these things.

Okay, so what is, what's the nuts and bolts of the implementation of this? So I'm just, I'm just hooking up irini and then it's just taking. Let's just start with the overflow, just taking the calls. That's pretty simple, right?

Yeah. So really the way the onboarding essentially works is part of the onboarding is putting all the basic details. You know, what time you open, what time you close, and then you start getting into what we call AI actions. What do you want the AI to take? What actions do you want the AI to take on your behalf? For example, book emergency appointments with this provider in this column at this time and ask these questions when someone wants to book an emergency. But if they want to book an initial exam and cleaning, then put them in that other operatory and so on and so forth. So you're teaching it the basics, just like you're onboarding a human to do the scheduling for you. Then you start taking some calls every day. Let's say we take five calls today. Let's listen to these calls and together let's see how do we improve it? Did it answer the question about insurance property? If it didn't, that's fine. We just tune it, improve it. So next time it gets a call, we know it'll for sure answer it properly. And that act of tuning the model and customizing it to your need is what makes it a very successful deployment. So we work very closely with the practices to do that. We have actually quite a few DO practices that have very complicated protocols. Like, we have groups that not only use Arini to schedule appointments, but also use it for their hygiene classes. They offer classes to hygienists. So Arini answers questions about this and talks to the potential hygienist that wants to take a class, what time they can come, and so on and so forth. So it's not just for booking the appointments. It can do a lot more than that.

So does the patient or prospective patient know it's AI? Like, how is this going to work? Is it. Is it. Because I've seen some of these on Instagram, I've seen some examples, and they're. They're unbelievable. But the way that the AI answers the phone is like, hey, I'm Abdul AI. You know, so. So does the patient know it's AI or they. Do they not know it's AI?

So it comes down to the practice. Do they want the greeting to disclose that it's an AI or not? So a few practices do that. The majority don't do that in practice. What we see that most people will recognize that this is AI and this is not a problem at all. In fact, the goal is not to deceive someone and tell them this is not AI. The goal is that it's fine for them to know it's an AI, but we want them to engage with it. So it comes down to, how am I going to convince this patient that this AI you're talking to is not the same AI that you encountered when you called your bank or your airlines and you absolutely hated your life talking to it. So how do we prove that to them? So we tune the greetings. The way we start the call is absolutely crucial to get right. So rather than start, for example, by saying, hi, how can I help you? We can say, hi, this is Irini from Mission Care. Who do I have the pleasure of speaking with? Oh, this is Jason. Oh, hi, Jason. How can I help you today? You made the patient not think and just say their name. You repeated their name quickly, and now they're like, oh, wow. It responded pretty quickly and it sounded pretty good. Let me try this thing. And then they go ahead and book the appointment. So these are the things that we're always working on improving and tuning to make sure that people are engaging and booking appointments with our AI.

Yeah, yeah. The banks really have made your life way harder than it should be. Right. Because when I think about calling one of those bank phone trees, blood pressure goes up immediately. Right. And so that's what we get into these calls with. But you're saying the technology has improved a lot. And, you know, my view is that people just want the job to get done. Like a patient just wants a smooth experience. They probably, over the long term, because we have to think about this, right? Like, how are people going to interact with robots like humanoids? How are people going to interact with AI Verbally? And I think at the end of the day, they just want to get stuff done, and if the AI can get it done, then they're fine. So that's what you're saying is people do know it's AI, but once they get the trust that this AI can get the job done, get them on the calendar, answer the questions that they don't care.

Absolutely. An interesting thing that we saw. So we ran an A B test. We have quite a few voices that we experiment with. We have a voice that sounds a lot more natural and it's a lot slower. It actually sounds more like a human. And then we have one that's just really fast. It just gets to the point and it books way more. The faster one that during demos, people say, oh, it's more robotic. That's the one that actually works. Which even further proves the point that what matters is it needs to get the job done as fast as possible. No one will be on the phone for seven minutes trying to book an appointment. The goal is within one, within two to three minutes. We want to get that appointment done, booked in the books on your practice management system, showing that's the goal and we can optimize these things and always change them.

So what are, what are some other hesitations or questions that people have? Like your clients, not the patients. You know, this is. I can't do this because this is not going to work.

Because, yeah, like the main thing they want to make sure, which, like, I think they definitely should make sure they have that is do they have visibility and are they able to improve it over time? So the way I would like, you know, suggest for people to test any AI system like you, you kind of want to think of it not as you're buying software. It's like you're recruiting or hiring a person. So what do you do when you recruit a person? You check the resume. Like, are they legit? Do they actually have customers? Are people using this thing? Does it work? Then you want to check, you want to observe what they're doing. Are they doing well? Are they doing a good job? And then you want to give them feedback and go back to observing. Did they actually update over time? So when you're buying AI software, that's the sort of stuff that you want to do. You want to check credibility. Is this thing real? Are they actually working in production today, improving people's revenue and so on? Do they give you visibility of the success rate of the system? Do they show you when it fails? Because they will be cases when it fails and you need to be aware of that so you can train it and improvement and can you give it feedback? Once you have these, you'll be set up for success without any of these things really start breaking down. Yeah.

So are the calls. The calls are obviously recorded, right? And can you. Because there's other softwares out there that kind of mine. They mine the data, they mine the words that the patients are saying. Are you like transcribing these? Am I able to. Searchable? Like, are you showing reports?

Yeah, so you have reports, Everything is recorded and transcribed. You have access to it, you have your metrics. So you see your success rate, how many new patients booked, existing patient booked, all that sort of stuff. Because that's how we show you that it has an incredible return on investment. Right. We Want to show you that this agent, this employee is actually getting the job done. Because you're hiring an AI company to get a job done, you know, so we have to prove to you that yes, the job is being done and that's through metrics. Yeah.

So what's, what's the, what's the future here? I mean, I feel like we're already in the future and this is like the future for most people. A lot of people don't. We haven't thought beyond this. Or like this is the end, right? Like what's, what's the future?

The future is going to be crazy. Or here's what happened. So basically in 2024, what we prove to the world is that we can build reliable agents that do well defined tasks really well. So we can book an emergency appointment, initial exam, cleaning and so on, or answer on a service. I believe in 2025 there's going to be three different stages of development. The first stage is that these agents are going to reliably do multi stage tasks. So what does that mean? It means Bob calls, hey, I want to book an appointment. I got a referral from Dr. John, he's going to send you my X rays. And then the agent said, oh, actually I'm so sorry, I didn't receive the X rays. Hold on for a minute, let me call Dr. John real quick and get your X rays. The agent calls the doctor, gets the X rays, gets the referral and then calls the patient again, hey, I got your X rays, let's get you booked in. So that's a multi step agent. So that's the world that we're going to live in literally in a few months from now. Like we're developing towards that right now. That's the first stage. The second stage which builds on top of it, is the ability to build with arbitrary software. As you may know, practice management systems are known to not have a lot of integrations. A lot of systems have very solid integrations and we work with them. But many practice management software kind of lock in the practice and their data and they don't allow you to work with like outside vendors. Even if you as a practice owner want to do that. What's going to happen is that even OpenAI recently announced this. We're starting to see models that can use computers really well. So you can check out OpenAI operator just for a demo. And the concept there is that these agents now, not only can they do regular tasks, you know, but they can also use computers. So even if an integration does not exist. That doesn't matter. The agent can go open up a browser, open up the practice management system and book the appointments directly over there. So you think you have practice management. X or Y or Z does not really matter. We'll be able to integrate with any software that exists.

Yeah, you know, we, we have, we have a couple, we have a member who has been working this for a long time. And they're, they're, they have AI that logs into the insurance portals, verifies, just does it all automatically. So we're, again, we're there, this is happening. It's just a matter of kind of going, okay, what was it? What was the third thing?

Yeah, so the third piece is getting to a state where you can start seeing that the existing legacy software kind of becomes invisible. So the humans, the dental practice owners and the admin staff, they're not engaging with the legacy software directly with the pms. Rather they're engaging with the agent layer. They're talking to the agents, and the agents go and figure out how the software works and how to book appointments and manage everything there. And that's where we get to a world where you can actually rip out that existing legacy software and replace it with modern software. It's kind of crazy that until today, a lot of dentists are literally using software from the 80s and 90s to run the most critical part of the business. That needs to change. And I think this is the year this will change because we're going to have agents operating on top of the software. People work with the agents, got the software out, you don't need it anymore. You need the database and an agent on top of it. And that's where IRINI is positioning itself. And the infrastructure that we're building is allowing us to become that hub where you can host all of these agents. One doing inbound calls, one doing messages, one doing outbound calls and so on and so forth.

Amazing. So we're there on the cutting edge of this, I can tell you, within the deo. So DEO members, clients of ours inside the DEO have been using IRINI for a couple years now. And they, you know, they kind of got ahead of the game and we're sharing it with each other. And lot of our members are using arena now. So if you're out there and you're not, you're, you know, some offices nearby may probably already doing this and you might not even know. So I'd say go check it out. And Abdul, I have to ask you, like, you are not AI right? Like, this is a real conversation. That's like a dad joke. I had to throw that in there. But, I mean, there will be a point, I feel like, where we have to ask each other that, right?

Honestly, maybe not too far from now.

No, there's a. There's a software that can. Can take a zoom call. Like, a zoom call for you. Like, have you seen this? And it's not you. Like, you're making the mannerisms, but it's an avatar of you. So, like, we're just looking at each other. You might be, like, cooking dinner or something, and just talking to me is presenting like this. So, anyway, we'll never know. All right, Abdul, thank you so much. I'm very excited to have your company and you inside of dentistry. I mean, that we've. You know, I think folks on the technology side, we've struggled inside of dentistry with technology. Like, a lot of it's nonsensical. It doesn't make any sense. Like, all the technologies don't talk to each other. And so I think we're making big strides in that. In that sense. So thanks for all the work you're doing.

Thank you for listening to deo's Growth Secrets podcast for dental organizations. Continue the conversation on our Facebook group, and meet other dental leaders. Get questions answered and the strategies you need to succeed. Plus, we have some free resources for you right there on the Facebook page. Remember, you're not alone. Let's do this together. Join DEO Growth Secrets group on Facebook. And until next time, keep smiling.