Loading summary
Transcript27 lines
- [00:00]
A
Foreign. Ricky Shockley and Lauren here from MedSpa Magic Marketing, Lauren Nettles. I should say people thought we were married. Somebody said that on one of the things, by the way. Oh my gosh. Probably stipulate. That's so funny. Yeah. And today Lauren thought it'd be fun to talk about AI use cases where AI can be useful in marketing for your med spa and where there's probably still a human advantage, whether it's internally or externally. But hopefully these will give you some things to help scale efficiency, effectiveness testing and your capabilities around marketing. I think big picture with AI, what we're seeing in our own agency and just in the landscape in general is it's just going to allow us to do a lot more. There are nearly an unlimited number of things we can all be doing to market our practices, our businesses and to grow our client base. There's an unlimited amount, right? You could put an unlimited effort, unlimited people putting an unlimited effort. And this is going to allow us to scale with our existing teams and create maximum efficiency. So let's go through a seven point breakdown here on where there's some use cases, awesome current use cases for AI and where the human layer might still make a major difference. And so we're going to go through these point by point. But number one is market research and data for AI. One of the things that AI can do, right, is it can scrape the Internet, find publicly available information and look at its training base and it can give you sort of a reference point for your research and so to like for things like competitive offers. If you want to go see what your competitors are running right now in terms of ads and offers, there are all sorts of AI tools, especially if you can find what we call the agentic AI tools. These are tools like Claude Cowork that will actually go out if you have open claw. That's beyond my level of comfort. But these are tools that will actually do the work for you. They'll act as a human, they'll go comb the Internet and they'll do the research. So use it for things like that to speed up your process. But again, I think there's a human layer to even the market research and data component, which is all of the stuff that's either not visible online, right. You're. If you want to kind of do like a secret shopping, we've had people on the podcast that are talking about the advantage of secret shopping your competitors and going in and experiencing those practices, seeing what they're doing, getting a boots on the ground, analysis AI is not going to be able to do that for you. Right. If you're trying to figure out what makes the most sense as your right next move based on data and analytics. AI is trained on what's publicly available in generic data, not proprietary information. And so the more that you feed it proprietary information and knowledge, the better results you're going to get from that market data and research. So I think that's a good item. Number one is market research and data. Some awesome ways to incorporate AI into that. But. But major human element advantage, especially by feeding it proprietary data and doing the little bit of extra girl work to give it the information it needs to give you informed decisions. Anything to add to the the research layer, Lauren?
- [03:05]
B
I think it kind of goes with it, but have we talked about or rolled out the Notebook LM yet on the podcast?
- [03:11]
A
Not sure, honestly. Go ahead and talk about it. I'll. We'll make sure we throw it in the show notes.
- [03:16]
B
So Ricky has trained a tool called Notebook LM with all of our background and podcasts and YouTube videos and also a growth equal equilib equilibrium document that kind of walks through all of our things. And I think it's super helpful because you can put anything into that AI system and it'll spit back out an answer. But what's so important is that it is trained on an extensive number of videos, documents and things from humans directly. So it's not like somebody at one of our clients could just go into a random AI tool and get the same answer. It's so important that those things are trained in the background too, to be able to give you the right next step. So I think that'll be a really cool one to throw in the show notes or to have people reach out if they're interested in.
- [03:57]
A
Yep. And I actually have a version of that as point number seven, so I'll elaborate on that one a little bit too. At point number seven.
- [04:02]
B
Oh, cool. Okay.
- [04:02]
A
Perfect. Cool. Okay. Number two is copywriting and creative assets. I know Lauren. She mentioned she was on a call with one of our clients and we were talking about spinning together a graphic and the client said, oh, don't worry, I'll just do it with ChatGPT. So this is an amazing use case for AI right now, our favorite tool. And this is really just like very new for us using this as an actual image creation tool because they've stunk in previous months. They've like really been bad. So this is going to be a very time sensitive episode. I don't think this Episode will be evergreen because the stuff we're talking about today might not matter in six months. So if you've watched other videos where we talk about marketing secrets and strategies, that's always evolving as well. So you see in our feed and our podcast and our YouTube feed, like a summer 2026 edition of some of our top performing strategies and things that are working best. This is going to be like our summer 2026 version of the State of AI. But copywriting and creative are great. I think a lot of you have been using AI for copywriting, but again, feeding it context matters. So if it's just writing based on generic information and insights, if you've seen this, if you've even sent an email with ChatGPT or Gemini or an AI tool, and if you really understand the importance of nuance in language, you understand that you can't just throw in a prompt and say, send a message. So the same thing is true with copywriting. It very much needs a co pilot and a pilot to make sure that it's steering in the right direction and that you're iterating in a way that's going to align with your brand and with what you want to say and how you want to say it. Very important. So don't just spin up a bunch of crud AI copy, really train it, maneuver it, mold it to what you want it to be, use it to speed the process and to give you better ideas for images. Like I said. Go ahead, Lauren.
- [05:44]
B
No, I was just gonna say, I think the challenge with using it for copywriting too, not only is you have to train it to sound like you, but so many people can sniff out what is AI versus what is. Not so fast nowadays. Like, we'll see one graphic that we send even in our chat internally, we're like, oh, that's a ChatGPT graphic. That's a Manus graphic. So there are things that kind of still, you have to be unique with and still have that element of personality within it. And I think copywriting is a huge one because as soon as I read something, it's even from team members, they'll send me stuff and I'll be like, come on, do you have to use AI for that one? Like, that's important too. So you still want it to sound like your brand and like it's coming directly from you and your practice as well.
- [06:23]
A
Yeah. Like the example I have here in our podcast outline is you ask AI for a lip filler ad and it spits out generic emoji heavy Fluff. It says, get juicy plump lips for summer. Right. Like a human operator. Somebody that's going in and auditing this might say, rewrite that, remove the emojis. I want the language that speaks specifically to an affluent 45 year old professional. Focus on subtle natural volume and frustration. You have things that you want to say that are unique to your practice and unique to your perspective. AI is not going to capture that. You have to feed it so you can help again, it can help mold your final product, but the source material that you give it, your opinions and your perspective are going to be the difference makers. Think about this. Alex Hormozi has said that he doesn't think, he thinks AI is going to change the game in a lot of ways. But at the end of the day, it's the same thing it's always been, which is people with tools versus people with tools. It's just we now have a really cool upgraded tool set in the form of AI. But if I can do everything for you, it can do the exact thing for your competitor. And we're back at the status quo. It's the human element that's going to be the tiebreaker. So you have to understand this, use it to create better output. So when it comes to copy, that's a great use case for creating ad copy. Speeding up your automation workflows. So the 108010 concept, you can use AI for a lot of 1080 10ing, right? Feed it the initial 10% with a good prompt, let it come up with the bulk of the 80% and then re prompt it to get it over the finish line. With that final 10% you can use that same framework to create extra efficiency and to scale this. So for like creative. One of the things that's on our radar right now is we want our ad graphics to look really professional on brand. We want them to have authentic images of staff and providers. We want to be strategic with the social proof on the ad. We want to be really intentional with the offer that we're promoting and the service that we're promoting. But creating the graphics still has always been an hour, an hour and a half. So our version of prompting was the instructions we give to our design team. And that's pretty time consuming to even just draft up the instructions for a good ad. But then it's another hour, hour and a half for the design team to put together version one of a graphic that now gets chopped down to minutes, seconds with AI. And so what this allows us to do, especially if you've heard Anything about meta ads specifically and how important Creative testing is one of the things we hope we're going to be able to do here and that you can do for your practice is, is you can scale your creative to test more, more frequently with less effort. So we can have maybe instead of only running two or three custom graphics, we can put together 15 to 20 AI generated images in the same amount of time. So that's a cool unlock, great use case for AI. But again, that prompting part is going to matter a ton. Number three Campaign and content ideas so I can be a great ideation engine. It can give you options so that I'm going to be very careful about this word options when I talk about ideation. Don't rely on the idea that AI creates for you. Use it to give you a bunch of options at the ideation level. So for example, it can map out a 30 day educational content calendar. If we wanted to promote micro needling in our practice and we want a content calendar on microneedling and you feed it your page, if you can come up with a really good blueprint that you can sort of 1080, 10 here. But again, where the human element is preferred, AI operates in a structured vacuum. It cannot see the full context context of your business and your goals and your nuanced perspective. So you know how many of you, if you did this, would feed at your microneedling page as like a starting point. And then you realize that that thing is thin as crud. It lacks context, it hasn't been updated, it lacks nuanced perspective. There's going to be 97 things you're probably going to want to edit in that workflow. So be really intentional again about using that 1080, 10 principle if you're using it for ideation and to come up with those initial frameworks. Anything to add to that section, Lauren?
- [10:09]
B
I think a big thing I've noticed with that too is because I've been testing it with certain graphic variations like we talked about. Can we get 20 variations? And I think a big thing that AI does is it'll just come up with as much as it can and just throw it on different graphics and different copy and things. But it's not things of direct value. So it comes up with as much as it wants, but it's not things that are going to actually make an impact or make that next next step that people want to take. For example, like we were coming up with a weight loss funnel for one of our clients and it just came up with so much fluff that just is totally not what people care about or what they want to hear.
- [10:44]
A
Yeah.
- [10:45]
B
And so it came down to, like Ricky said, the 108010 rule. We had to prompt it with. No, people don't want to hear about that. Like, they want to hear real statistics that have to deal with why the weight loss drug can impact your hair loss or your hair growth and how can we overcome that. So it's so important still, even with content creation, to have it be able to back and give you more than, you know, what it just spits out as what it thinks.
- [11:08]
A
Yeah. And I think the cool thing with like ideation and content is if we can just get 50, 60% of the work done, it gives us a little bit more time to spend on things we would normally ignore. We built our first site recently with AI for a client, fully transparent. Our client knows that we did this with AI and it was pretty time consuming and we did it with multiple tools. I'm not even sure we save time at the end of the projects, but part of that is just using multiple different tools together and patching it and iterating the crud out of it. But one of the things I noticed is we didn't have to spend as much time with a designer on design elements. So I was able to spend a little bit more time on actually helping our client do something that our clients never want to do, which is go back through their site copywriting and content. We were actually able to feed it again, context and feeding it good information. We collected from the client, a transcript interview. So we were, the client was just talking and giving us a bunch of good information and context on what they do and why they do it and how they think of things. And it was the doctor and the injectors having a very long conversation. We were able to feed that transcript. They gave us their four pillars, which is like their framework for how they treat and serve clients. They gave us their core values doc. They gave us their like SOPs and their consultation SOPS so we could feed all of that to the AI and use it to create better copy on the site. So that's something we would never been able to incorporate into a website package before. We're able to incorporate, incorporate it with the use of AI. So another good use case there. Today's episode is sponsored by Gloss. Genius. Gaps in your schedule and patients who don't rebook. Tight margins, high fees, and a clunky EMR holding it all together. Sound familiar? You'd love to be growing your practice, but you don't always have the time to figure out your right next move, let alone to sit down and eat lunch. That's why I want to tell you about Gloss Genius, the System More than 5,000 owners use to grow their practice in the background. Unlike others that take more effort on your end, Gloss Genius automatically fills your calendar by getting you more traffic, more reviews and more online bookings. It finds hidden revenue for you using AI and auto sorts your treatment photos with a HIPAA compliant EMR your team will actually enjoy using. And Gloss Genius's built in point of sale doesn't just take payments, it earns you more every day, rebooking patients at checkout, increasing what they're willing to spend and saving you thousands with the lowest flat rate and zero hidden fees. Ready to earn more and take back your time with the software that does the work for you? Use code strategies@glossgenius.com MedSpasuccessStrategies for 50% off your first two months of any aesthetic practice plan. That's code strategies@glossgenius.com MedSpasuccessStrategies hey practice owners and marketing directors interrupting this episode to invite you to schedule a one on one strategy call with me to discuss how we might be able to improve and level up your digital marketing efforts. So we're rated five stars on Google, we're HIPAA verified by Compliancy Group and we have a track record of taking clients from 30,000amonth to $120,000 a month and adding multi millions of dollars in additional revenue for some of our bigger multi location med spas from more effective marketing strategies. So on the free strategy call it's really educational. I basically spend an hour going through detailed reviews of all of our best performing plays that we run for our clients. You have it to take and run with it if that's what you want to do and if you think it might be a fit to work together then we're excited about the possibility to partner with you. But if you're interested in better, more effective digital marketing solutions for your Med Spa, visit Medspamagicmarketing.com that's Medspamagicmarketing.Com to schedule your your one on one strategy call would be all right. Number four on our list is ad Campaign deployment and optimization. What AI can do
- [14:52]
B
Ricky, I think your audio just went out.
- [14:56]
A
I don't know what's going on with that. Can you hear me now?
- [14:58]
B
Yep, you're back.
- [14:59]
A
Cool. Did you hear me say number four?
- [15:01]
B
Yeah, you just said number Four. Yeah.
- [15:03]
A
Okay, perfect. Number four is add campaign deployment and optimization. So what can AI do here? So there are modern AI tools, some, some native to your ads platforms, some third party tools that can do things like automate your bid adjustments. You've had that as an option in Google forever, to be honest. Build new audiences, run meta and Google performance, max campaigns, scale budgets. We I just got off a call with an agency group that we're in today. A lot of people were using a tool called Victor V I K T O R to do a lot of agentic. Again agentic just meaning the AI actually logs in and does some of the work for you. They were using it to build the initial frameworks for Google Ads campaigns so they could preload all of their SOPs, their frameworks, their keyword list, their copy and Victor can go in, get into the ads dashboard, build everything out and then all you have to. They weren't letting it deploy yet, but it was in there as a draft and then it had a human QA person on the back end making sure everything looks good and launching, but just speeding up the process dramatically. So another really good use case for potential AI deployment to just make sure that the actual building of campaigns is quicker. Building your ads assets and managing some of your online dashboards.
- [16:16]
B
With that though, I think having the human eye kind of as the next step of this is the number one importance there because there is so much that these dashboards will recommend as great suggestions or solutions that are going to make your campaign so much better. And it's funny because a lot of times our clients will come to us and say, oh, I got this pop up from Google and it's saying, you know, turn on Broad Match because it'll be cheaper and that'll increase our, it'll increase our number of clicks, et cetera. And then we have to take 20 minutes to explain why that's not something that we want to do. So it's still incredibly important to have that human eye then on the campaign or on the build and make sure that we're approving those changes or knowing why we're making those changes because it really can make or break a campaign. For example, if we allowed that auto recommendation to turn on the Broad Match keywords, I'm sure you've heard us talk about it. If you watch the podcast before, you will have so much junk search that's coming through that you're basically wasting all of your budget. But you're hitting Google's benchmark of conversion rate looks great or your Cost per click is super cheap. So those kind of things that the system wants to prompt for aren't necessarily things that a human is going to say. Yes, that's something that we want to do in our campaign.
- [17:22]
A
Yeah. The advice, I would say, like the advice from AI is less reliable. It's really the use case across the board from this list that we're going through is once you make a really a decision based on full context and full data, then the AI is good at just doing the thing for you faster, actually just going in and pushing the buttons and running through a workflow. But it's again, like Lauren said, be careful. Like you even look in Meta ads. They'll do things like the AI tool, the AI recommendations and something like Meta are telling you to turn on Advantage plus Audience and audience expansion that might reach people that literally aren't even in your local area. But Facebook knows that there's a lunatic three hours away that fills out every form that pops up in their feed. And you've told Facebook that your goal is lead form submissions. So Facebook goes out and finds lunatic three hours away to fill out your lead form. Right. So you do have to be very careful with, with like the AI recommendations. So the thing that I had put on the ad campaign deployment and optimization number four here that I wanted to go over was like the football analogy. I love sports analogies. If you've listened to the podcast, but in football, for example, there's not really a formation or a play that is so novel that every other coach in the league doesn't know that it exists. So if every coach has the same information and the same options in terms of plays and formations, why is there such a difference in when you call a play and at which time? Well, that's where making the call makes all the difference in the world. And having a human make the call. Get a list of ideas like we talked about, but you have to have that qa, that person that's making the call that's going to be responsible for the decision on the back end. Cool. Okay, so number five is our tech stack, connections and automations I can handle. Again, caveats here. Well, we're going to, we're going to talk about this a little bit, but it can fix speed to lead issues. Right. There are all sorts of AI assisted chat bots and tools that will converse with your lead, so they're not waiting around for three hours. A lot of you maybe have heard of podium or you've used the leads AI in high level. I don't know if some of these other tools, Lauren, have AI, but a couple things that are the problem here is AI has no accountability when things break or if something is wrong. Right. Who's going to be accountable for this? At the end of the day, if the system glitches, it misinterprets data, it messes something up, you're going to run into major challenges. So if you're using this in your tech stack or in your automations, you, you need a human systems captain at the very least to monitor and to fix it and to train it to continuously improve, especially for something like leads management. But even then, for leads management, I think it's the human element in the context is the thing that makes this work because we've had clients that have tried it. Almost no version of this is plug and play. And that just speaks to the idea that these AI tools are not super valuable off the shelf. You have to train them extensively to do what you want to do, even if that's ad graphics or images or in this case, your tech stack and your automation workflows.
- [20:27]
B
The other thing with AI leads management that I've been noticing lately too is you lose out on so much information that you could get from a person actually doing that role. For example, somebody might be really confused about the way that a promo is laid out, or somebody might be actually really far away and something's wrong with your targeting that you wouldn't have caught otherwise. Like there are certain things that you'll be able to pick up on and gather through conversation that you wouldn't have otherwise if you just had a manual tool or an automatic tool taking care of it. Example, we ran a weight loss campaign where we tried a first month for $40 and then the next month was $250. So there was a ton of disconnect between, well, why is the first price so cheap and the later price isn't. We ended up restructuring that campaign because of how many people ended up responding with that confusion. So that's something we would have never picked up on if we weren't actively in there. People are actively in there managing those campaigns and responding to people.
- [21:21]
A
Yeah. And keep in mind, I think one of the coolest parts about this AI revolution is you're going to have an opportunity to differentiate with something that was just the status quo before. So there's two sides of every coin. And as more and more people lean into AI to create more efficiency and to increase speed to lead and to create better consistency and communications through leads management, well, just Producing a white glove, authentic human to human interaction, is going to be a point of differentiation. The thing that used to be a default because there was no other option is now a point of differentiation. So when people start to wonder, am I talking to an AI? You know, I still think at the end of the day, med spas are a service business. There's a lot that AI is going to be able to do in these conversations and some of you are going to feel comfortable with it, and that's totally fine. And especially if it's trained well, I think it's going to be a very valuable tool and it's going to have its place. I 100% believe that. But at some point, if you're able to have your leads conversing and connecting with a real person at the business, it's now a point of differentiation just to say in your workflow, hey, it's Lauren, by the way, I'm a real person, not. I wanted to see if you were looking at the schedule for next week, if you wanted to schedule that appointment right now. People are. There's something called the reciprocity principle, and it's that if it's essentially if someone does something for you, you feel an obligation to do something back for them. Well, that can even be boiled down to the communication layer, where people feel more of an obligation to respond when a human is making the effort to reach out than if it's an AI engine. If people know or start to feel like this is just a computer, even if the computer is so good, you can't tell, but they just start to assume by default that it's the computer talking. People really don't feel any sort of obligation to get back to you and give you a status update. If it's a real person. Hey, Lauren. A real person. Noting that in the chat thread, you create more of an obligation to respond, which I think is going to increase results. So again, it's not saying that AI is not useful. I think that it's definitely is can handle the bulk of this, but you have an opportunity for differentiation by building in a human layer to that as well. Hey there. Wanted to briefly interrupt the episode to make a quick ask. If you're a podcast listener, it would mean the world to us if you leave a review for the podcast, whether that's on itunes or Spotify. It's something I hadn't really remembered or thought of asking for, but it does help us show up more frequently so that we can reach more people with the information that we're Providing. So it mean the world to us if you'd leave a review on itunes or Spotify, if you're listening on audio, if you're watching on YouTube, make sure to hit the subscribe button so you're in the loop for future videos and you don't miss any of the content that we're putting out. A couple more. So number six on our list is scaling visibility and client acquisition. So we talked about all of the things that it's going to be able to do to help you build more ads, create more creative variations, streamline, copy, create workflows, give you seed ideas for whatever marketing and promotional things that you want to do. If you feed it the right context in 1080 10, it's going to make you a lot more efficient and a lot faster so that scalability will become easier in terms of scaling client acquisition efforts. But what it's not going to be able to do is fix the things that are going to inevitably make your marketing less, more or less effective, regardless of what AI is able to do. And we see this over and over again across our client roster, which is the businesses that are dialed in at the service level see the best success and the best results from their marketing. And there's no version of AI that's going to fix this for you. And so when we look at our client roster and we have clients that are getting, you know, new bookings for 1:50, $160, and they're thrilled with the results. And then we have other clients that are getting 156, 150, $160 client acquisition on ad spend and they feel like they're in a cash pinch and everything's a struggle. We start to look at things and things fall apart. At the consult the service layer, it's how are your consults handled? What are you doing about your rebooking process? What is the difference from your top provider to your mid tier provider in terms of retention and client satisfaction? Those things make all the difference in the world. Those are problems AI is not going to solve for you. And marketing at its finest is just going to be fuel on the fire of an already successful business. It's putting a bandaid on a bullet hole of a business that's struggling foundationally. We just had Hayley Wood on the podcast. She Talked about her A Plus providers have something like 90% retention in her practice. Now again, your source of client acquisition will impact that, no doubt, but for her it was 90. The difference just in provider quality from the same types of patients was 60% with certain providers all the way up to 90. And her practice will build and thrive, build momentum and thrive with the 90% providers on the people that have 60% retention. She knows it's going to be a struggle and she has to coach those people up until they get to the A plus caliber. The difference between A plus and A minus teams is all everything in this industry and everything in this space. And AI is not going to fix that for you. So if you get a bunch, if AI writes the perfect ad and helps you spit out a bunch of different creative and test and get your cost per lead down and it doubles the amount of new bookings in your clinic. But you have a rude receptionist and a subpar like in office experience. Your waiting room is kind of sloppy. It's not super professional. The providers are just a little above average, but not exceptional. Everything will fall apart. So humans fix operations. AI is just here to help amplify what's already going to work. I know we've talked about that extensively. So PSA number 976 on that. Anything to add to that, Lauren?
- [26:47]
B
No, I think that's perfect.
- [26:49]
A
Okay, now the last one is the strategic brainstorming and context layer we just talked about. So AI can synthesize and analyze a massive amount of data if you feed it custom context. So we've, we did this. For example. One of the things that we try to do on our quarterly reports for our clients is we talk about the reason we don't like these patient guarantees. You see, a lot of marketing agencies do is it's optimizing for the wrong data point. It's just optimizing for the leading indicator of butts and seats. I talked to a person a couple of years ago that ran a facials business. She was an esthetician, didn't have injectables, which I think is a challenge in and of itself here. But she was an esthetician doing facials and she worked with an agency that was guaranteeing her 30 booked and paid new facial clients every month. And she scheduled a call with us to talk about marketing. I said, well, are they getting you the 30 every month? She said, yeah. I said, so I'm confused. What, what is the problem was? Well, of the 30 that come, they're coming from weird areas. They're really like the fringe market of discount shoppers. And my retention, my rebooking, my cross selling is virtually zero on these people. But as far as the agency is concerned, the incentive structure was to get 30 bucks and see so you have to make sure that you're feeding this data. Like one of the things that we do as a result is to make sure we're optimizing for the data points that matter. We feed data multiple different reports into Gemini to give us specific findings so that we can calibrate strategies to offers that are not just getting butts and seats. Butts and seats is important, but how does that actually play out in terms of retention and lifetime value over time? AI is really good at synthesizing and giving us that data if we feed it the right information in the right context. But the human element is still super, super important because you're going to have to 1080, 10 that you have to feed it the right data, get the outputs in a way that makes sense to give you the capability to make informed decisions. And if you don't make informed decisions with this data, it's not super useful. So having a pilot that tells these things what to do, how to talk to each other and how to find the right answer to the right problem so that you can make the right strategic adjustment makes all the difference. We talked about the notebook. LM Lauren mentioned it earlier. This, this thing that we have. So I'm going to make sure that's in the show notes. If you have not heard from us about this before, I strongly encourage you go check this out. It's completely for free, it's ungated. You will just have direct access via the link. The answers you get from that tool are going to be dramatically different than the answers you get from just going to ChatGPT. And one of the major reasons that I think this is really important is because ChatGPT, Gemini and all these other AI tools at a default off the shelf layer, they don't know what good and bad advice is and that's dangerous. Go back to the email examples. I was just using AI. I think all of you, if you use AI, have had this exact experience where I told it to do something. I gave it what I thought was pretty good prompting that first 10%. It gave me back something that just didn't make any sense. Like it was just, I think, foundationally incorrect. And I pushed back and it said, oh yes, you're correct, that wouldn't make any sense. I just remembered what it was. It wasn't even work related. I was planning a golf trip, trying to plan a golf trip and I was going through all the different options of trying to make this work logistically and it gave me this itinerary and I said, but that wouldn't work if this, this and this. And it's like, oh, yeah, you're right, that was a horrible idea. And so at the end of the day, you need someone to feed it the right information and to help make the right call. So, like, when you use our notebook LM that's trained on hundreds of our internal SOPs, our internal documents and our playbooks, the things we've talked about on this YouTube channel and this podcast, not only internally with Lauren and I, but some of our guests, top performing, top experts in the industry, that's going to give you information that we have conf confidence is the right next move that fully digests all of the trade offs. Okay, so we're going to do a little exercise. We're going to look at a generic Google Gemini result and answer verse like the Notebook Notebook LM that we've created from the training of all of our podcast episodes and YouTube videos and some internal documents to show why context matters so much. So I'm going to paste this question word for word, and the question is, my cost per lead is high on my meta ads for my med spa Botox ads help me fix it. And so the generic version of Gemini is referencing a multitude of sources from multiple different companies. And the problem with this is you don't know if the data that backs these recommendations is legit. This is a major problem with AI. It's trained on information that's publicly available with no ability to verify the effectiveness of the strategy. And so I spoke at a conference a couple years ago and I spoke right after someone who was preaching a fundamentally different strategy than we typically preach and coach on in our marketing advice section. And when I talked to the owner of this conference, I was able to provide data that shows, hey, here's what I'm actually seeing in my data. Hers was all theory. Why would you compare hard data to someone's theory that has no evidence to back it? And that's exactly what these search tools or these AI tools are going to do by default. So it sounds really good and some of this is good, but the problem is if 85% of this is right and you use the whole thing and 25% of it is rat poison, right? Like a pizza that's poisoned with a few drops of cyanide is still going to kill you. And the same thing is true with the bad advice from these tools. So it gave some recommendations here that are pretty good. Like it said, Use lead forms, native lead forms, instead of landing pages. Your offer might be too generic. So it was giving some like decent introductory advice on where to look. But then the specific recommendations might make absolutely no sense. Like soften your language. Swap Botox out for wrinkle relaxer treatment. That's the exact opposite of what we found to be effective. Your offer is too generic. So in this example it was very it was a get 20 units of Botox plus free skin care consultation for only 199 new patients only and then bank your Botox lock in $10 a unit for the next 12 months when you quit. They have no idea whether offer a version A is better than version B at all. And I would be willing to bet just from first glance that probably was the opposite of the truth. They give you an idea on different types of creative to test, but if we actually have data to know which creative version is going to perform, we don't do this. A video walkthrough of a treatment, a selfie review, a simple high contrast text card. Let Meta's algorithm decide which one gets the lowest cpl. Again, you're kind of optimizing for the wrong point too if you're just looking at that. So this just lacks so much context on what the right decision actually is. It says like stop doing over produced graphics. And then it I don't even know where it's pulling some of this source material from, but it's a hodgepodge. I'm not going to list the sites, but I can see just in the first section there are three different links that it's referencing for information that it's putting together in one answer that might be completely contradictory. So if I go into our tool and I ask that same question, you're going to get a framework that has some data behind it, right? If your Meta ads are pulling high cost per lead, you have a top of funnel bottleneck. While a highly optimized Botox campaign could yield 7 to $15 and sometimes lower, a CPL creeping up to $30 means your campaign might be bleeding cash and driving up cac. In our growth equilibrium framework, top of funnel CPL is a direct reflection of marketing friction. If your leads are too expensive, the digital friction of your ad. Here's our agency's blueprint and it goes through an entire thing. Audit the offer and its price points and it's trained on the data and the nuance of everything that we've ever run for clients for four and a half years and the understanding that modifying one part of this funnel, improving cost per lead might come at the expense of another attribute. So it also gives you a whole coaching on the context of you might do these things to lower cpl, but you might create this other secondary problem on the back end and so the full context matters a ton. I hope that gives you a good idea for the use case of AI. I think the big picture of this is AI used well is a classic case of 1080 10. You need to feed it the right information, the right coaching and the right vision. Let it help you with the workflows, to streamline and maximize your outputs and to do things faster and then have good qa. Have somebody that's responsible to make the call and to iterate, just like you do when you ask chat GPT to write an email on the back end and I think you'll be better for it. If you're using AI in any cool, innovative ways, let us know. That was just a little bit of a look on some of the things from a marketing standpoint. I know there are so many cool use cases for AI. We've been using Claude cowork a ton, so if you've got anything cool you want to share in the comments, share it below. We'd love to make sure everybody else has visibility and we're sharing ideas there. Anything to add Lauren for we wrap?
- [35:32]
B
Nope, that's all awesome.
- [35:33]
A
Well, thank you all. I hope that was a good little insight as we stand here in summer of 2026 on use cases for AI and the importance of the human layer. Thanks and we'll see you on the next episode.