Welcome to Social Geek, the most popular franchise podcast!
It doesn’t matter what you do with AI.
There…I said it. Come and get me, snake oil salesman.
What DOES matter is what your candidates and customers do with AI and how they’re finding you.
It’s the Marketing & Innovation Update for October on the most popular franchise podcast, Social Geek!
Patti Rother, Founder and CEO of Root + Rise, is here to help Fran Dev pros feed the Large Language Models with answers for their brands.
We’ll cover:
* Why Reddit is now so important and how you can get started
* Using Wikipedia as your guide to sharing your facts and data
* Other places to share your brand’s unified messaging to gain authority
* How video content is currently feeding answers
* Why PR and earned media are more critical than ever
Shout outs to our all-star team members Jacqui Blair, Islam Tawfik, Willem van Daalen, Madison Leavy, Grace Gorman, and Jason Weaver.
Next, I’ll chat with Jeffrey Leach of Vendasta to dive into AI for local franchisee marketing.
Jeff shares the results from Vendasta’s new study of 20 franchise brands looking at their franchisees’ local AI marketing services. There are some amazing conversions via web chats, SMS conversations, and booking services.
Connect with Jeff to learn more about the survey results.
Thanks to Northeast Color, Hughes, and Thunderly ⚡
Transcript
AI was not used in the creation of this podcast. This is the Social Geek Radio Network.
Hello, Geeks, and welcome to the Social Geek Podcast. I’m Jack Monson, your host and marketing co-host, Celia. It doesn’t matter what you do with AI.
There, I said it. What does matter is what your candidates and customers do with AI and how they find you. I’m chatting with Franchise Pro Patti Rother on the AI effect on Fran Dev sales, and Jeff Leach on AI for local franchisee marketing.
All right, today’s show is brought to you by Northeast Color, Hughes and Thunderly. If you’re planning your 2026 Franchisee Convention, don’t forget to shoot some video. 65% of brands don’t have franchisee testimonials on their website or YouTube.
Don’t be them. Bring in the best storytellers in franchising to film your franchisees. Contact Thunderly at thunderlymarketing.com.
Patti Rother is the Founder and CEO of Root & Rise, Franchise Development. She’s back to talk a little Fran Dev Marketing and AI. What’s going on, Patti?
How are you, Jack? It’s so good to see you.
We’ve had some really good conversations dating all the way back to the early days of ChatGPT, which was, what, God, three years ago now, I guess.
First time we’ve been on a panel at IFA together. About AI.
Yeah, a big anniversary of that coming up, I guess. So I wanted to check in with you on what you’re working on these days, specifically when it comes to Fran Dev. Over the summer, my team, and a shout out to Maddie, Jason and Grace, it seemed like they were the only people I knew in marketing or Fran Dev that were looking at Reddit as one of the places that all brands need to be.
Then I chatted with you a couple of months ago and found out that you are going down the same road we are. So let’s talk a little bit about Reddit. What are you seeing right now when it comes to the importance of Reddit for not just Fran Dev marketing, but for any type of marketing in AI?
Yeah, so it was really interesting. I actually posted about this a while back on LinkedIn, talking about how OpenAI specifically, but a lot of the large language models are pulling their information directly from Reddit and Wikipedia, as well as their main sources of information. That was wildly surprising to me at the time that I read the article.
I’ll make sure we have a copy for everyone to read it themselves, but super interesting. I started to dig in as to why that might be, because we know that when ChatGPT specifically was trained really early on, it was basically trained on books and articles, and the Internet, and web pages, and things like that. Everyone’s been trying to figure out, you know, what’s the silver bullet to make sure that when people ask questions into these large English models, that somehow my brand shows up, and everybody’s been trying to figure it out.
So once I found out that Reddit was a main source of information, I figured I should dig in, because if that’s the case, I need to figure out exactly how that’s working, and why it’s so important for specifically OpenAI to be linked into Reddit and Wikipedia. So that’s where the conversation kind of started with you and I as well.
It’s a different type of social channel or platform in that you can’t really go on Reddit as, you know, Jack’s pizza franchise and start talking about how great your pizza opportunities are, right? It’s got to be a little bit more, I guess, genuine, maybe a little more nuanced as well.
Yeah, I mean, for those of you who don’t know what Reddit is, it’s essentially a place where real people talk about real problems. So you can post a question or a dilemma that you’re having, and everyone has the opportunity to answer and support and comment. And so these long threads of conversations are happening, and it’s full of this unfiltered lived experience data.
And so OpenAI specifically, large language models are so interested in this because it’s teaching the large language model how people ask, answer and decide things. And it’s gold for training AI to sound more human and understand context. And so when OpenAI essentially has that direct link into Reddit, which is what it is, it’s actually not just scraping the information for context, it’s being fed into the models with permission, which obviously is also raising accuracy and ironically reducing bias, which is also super fascinating.
So it is a direct link and means that the way that it’s being fed into these large language models is wildly different than the way that it’s kind of scraping information and kind of finding out. I always tell people when I’m trying to explain what a large language model is in general, I’m like, I always picture this like mini librarian. And when you ask a librarian a question, they like go throughout the library and they grab books and articles and all these things and they do the research for you and then they form this unique answer based on everything that they found out.
The thing that’s missing is the context around why you’re even asking the question in the first place. So with Reddit, the difference might be like, if I Google and say, best dog breeds for children, I’m gonna go through all those articles, I’m gonna read different things, but they don’t know the age of my kids, they don’t know where I live, they don’t know if I work outside of the home, how many times I can walk said dog, and so there’s just missing context. But when I go into Chachi BT, I might say something like, you’re an expert in dog psychology and dog breeds.
I’m thinking of getting a dog for my children for Christmas this year. My kids are nine and seven. We all work in the home.
My kids love going outside, so walks should be fine. And we’re looking for a dog that maybe is on the smaller side instead of something large, right? So I’m going to give it way more context.
And that’s why something like Reddit is so valuable, because on Reddit, those are the same types of questions that people are asking of the community. It’s very different from Googling things or using a search engine. Now I’m trying to find a very specific answer.
And so the link makes so much sense once you dig in, because that’s exactly what these large language models are trying to do. They’re gaining all this context, all this information about the asker, and they’re trying to figure out what the best answer is for them. And in that sense, if you and I both ask the same prompt, it would give us wildly different results, because it knows all this stuff about us already from a previous conversation.
So it’s so fascinating.
And someone who asked the same question about a dog, but maybe they live in an apartment in Manhattan, they’re going to get a very different answer than what you got. Something you said just jumped out at me about why Reddit is so important, because that was one of the things that I’ve been trying to wrap my brain around, like, why Reddit? You know, and we’ll get to Wikipedia in just a minute, which is even a bigger question, but why Reddit?
And I think you said something about open AI is looking at real, genuine conversations. So in essence, they’re kind of looking past all of the marketing speak, that they might see on something like Facebook, which has a very low score, I’m guessing, when it comes to citations by the large language models. They’re really looking for those real conversations.
And I guess Reddit is about as real as you can get.
Exactly. Yeah, it’s the new gold. It’s where people are solving real problems, but it’s in plain language.
And that’s exactly what these large language models are trying to do. And they treat it more like a knowledge channel instead of an ad channel, right? Like, who worked the system the best to get that article to be the strongest authority versus what are people actually asking the most about?
And what answers are they liking? Because in Reddit, you’re also benefiting from knowing the top answer, the one that has the most interaction, the one that’s been upvoted is the one that most people are aligning to. So then OpenAI is also benefiting from that and understanding how humans are interacting with all of these conversations and they’re essentially telling the AI tool what they found most valuable.
And therefore, OpenAI can pull that in and say, most humans who are having this conversation are benefiting from this type of answer, this type of context.
What advice do you have for anyone in franchising that just has not touched this thing, or maybe they haven’t touched it in five years like a lot of people have, and kind of diving back into Reddit and engaging with people, not being salesy, but engaging with people and trying to lift our brands up? Where do we even start?
So I think there’s two different approaches. The first one is obviously get into Reddit in some capacity. It’s super easy to sign up.
You could have a very generic username. I’m testing out various things where I’m kind of like a wellness influencer who knows all the wellness trends because I’m working with Perspirus on a studio, let’s say. But I’m also in there as a franchise patty, somebody who knows the franchise industry and can really answer things about entrepreneurship and the whys and how and all of that.
So the first one is just to get in there, figure out the platform, start answering questions and having some kind of authority in the space. But the second thing, and I think actually the more important thing is understanding why somebody like OpenAI would have a formal licensing deal with Reddit. So the same thing that they have with Google, they have with Reddit.
And the reason they do is because they’re looking for that high intent, fresh problem solution text. And what that tells you as a brand is that in every place that you are online, you should be thinking exactly the same way. So even if you’re not on Reddit, even if you don’t have a marketing team that can go and comment and answer and figure out exactly how to get authority in this space, you need to pay attention to the fact that large language models are pulling in this information because of those three things.
And so you want to be able to feed the data or the models with the same kind of like credible human Q&A and lived experiences. And you want to be showing up in that same capacity. So if you think about your website, like this is a huge one.
You have to have structured, linkable content. You have to have Q&As and FAQs and figure out like, what are people talking about on Reddit? And how can I answer it in my own earned or own space, right?
If you think about the articles that you write and the PR that you’re pushing, how to also have it be pulled in to AI the same way that they’re pulling information from Reddit. And so it’s just a clue to what’s coming. And I think that’s where a lot of brands are kind of missing the AI train.
They’re focusing on the thing, right? Like, oh, I got to learn ChatGBT. I got to be on Reddit.
And you’re like, well, yes. And pay attention to the signals it’s giving you about what’s coming in the future on how to also get fed into these models.
And that same way, good advice there. Whatever people are talking about in your industry and specifically maybe in your category or sector, what people are talking about, that should be part of your frequently asked questions on your website. We used to always say those FAQ should be all about what are your customers asking you about the most?
Like make a list and let’s make an FAQ that has all of those things on it. Now we need to look at what are people asking and what are people talking about on Reddit because that really needs to align with your website. That’s a great point.
Thank you for that. Now let’s talk about Wikipedia because I got to tell you, I got a problem with Wikipedia. I’m a little nervous about how important Wikipedia has become to the large language models when half of the time I’m looking at something on Wikipedia, I can point out the mistakes.
It’s as problematic as AI sometimes. If Reddit is feeding about 40 percent of when you see linkable content, 40 percent of it will be Reddit. Twenty-six percent of it is coming from Wikipedia, which again, you and I both know, Jack, those are human-built, anyone can edit, it’s open, and so things are on there that are wildly inaccurate.
The other thing that’s difficult is to get on Wikipedia is actually not super straightforward or easy. So if you don’t have a Wikipedia post or link or landing page, that becomes an issue. I think when OpenAI and AI in general is looking at Wikipedia, it’s more about the canonical facts, the historical play-by-play is really what they’re going to Wikipedia for.
And then what’s really important is that it’s machine readable. So the structure of Wikipedia is exactly what large language models need in terms of how they learn, how they feed the model. And so kind of the same thing as Reddit, right?
It’s like the Wikipedia itself is important, but it’s not the most important. The thing you have to be paying attention to is why is OpenAI and all these other large language models seeing this as an authority? They’re not actually seeing it as an authority.
They’re seeing it as a very easy, scrapable, learnable way, right? And so if you look at a Wikipedia article and you’re thinking about your own website or lots of pages within your website on how you want to do long form copy, Wikipedia should absolutely be the thing that you look at for how to format long form copy because it is so easy for AI to read it. Because of the structure and the formatting.
So even if I am thinking about like, it’s not going to be correct, it’s still going to be seen as a top source because it’s easy for the AI tool. So what I’m thinking about is how do I also make everything that I create easy to read for an AI tool and structure it in a way that literally is this like machine readable structure.
Yeah, still the format from Wikipedia because it feels like that’s, if that’s what the large language models, if that’s the way they want to read your stuff, then terrific, we’ll put it into that template. I always thought that one thing that Wikipedia got right was even for humans, it’s readable, right? If you look at any, pull up any movie on Wikipedia and your favorite movie, it has the same format for every movie, right?
There’s a little headline, then there’s the, a bit about the production of the movie, and it goes into the plot of the movie, then the cast of characters, and then the awards, right? And it’s like, maybe think of your brand like it’s a movie, and you’ve got pre-production notes, you’ve got the plot, you know, this is what we do, then we’ve got the cast of characters, which is who we are, and then we’ve got the awards and recognition at the end. And every movie on Wikipedia has the same structure, maybe every business should as well.
Well, and the way that I’ve been thinking about it too, is if Reddit is just that ongoing conversation and that real human information, like Wikipedia is just that reference shelf. It’s where I can go to check a fact, see a timeline. The timeline is actually the most important piece of Wikipedia because it’s always like past, present, future.
It’s just a list of all the facts about what happened, when it happened, and what order it happened in, and you can just confirm definitions as well. So if I think about what am I learning from Reddit and why AI and large English models are so attracted to it, and then what is it about Wikipedia? It’s like this is human conversation, ongoing constant learning.
This is reference, just quick fact pulls, quick timeline checks.
And that reference could be for franchise brands. It could be things like growth and acquisition and number of locations opened. You know, by 2025, we had 75 locations or something like that.
You know, who the CEO is and maybe a little bit about the brand and what they do and products and things. So yeah, it is…
I think sometimes we overcomplicate things when we’re trying to like do marketing, right?
No, not in marketing we don’t!
We try and make it like sound really cool. I always go back to Donald Miller because he’s such a king at this of just like, keep it simple, stupid. Like it has to be so simple and Wikipedia is that.
Like it’s not the greatest thing to read. It’s not the most exciting thing to read. But if I’m just trying to figure out like a brand’s history, I’m gonna get it done very quickly and very simply.
It’s not gonna be like…
Where is headquarters? Oh, Kansas City, got it, right?
Exactly. Like I shouldn’t have to be like, in the small town of Main Street, back when the tumbleweeds were blowing through, it’s like, just say Kansas City, gosh, Mike.
And that’s the thing, you go to most companies’ websites and they don’t say where their headquarters is. Because anymore they don’t want you to show up there. So you have to go to something like Wikipedia to actually find out where does this go?
From Social Geek: Marketing & Innovation: Franchising AI Updates from Patti Rother,
Where are these people at? Like what’s the city of origin? That kind of thing, so.
Well, and sometimes we forget some of the simplest information that people are actually seeking. And that’s also so interesting. When you look at somebody’s FAQ page sometimes, I’m like, they’re missing the first five questions that someone would maybe be asking.
Because when we’re in our own franchising fishbowl and we all understand the context and the world in which we swim, we forget that people aren’t in the fishbowl with us and don’t know key terms or information. Some of the conversations that I have with first time franchisees on those initial calls, I’m asking them very simple things because I need to understand the context that they have before I dive into the actual brand. And I think sometimes as brands, we just forget that.
That where headquarters is might actually mean something to somebody and it might seem so inconsequential to us.
Very good. So we’ve talked Wikipedia, we’ve talked Reddit. Are there other places right now where some brands just aren’t spending any time or resources that maybe we should think about being present on any other platforms or places?
Yeah, it’s been super interesting to watch this. Some of the things that are interesting, especially if you’re a brand that’s trying to be like a, I hear so many brands say like, we’re a technology company disguised as a blah, blah, blah, right? If that’s the case for you, think about something like Stack Overflow or Development Forums, because there’s a direct OpenAI API partnership with Stack Overflow.
It’s piping all of that technical knowledge into the models too. If you have a proprietary app and you actually have developers on your team and you’re doing all of that work, think about there. News publishers with licensing deals, like PR is more important than ever, which is probably music to your ears.
But there are a ton of deals now with things like News Corp, Financial Times, the AP. And if your brand is earning coverage there, those mentions are more likely to surface in the large language model specifically. So like PR and thought leadership, we’ve been hearing about it for ever.
Very few brands are actually doing it or taking advantage of it. Now is more important than ever because opening eyes specifically is striking deals directly with those news sources. And things like Financial Times, probably not an outlet that you would ever think about as targeting.
When we’re thinking about franchise brands, like we go to a lot of the franchise specific outlets or places, maybe like entrepreneur, things that make sense. But trying to think something outside the box about like how would I get on that specific news outlet?
Yeah, Financial Times and Wall Street Journal, always important, but probably more important now to the large language models than just to the people who may actually read those publications still.
Well, and it’s so important to stay up on all this AI news. Like if you don’t have a great rundown or something that feeds into your either email or something that’s telling you all of these things, paying attention to who open AI has formally struck deals with is so important because it’s just telling you where to be and who to pursue only if you’re trying to get into those large language models, which we see the behavior of humans is already starting to shift away from search and into answer. And everyone’s just wanting the easy answer.
They’re unfortunately not doing a lot of research to verify that answer, but they’re not wanting to click through a bunch of links to try and form the answer themselves. We’re just all to instant gratification at this point. So if you know that people are going to stop searching on search engines pretty soon and it’s already happening and they’re going to move into just living in their ecosystem of whatever AI tool they like best, then it’s really important to pay attention to who these people are striking deals with directly.
The other thing is just like open crawl open license places. So publish white papers and documents with permissive licenses. This is a huge one.
So many brands use kind of like a lead form or a paywall or something to block content. And that used to be really important five, ten years ago. Now it’s actually blocking that information from getting into the tools.
So rethinking that strategy for a lot of brands will be really interesting. But they want this like information that’s really pertinent and kind of the formatting that we were just talking about. But like clean metadata, stable URLs, all of that’s going to be really important for the large language models and the search engines.
And then your own site, like we’ve talked about it, but you should treat your data like an API for large language models. That should literally be the way that you’re thinking about it. So your FAQs, your Q&A markup, product and spec pages, a lot of people are missing like detailed information about what it is that they sell or their membership models.
Pricing, this one’s super interesting, Jack. Like if I’m in Oklahoma and your brand gets linked or some information gets pulled from your website, OpenAI actually wants to be able to offer that person as much information as possible. So it might say, like, hey, here’s three locations where you can do sauna in your market, and this one costs this much, this one costs this much, this one costs this much, because it’s trying to help that person make a decision.
If you’re blocking pricing, which is another really interesting thing that brands will do, is like contact your local studio, or some kind of information that they don’t want to share. That’s actually limiting the ability for you to show up in those answers, because you’re not getting enough information.
You may have just blocked yourself out of being part of it. Yeah.
Exactly. Yeah. So thinking about what’s on your own site, and we already talked about, but like canonical links and timelines, and think about just how to present the information so that AI can read it and understand it and start pulling it in.
So the more information on your website and the more valuable it is from a question and answer standpoint, that’s what brands should start thinking about.
What about video? We’ve been sort of through this phase where we think we need to transcribe everything, every video, every podcast, everything needs to be in text format as well, on the website, on YouTube, on all of the other video platforms. Where do you see that going with the large language models?
Are we going to get to the point where you don’t have to transcribe everything that happens in a video?
I mean, I think that might be a ways away. I would say if you’re a brand that already has a great video, you know, platform, especially if it’s on YouTube, because YouTube metadata is already informing search in Gemini. And as those overviews expand, the well-structured video content is going to get summarized inside it.
Like that we know 100 percent. What’s really interesting is Rumble just did a new partnership with Perplexity. And so there’s like this video AI search partnership that’s starting to point to something that will probably start happening with YouTube as well and Vimeo and other platforms.
But the key is that you need the transcription, you need the subtitles. You know, it’s a huge knowledge well for all of the multimodal models, for sure. And practicality wise, you have to publish your videos with really accurate transcripts.
Chapters are actually really interesting if it’s a longer form video having kind of those breaks in those titles so that it can just pull that one context, that one piece of the video that’s most important for whatever it’s answering. And then the structured descriptions. So oftentimes people will again, like do a marketing description, which might be great for the consumer potentially.
But even rethinking those descriptions to be really structured, kind of Wikipedia-fied to make sure that the AI models will actually pull in that information the way that they think is most valuable to whatever that person is asking. Because it’s the text that the model is actually digesting. As much as like people will say you can pop a video in and it’ll understand the context, it will not unless there’s a very clear transcript for the model to actually read and digest.
Wikipedia-fied, I think I have my new favorite word of the week. Thank you for that. Before we go, I wanted to get an update on what you’re doing these days with Root & Rise.
Tell us about the company and the team and the work you guys are doing these days.
Yeah, so it’s been really exciting. You know, I very much know that my superpowers in franchise development, everything from initial inquiry all the way to grand opening and really just understanding that world so well. So, I knew I wanted to create a company that helped brands do just that, take that travel or that journey from initial inquiry to grand opening.
I partnered with Willem van Daalen and Islam Tawfik because they’re incredible developers and know the AI world incredibly well. They had built out Franchise.ly and had done that for four years, which is how I met them. I was on the board of Franchise.ly with them and was just so impressed.
I love being around people that are smarter than me and they’re 100 percent smarter than me when it comes to all things AI and development. What we knew we wanted to do was essentially build this platform where we’re helping brands scale through Franchise sales and Franchise development. But what I knew for sure is that I could not do it without a huge tech stack behind me.
The way we describe it is, I’m there, I’m on your leadership team, I’m embedded into your brand, I’m very much the Chief Development Officer. You might have team members report up to me just like somebody on your leadership team would. But the way that I’m able to do is almost like magic in the background that’s happening that you benefit from as a brand, but you don’t really need to know a whole lot about.
It’s just happening. We very much rely on Claude MCP as our spine. That’s Anthropics Model Context Protocol.
Essentially what it does, it connects agents to our actual business systems. HubSpot or Salesforce or Salesforce or Google or the mapping software. The way that I picture it because I’m such a visual person is like, I imagine it just has these hooks into all of the systems that a brand uses in its entire ecosystem.
That way when we’re asking to create things or do things or understand things or make a template for something, Claude has all of the information that needs to know about that specific brand. Instead of it going off and doing research on its own, we’ve essentially created this very safe sandbox that allows us to do brand-specific work. We also teach it our own thing, so it always has access to our own drive where we have a ton of tools and resources for ourselves.
It’s been really incredible. We also have Jacqui Blair on our team, who’s amazing. She’s doing a lot of the initial calls and a lot of the admin work as well, working with brokers, territory checks, all that fun stuff.
It’s been amazing to build, and it’s really resonating with people. We have four incredible clients. We have a couple on a waiting list that are going to start before the end of the year, and it’s been awesome to see the growth and how much people are needing kind of what it is that we’re talking about doing.
I love the direction you’re taking. It’s so different than the old days of fractional Fran Dev sales, and by the old days, I mean five years ago, where it’s like, send me some leads and I’ll call them. Right?
Like you guys are so deep into what the brand is doing and what’s actually happening out there in the world within all of these platforms and everything from a technological standpoint. I love that. If anyone would like to get on that list to do some work with you, where can we send them?
Yeah, so I landed, I think, the best URL possible.
I know it’s coming.
It’s franchise.dev, and all of our emails are just our first name at franchise.dev. So you can see me, you can reach me at patti at franchise.dev. And I’m obviously all over LinkedIn.
And so feel free to reach out there. I always have a book appointment button on my profile on purpose. Grab some time on my calendar.
I’m always willing to have a chat.
All right. Well, very good. Thank you for being willing to have a chat with me today.
And we will chat again soon about what’s new in the world of technology and franchise development. Thank you, Patti.
Thanks so much, Jack.
Back in a moment after a word from Northeast Color, here’s a franchise reality check. Your brand standards mean nothing if every location looks different. That’s where Northeast Color comes in.
For over two decades, they’ve specialized in turning franchise brand guidelines into reality. Their value engineered approach means you get consistent results without breaking the bank. From design and fabrication to installation, Northeast Color handles the entire process, so your brand looks exactly the same whether you’re in Maine or California.
As they say at Northeast Color, they’re not just making signs, they’re building partnerships and creating spaces where brands can thrive. When you’re ready to stop playing signage roulette with your brand, visit northeastcolor.com.
Using Digital Menu Boards is a strategic decision that can boost revenue and improve customer experience. Hughes Digital Signage delivers immersive digital experiences that engage customers, employees and visitors alike. Show your customers’ targeted communications, in-store picks or video of menu items they can order.
Exhibit promoted items that match in-market ads or highlight items that meet inventory availability. Hughes Digital Media Solutions are easy to use and managed via an online administrative dashboard. Reach out now to find out how Hughes can become your Digital Menu Board and Digital Signage Provider.
Visit hughes.com/franchise.
We’re going to dive deep now into a study from Vendasta that uses real customer data from franchise organizations on how franchises are leveraging AI. Joining me now is Jeff Leach. He’s the Head of Enterprise Sales at Vendasta to look at some insights.
Hey, Jeff, how are you?
I’m doing great, Jack. Thanks for having me.
I’m really interested in this study because, you know, everybody, everybody’s talking about AI, but I don’t know that a lot of people are doing so much about it right now. So tell me about the study. What are you looking at and over what time period were you researching?
Yeah, great question. And you’re exactly right. I mean, you hear it in every presentation, every webinar, everybody’s talking about AI.
But what I haven’t heard is a comprehensive study to say, is this really having the effect that it’s supposed to be having? And so I am always skeptical of anything new. And so our company is pitching this, we’re selling it.
And I’m like, well, is this really having the positive effect with our partners? So we have many, many thousands of people using our AI now. So what I did is I took our top 20 franchise brands and multi-location brands over 30 days, actually 28 days in August.
And I pulled data from our system, largely around AI web chat engagement and AI SMS engagement. And I’ll talk a little bit about what those two things are.
Okay. Yeah, tell me about web chat engagement first.
Yeah, so when we talk web chat, you know, as I mentioned, I’m skeptical. So when I used to use web chat, it wasn’t web chat. It wasn’t chatting.
It wasn’t doing anything that was giving me any information. It was, hey, you know, would you like to talk to sales or would you like to talk to support? And I’d say yes.
And then it’s a great, we’ll have them email you. And that was it.
Right?
So that’s what we knew of chat. And it caused me to be a little hesitant in adopting, and I’m sure it caused others to be hesitant. But now we have true AI web chat, where it is a conversation back and forth.
And so when I say AI web chat, I mean something that’s conversational. You ask a question, it answers the question. You ask another question, it answers.
Ours really doesn’t ask for name, number, email, first born until into a natural place in that conversation, where it makes sense to ask for that information. And we found success with that. So as I mentioned, I think people are still skeptical, but we are seeing about 8% of web traffic across these 20 brands.
8% of their web traffic actually engaged with the AI web chat in terms of engagement.
So that’s 8% of the people who went to a web site actually engaged with the chat on that site, is that what I’m hearing?
That’s exactly right.
That’s actually pretty good. That’s a pretty high number. I would have not guessed that it was much more than 1% or 2%.
So 8% is that makes me kind of optimistic about the future of web chat.
Yeah, and so I initially started this study by doing 6 months, and I saw such an exponential growth.
Really?
That is, let me take August because to your point, it is growing. Yeah.
Very good. So what happened with that 8% who actually engaged with the chat on the website, where did they go from there?
Yeah, so the 8% actually translated into just over 40,000 conversations. And when I say conversations, the data also told us that a conversation averages about 11 messages back and forth, which we also felt was great. That’s an engagement to us.
Yeah.
But then out of those 40,000 conversations, roughly half, 19,704, turned into leads generated. Meaning in this case, we collected the contact information and we were going to follow up. They went into a CRM and we were going to follow up in some way.
So a 50 percent conversion rate blew me away. We actually had a few brands that were one that was as high as 65 percent conversion. Wow.
50 percent is amazing. If you can get 50 percent of anything to convert into a lead, you’re doing something very right there.
100 percent. So that was the tip of the spear. We saw the 8 percent of web traffic and converting at a 50 percent rate.
Then what we looked at is a different set. I mentioned this at the beginning. We also looked at AI SMS.
The old SMS is we text a one-way campaign out, hey, you should be using my air conditioning service. That was it. Hopefully, they call in or they take some action from that campaign.
But we now have SMS that is truly conversational. It’s not one way, it’s actually two ways. An example of that, we ran with one company is air conditioning service.
They’re in the CRM, they came in through many different ways. We did an outbound SMS message that said, hey, you saw that you’re interested in our air conditioning maintenance package. Timing is perfect, we’re running a special, would you like to learn more?
Now, I’m asking a question, now I’m engaging. And so we saw a high rate of engagement there. Actually, 40% of the customers that we reached out with, with that AISMS said, sure, we’d like to learn more.
Wow. And that’s just from a cold text going out, right? So 40% of the people who got a text from these clients started engaging with that text, is that right?
That’s right. And they were in the CRM. So it was a pure cold where, I don’t know who you are.
Yeah.
CRM for some event that happened. But they were getting almost no engagement via email. So this was just a huge lift in conversion for them once they started with this AI conversation via SMS.
I know you’ve done email marketing for several years over, over, you know, lots of different campaigns and things. Are you seeing better results now with the SMS stuff that you’re doing now than you’ve seen in the past with email marketing?
It’s night and day, Jack.
Really?
So the conversions we saw through email were roughly about a half a percent. The conversions we’re seeing in AI, and this is across the company now, not part of this study, but we’re seeing about 14% via SMS and AI. So it’s night and day.
I can’t do the math, but I know that’s a lot more than half.
Yeah, that’s amazing. I guess it comes down to consumers are more likely to respond to a text than they are to email. I mean, maybe text is where email was 20 years ago or something like that, right?
Yeah, 100%. It’s also the generation, right? Its customer behavior has just shifted dramatically.
Even over the last few years, it’s this whole shiny object syndrome, TikTok syndrome where I need to do it now, right? I need to go on Amazon and order it within seconds of me seeing an ad for it. SMS gives them that window, that avenue to go react quickly.
Got it. So, how did the SMS text outreach, how did that convert into new leads or appointments or jobs booked or anything like that? Were you tracking the results on those texts going out?
We were. So, the 40% engagement with AI, that’s like a 5X improvement over what we saw on WebChat. And then of those AI SMS conversations, we saw 35% of those engaged leads actually book appointments.
And what we love about the AI is it can book appointments. So, I know us and others, we integrate with many of the different booking services. So, that whole AI conversation was never touched by a human.
It was, hey, would you like to schedule your first service? Yes. Great.
What days work for you? Friday. Great.
We have Friday at 2, 3, and 5 available, which works best. And then it goes in, books the appointment, and then it even texts the business and maybe even the contractor and said, hey, you’ve got an appointment Friday at 5 with John to do an air conditioning maintenance service. Wow.
So, 35% conversion based on those AI SMS conversations.
Wow. So, overall, the amount of business flowing in from these outreaches and from the web chat even, this is a whole new world that we weren’t even looking at this, let’s say, what, three years ago. There really, I mean, there was some marketing going out via text and via web chat, but consumers really weren’t engaging with it at this kind of level, right?
There was still that mistrust, that legacy used AI, which wasn’t AI, and it was crap. Yeah. So it took, I think, this long to really see the engagement, and that’s back to that exponential growth I talked about.
It’s really hitting its stride now and frankly, blew me away when I saw the results of the study that we were doing.
Is it because do you think that people are just more comfortable engaging with some of these technologies, or maybe the AI has gotten so much better than in the early days, that it’s actually, the AI is actually engaging with consumers better than it used to?
Yeah, I think it’s a mix. I think it’s both. I use it every day.
I know you use it a lot. I try out the different models. One model is better one day, and one model is another, and now we’ve got Nano Banana with Google, and I can now create images.
I think there’s a familiarity and maybe a trust that wasn’t there, even six months ago. Certainly, three years ago, it wasn’t there. I think that’s a good part of it.
I think it’s back to those customer expectations. They want answers now, and so they expect it.
What’s next for all of your research and putting all of this data together for your clients and future clients and everyone else in the franchise industry? What are you looking at next and what’s the next six months look like for you guys at Vendasta?
What I wanted to do is see how I can tie this to revenue. The end of the study was, I looked at what we call missed call text back data. Missed call text back is self-explanatory.
If somebody calls a business, maybe it’s a hair salon and the stylist is busy, they don’t answer the phone. They get an immediate AI SMS text back and says, Hey, we’re on the phone. Can I answer any questions for you or would you like to book a service or write the AI takes over? From that, I was actually able to derive some revenue numbers.
We actually were told, believe it or not, and this blew my mind, anywhere between 20-40 percent of their leads were being missed because the staff was too busy to answer the phone either during hours or the calls came in after hours. Back to the shiny objects in general, people call whenever at two in the morning because they expect an answer. The real quick numbers there, for one particular brand with 40 locations, that was almost 3,600 missed calls during a 28-day period.
Doing the math there, it’s about 89 texts that went back per location of those 40 locations.
Wow, that’s a lot of missed revenue, yeah.
Yeah, it was. And what it translated to, it was an uptick of revenue of 200, almost $250,000 for this 40-location brand.
Wow.
And while that wasn’t incremental, because some of those people would book.
They would book through other means, maybe. Or, but they could also book with a competitor, right?
They could book with, right. And that’s the big thing, right? And we know that 80% of people don’t leave voicemails.
These are industry stats.
Yeah.
So if 80% aren’t leaving voicemails, they’re not coming to you. They’re going to the next person.
Finding whoever else came up on their search, yeah.
Yeah, right. I mean, we all do that, right? We’ve been doing it for years.
So running the numbers, we saw in this particular brand, it was about a million and a half dollars in new annual business, just from implementing this AISMS and this Misscall TechSpec feature that we had. Wow. So it blew my mind.
So you asked another question, which is, what’s next? This is just one example, one aspect of AI. What I want to go back and look at is efficiency.
Why did this do to organizations? It’s no secret, Neighborly is one of our partners. They use us for AI among other things.
We keep hearing from all of them, what you’ve done from an efficiency standpoint to help us out, has drastically changed our organization. We now can do the things we’re supposed to be doing, rather than the mundane or the very manual tasks that AI now replaces. So I’d like to go back and actually see, how does that impact the organization, whether it’s revenue or any other KPI?
Yeah. And I love the idea of looking at this from a revenue lens, because whether you’re talking to the small business owner, or the franchisee, or you’re talking to the Chief Marketing Officer, Chief Revenue Officer, or CEO at a big franchise company, the engagement numbers are really interesting to geeks like you and me, right? But the revenue numbers and, you know, missing out on $250,000 of missed revenue or something like that, that’s where the business owners and the CEOs really want to hear more data from you and your team.
Jeff, if someone would like to learn more about this study or get some more of the results, where can we send them? And also, if someone would like to have a conversation with you about the technologies that you guys are working on, where would you like to connect with people?
Yeah, thanks, Jack. So please reach out to me. It’s J Leach, J-L-E-A-C-H, at vendasta, V as in Victor, endasta.com.
JLeach at vendasta.com or LinkedIn. I’m connected to almost everybody in the franchise community, not as many as you, Jack. But I’ve been doing this too long to admit.
Yeah.
So please reach out. We’ll give you a demo. We’ll start a pilot at no cost, risk-free, just so you can start using it.
And whether you’re using Vendasta or somebody else, it’s really time for you to step in with both feet and just try it. It’s really, I mean, you see the results of the study. It’s having a huge impact.
Excellent. Well, thank you so much for sharing that. We look forward to hearing more from you over the next few months on where you’re going with your next data survey project.
And we will see you soon at a franchise show down the road.
Thanks again, Jack. Appreciate you.
Thanks again to Patti Rother and Jeff Leach for joining me today. Thanks also to today’s sponsors, Hughes, Northeast Color and Thunderly. And thank you for listening, telling a friend about us, and staying connected on the Social Geek Radio Network.