← Unemployable with Jeff Dudan
Unemployable with Jeff Dudan
Franchise Tech Strategy: When to Build, When to Buy (AI Changes Everything)
00:12:34
Loading summary
Transcript8 lines
- [00:01]
A
You're going to have a real strong bias to this question because you are a provider of technology solutions. How does a franchisor decide whether today they should buy something or build it themselves?
- [00:20]
B
Yeah, that's a good question. You know, you have finite resources, both in terms of, you know, money, but also on mind share, right? So anything that you build yourself means you have to spend resources in thinking about what you want to build. And this is true also, like, you know, I use AI all the time, you know, to build applications on my own, to test them out, right? And also internally, like when you like friend Connect is a tech company. Like a lot of the SaaS products, third party products we buy, we could have built it ourselves, you know, tracking what code is built where, like some marketing automation stuff, sending emails and all. You could probably build it yourself. But the problem then becomes like, is it better for me to spend my time and money and resources into building this versus focusing on something that creates a different sheet or something that creates more, where the ROI is going to be higher, right? So it's always a decision. If you have unlimited budget, unlimited time, you know, there's no difference, right? You can build stuff by yourself because it'll be exactly what you want it to be. But the world changes, right? So what happens, like, you know, six months down the road or one year down the road, you want to make some changes to it or you want it to be different. You know, how are you going, you know, how are you, you know, who's going to own it? Like, what are you going to do? And the person who built it, if you build it internally and the person isn't there, you know, what are you going to do about it, right? So there are all of these, these pros and cons. So I'm not saying that, you know, you should never build something or you should always buy something. It's like for most organizations, given finite time, finite resources, it makes sense to buy stuff, some things and it makes sense for you to build certain things. Like that makes you know that that is uniquely yours, right? So if it's uniquely yours, maybe, you know, try to, and you think the ROI is going to be really high from you building it, then you should build it. I'll also give an example, right? With AI, like often people show these examples. You can do this. It's called vibe coding. I just go into replit, enter some text in there and it build me an email sending app, blah, blah, blah, it'll build it for you. It looks pretty cool. You can Say use Google's design and it'll even do that for you. But in the process of building it, there are two things that happen. One of the things is sometimes it'll run into an issue. When you go to the app, it'll say, hey, it doesn't do this thing that I want it to do. The app will then say, oh yeah, I see, I made a mistake, let me go try and fix it. It'll then try to fix it and then introduce another thing. Suddenly you're looking at it like, hey, you didn't fix it. But you also made this mistake over here. Like now it doesn't allow you to like click on this and drill down on it. Like when you're looking at it, it'll say, oh, you're right, this thing doesn't work. Let me go replace, I think I used the wrong module for this thing. Let me go replace it with this other library in Python and it'll explain, really. And I think I'll fix it that way, right? And then at that point you're like, okay. And then it puts this Python library, whatever it does, and it says, hey, now it works great. But then you said, okay, now I also want to do, you know, role based sign in. And then you said, when you try to log in, you say, hey, my login isn't working. Oh, this thing that I've put in here doesn't actually work with this login. I need to update the login. Like it doesn't support X, Y and Z. And you're thinking, as a developer, I've spent like one week on this thing, I'm so close. The only thing left is this little login thing. So then you tell it and you say like, hey, is there a way around it where I don't have to provide this like, you know, role based, like login whatever credentials can you, like, relax it. I can fix it later and the app will say, that's a great idea, why didn't I do it now? And then later on you can fix it. I build the app, I publish it, I tell all my colleagues, hey guys, I built this amazing app that's available that does this thing and people love it. But what they don't realize is that I have put a security thing deliberately in there to just get this thing to work. Because the criteria that I originally put on access control was too hard for this app to implement it, I had to rebuild it all over again. So it's done and it's available. And so there are many examples of where you know, these, like shortcut, like AI allows you to take these shortcuts, you do these shortcuts and it's there and you don't see it until much later. That's why I call this like the last mile problem.
- [04:42]
A
Right?
- [04:42]
B
AI is something that is serious. So that's why when you're trying to build things like, you know, what are the last mile, what's called shortcuts that you took. When you buy stuff, remember the vendor has gone through certification, SOC1, SOC2, like all of these things that are really hard to bypass just because you wanted to move that into production or just make it available, right? So that's the way I think about it, right, is that it's an ROI cost benefit analysis and you know, what makes sense for you, the question is
- [05:15]
A
still relevant, what business am I truly in? And with companies like Fran Connect, like yours. And then we had Sean Clark from Go High Level, which is a rapidly growing, very cool CRM platform. His view was comforting to me and he said, look, you've got all these startups, all these single modality, like you said, somebody built an app, now they're trying to sell it for X dollars per seat or per location to a franchi system. And then what I, what I, what I saw in 25 was like you had, let's say you had five single modality applications that all had a price tag to it. And they basically these five applications, com completed your customer conversion journey. Okay? It was automated response text, it was automated voice, it was all of the things it was, you know, and it was getting the information into your CRM and callback and drip email and you know, you, you put these things and it was a kind of an expensive little Frankenstein. And then all of a sudden, within about 90 days, each one of those single modality softwares had all the features of the other ones, okay? Because they, because once they knew how to build something, now they had programmers and they just, they, you know, they went that way, they went, you know, they, they went vertically up the chain and they started building more robust apps. And what Sean said was he goes, look, he goes, there's going to be all of these things. He said, but what good companies do that already have a customer set is they're going to use AI as a utility to improve their product and not continuously try to upcharge their clients for it. Because you could try to get them for it today, but it's going to be table stakes in three months or six months. So they rolled out some call center features and functionality and AI and you know, and they're not charging, they don't charge for these things because it's like we're just using AI as it's intended to improve our products. So at the end of the day, the companies that have scale, that have loyal customer bases, that have, have built those relationships are if they use AI appropriately, they will so quickly be able to expand their products, improve their products, that there won't be a reason for other people to go and flirt with these other single modality type, you know, shiny objects.
- [07:42]
B
Yeah. By the way, that's a really excellent, that's a really excellent point. And what he's touching on is how AI is changing the cost of development. I mean, the big thing in the old days what happened was you built a product and if you wanted to extend the product because you had to maintain what you already have is you would have to hire more developers, your cost would go up. So there was a additional cost to adding features onto an existing platform and that's why they built these different modules. I think what's happening, I mean, not, I don't, I think, but what, what is happening with AI is that AI is allowing you to incrementally add functionality and features into a product without adding more development costs associated with it or just it's a very low cost. Now what's interesting about this is two ways, right? So the way products that are AI based in AI, the cost of building something is really low because AI is building it. But then to use the AI is actually really high because that's why they're building up these data centers, right? Because you can't actually predict how much it's going to cost to a vendor for a customer that is using AI. Because right now the benefit we have is that the AI access to the infrastructure is pretty much subsidized, right? Because all these big AI companies, these foundational companies, don't need to make a profit out of it. But in a SaaS company, what happens is your development costs are costs that you have to account for. But then when you bring in customers, the marginal cost for every customer is very low, right? Because you know exactly how much, you know it's processing time, power it's going to take you to use that application. So we are seeing this evolution. We'll see between, you know, low, like we are going from high development costs and low operational cost to very low development costs to unknown operational costs, right? So it's, I'm talking about these products that have AI built into them. And so that's why it's a huge impact on the market. And I think that with a lot of AI companies will probably have to move to some of the usage based model because the development is really, is going to be very low. It'd be table stakes. Right?
- [09:52]
A
Yeah. I even hate to ask this question, but what do you see over the next three years? Is that too long of a time horizon to ask about? Maybe what do you see in the next year? In the franchise industry you've got a bird's eye view of the data, the insights. You, you see what's happening before anybody else sees what's happening. And now particularly with the AI tools, you can probably make sense of it. My view is franchising is accelerating at an accelerating rate. People are, are looking to entrepreneurship as a way to build wealth for their families. We' the service based industry, so we're relatively low cost. And then you know, with a, you know, we, we, the way we scale is different because you don't need another box, another gym, another restaurant to get more customers. You could just scale these businesses. So I think we're positioned in a really good place as long as we're smart. What do you see?
- [10:50]
B
Yeah, I, I agree with you. I think the, the, the, you know, I mean a byproduct of AI, the impact of AI is that I think more people will become entrepreneurs. I think they will move to businesses that are relationship businesses. Franchises tend to be relationship businesses. So you cannot really be replaced by an AI in most cases or by a robot in most cases. I mean the robots can work within it, but I think the business model is still a relationship business. So I agree with you. I think my take is also that that's going to be the case. But what's gonna happen is that, you know, and I think you're right in here is that, is that people won't just be, you know, like new franchisees. So because people are gonna move to, you know, to be, they wanna be, you know, business owners, entrepreneurs is, it's not that I've been looking at, hey, do I need to get a McDonald's or a burger King or whatever it is, they will use AI to do their research, like I said, right. They will find like all of your, you know, the FDD data and all of your growth and like all of the information available, like your franchisee, you know, Google reviews of your business, your nps, all of that avail information is now easily available. So people who are making these decisions will have more data, more information. And so those franchise businesses that are well run, like I went back to, like high engagement, happy franchisees and others will benefit much more than those that don't because all of this information will now drive into the decision making process. Right. So that's going to be, I think that's what's going to happen. Like you'll have better, more growing market. But I think they will be the winners and the losers. The winners will be the ones that are operating successful brands.