The Koerner Office - Business Ideas and Deep Dives with Chris Koerner - How to Make $864K/Year Profit with ATMs. ATM Biz Deep Dive. Ep #072

Episode Date: September 30, 2024

I sat down with Mitchell Sorkin, the ATM guy from Twitter. He owns 900 ATMs across Los Angeles, bringing in $864,000 a year in net profit. We talk about how he bought these locations rather than using cold calling to get them, and why this is an approachable business for anyone. The multiples are much lower than you'd think, making it a great opportunity in a well-established industry.At the end, we talk about few different business ideas beyond the ATM industry, including a unique AI business idea that you probably haven’t heard of.  Timestamps below. Enjoy!---Watch this on YouTube instead here: tkopod.co/p-ytAsk me a question on or off the show here: http://tkopod.co/p-askLearn more about me: http://tkopod.co/p-cjkLearn about my company: http://tkopod.co/p-cofFollow me on Twitter here: http://tkopod.co/p-xFree weekly business ideas newsletter: http://tkopod.co/p-nlShare this podcast: http://tkopod.co/p-allScrape small business data: http://tkopod.co/p-os---00:00 Highlights00:14 Introduction to Business Ideas and ATMs01:12 Meet Mitchell: The ATM Guy01:48 Starting Small: The First ATM Deal02:35 Scaling Up: From 3 to 900 ATMs04:45 Challenges and Strategies in the ATM Business06:18 The Importance of Location and Contracts16:25 Revenue and Margins in the ATM Industry17:44 Acquisition Strategies and Market Insights20:12 The Best ATM Locations: A Case Study24:09 Managing High-Volume ATMs25:06 Cash Loading Strategies26:40 Surprising ATM Locations28:19 Dealing with Theft and Insurance29:25 Challenges in the ATM Business31:00 Advice for Aspiring ATM Owners32:51 Future Plans and Market Insights36:22 Exploring New Business Ideas38:54 Diving into Comfy UI44:14 Resources and Tools for Comfy UI46:57 Conclusion and Contact Information

I'm very curious about locations that you just wish you could copy and paste, right? Or locations that are surprisingly or unexpectedly high volume or very profitable and what those look like and why you think they're like that. Welcome to the Kerner office where I primarily do two things. I talk about business ideas and execution strategies. And then I go deep on different industries that are approachable for most people. Today, I wanted to talk about ATMs because ATMs are like vending machines and everyone loves vending machines, except ATMs are even more passive than vending machines.

So what better guy to talk to about ATM machines than the ATM guy from Twitter, Mitchell. He's great. And he has 900 ATMs across Los Angeles that make him $864,000 per year in net profit. And he didn't cold call to get all these locations. He bought them. And the multiples are much, much lower than you might think this is an approachable business for anyone. It's well-established industry, lots of room for disruption.

And then at the end of the episode, we talked business ideas that have nothing to do with the ATM industry, one in the AI sphere that I think you will find interesting and you probably haven't heard of. So please like and subscribe, share with a friend that might be interested in businesses like these, and we will see you next time. So why don't you tell us who you are and what you do? Yeah. First of all, thank you for having me on. Really appreciate it. My name is Mitchell. I am the ATM guy on Twitter, and I've spent the last three years kind of spending all of my time in the ATM business. I rewinds like a little bit, I spent most of my 20s building internet businesses, a polar opposite world.

And at some point, my late 20s decided I wanted to buy a cash flowing business. I just fell into the world of SMB, Twitter, just started reading about search, trying to like soak up as much information as I could. And kind of randomly, purely out of serendipity, stumbled into an ATM deal. And the seller originally was selling like 400 machines. It was a much larger deal than I wanted to do. And my brother was my partner in the business. I told him, why don't we just go and ask this guy if he'll just piece off a few machines to us.

And we can say, like, this is our way of getting our feet wet. We'll come back to you for a larger acquisition afterwards. It's a good idea. And to my surprise, he said yes. And so we ended up buying three machines. The deal is $36,000 for $1K a month in cash flow. So it's basically exactly a three-year multiple. And there's some details about the way that people present ATM deals that we can talk about later.

but that was how we perceived them multiple at the time and of being slightly higher than that. But yeah, so we bought those three machines. We ran them for about five weeks, maybe six weeks. We fell in love with the business and we ended up buying 20 more a few months later than another 40. And we did a bunch of smaller deals and now we're up to over 900 machines. Okay. Okay. Amazing.

I want to back up. You were in the tech world and then you said, I want to buy a cash flowing business, which is a great idea. Were you just like window shopping? on biz by sell? Or how did you get from that thought to ATMs more specifically? Originally, we were just like shopping around. We made like a short list of qualities that we wanted to avoid. And the main one was like high labor intensity. So in other words, we decided that like we did not want to buy a landscaping or an HVAC or a plumbing company and have to manage like 30 or 40

people. We just wanted to remove like the key man risk out of our acquisition. And my brother is really good at managing capital. And so it was just, we'd rather remove that, remove that unknown of human capital out of the equation as much as possible. And so that led us to laundromats and self-serve car wash and vending and like businesses that fall into that camp. How we stumbled on an ATM deal specifically was we originally were looking at a vending deal. And I think my brother asked the seller why he was selling. The guy said, oh, I want to focus all my time on my ATM business. And so that kind of piqued our interest that we started going down that rabbit hole. That sounds like Quinn Miller, right?

I'm sure you follow him. Yeah, yeah. He's a good friend of mine. That sounds like his story. It sounds like his preference because I had him on the podcast a few months ago and he's got all these vending machines and a few ATM machines. And one of the questions I asked him was, what's next? And he's like ATMs.

Interesting. So, yeah. He's bullish on those. I didn't know that he was that bullshit. I knew he liked it, but he must be like really focusing on that more now than he was previously. I think he's carving out a niche with ATMs in hotels specifically. He's finding some good success there. Yeah.

Yeah. Quinn is a beast at cold calling. So like his whole growth strategy is completely different than ours. In a lot of ways, like I'm jealous that you just sits there and pounds the phones. And I think he'll build a big business that way. When it came to you requesting from the broker to peel off a few machines, why do you think it was that they said yes, because I could see this one of two ways. They say no, because it's messy, it's hairy, it's too small for us. They say, yes, because it's been sitting for a while,

better than nothing. Had it been sitting for a while, or why do you think they said yes? The guy that we bought from is actually quite known in the city as the guy who builds routes and sells them. I think we got really lucky that he listed it as 400 machines, but in reality, he was willing to sell one or 80 or 180, right? So it was less that, like, I convinced him to sell three instead of 400 and more, that's just kind of his business. The cool thing about ATMs is that you can piece off a part of your portfolio, and it's not like there's specific overhead that comes with that, right?

So you can have 15 machines or you could have 1,500, and really it's just like more admin work. But it's not like, you know, the labor intensity doesn't say. scale with the size of your portfolio, if that makes sense. Which is probably rare, right, for any other industry that would probably not be the case. Yeah. Maybe software. What city are you in again? Okay.

Now, when it came to the seller, one thing that might worry me is if I ask him to peel off three out of 400, he's going to give me the three dogs and save the best for the big seller. But on the other hand, he could say, you know what, I'm optimistic. This guy's going to come back and buy more. I want to make sure that he has three. three good machines or at least one or two really good locations because I imagine the location matters quite a lot in this business. It does. But the thing is, ATMs are incredibly predictable in terms of looking at their historic performance and then projecting that six, 12, 18 months out.

So we looked at the last two years of transaction history on these machines. And that is what we based the multiple off of. And so you don't really care whether it's like, good locations or bad locations because you're just paying a multiple on that ATM's performance. So there are some details there where it's probably riskier to buy a location that does a thousand transactions a month than it is to buy 10 locations. Step back for a second. It's definitely riskier to buy one location that does a thousand transactions than 10 locations that do 100 each just because of customer concentration. But then also the locations that are really busy are incredibly competitive and the owners know that they have really, you know, high value real estate.

And so they're constantly trying to get better deals. And so I figured off the path there. But yeah. Well, I do that all the time. I have ADHD. So it's going to happen like 12 times. There's an interview. Yeah. I would also think that like the location matters based on price sensitivity, right? Maybe some locate, because part of your strategy, which I'd like to touch on at some point, is increasing the fees. And is that less possible at certain locations compared to others? It's definitely less possible and not because of price sensitivity and only because of the owner of the location. So it's actually like one of the most annoying and frustrating parts of this business is handling like, you know, depending on how many machines you have obviously, but like dozens or hundreds or thousands of micro relationships with these store owners. And these guys who own a mini mart that probably makes like 35, 40 grand a year.

and they are the employee and the bottleneck in the business and they just sit there all day, they will see you raise the ATM fee from 3 to 3.50, and they'll call you and tell you that nobody's going to come to their store anymore because it's 50 cents more, which is complete nonsense. But it doesn't matter. Because if their call. Correct. It's not their call. Most of the time, if you're smart about the way that you draft your language and your contracts,

but when you inherit a location, when you buy it, you're inheriting whatever the contract was, previously or if there is no contract, then it is their call because they could just tell you that take your machine if you're going to raise the price. Yeah. Well, like in any business deal, even if contractually you can do that, you still want to maintain a good, healthy relationship, right? Yeah. Of course.

Okay. All right. So how much due diligence were you able to do on those three locations? Did he let you talk to the owners? Did you even need to talk to the owners? What did that look like? Yeah. So typically it's just looking at transaction.

history from the processors. It's looking at the contracts and the language in them when they were signed, when they're going to expire, and whether they auto renew or not. I'd say that's like the bulk of the diligence, especially when you're just buying three machines or eight machines or whatever. In terms of talking to the owners, you actually don't really want the owners to know at all that it was sold because again, it just opens up this situation where they now want to demand to hire commission or they start saying like, where's the other guy? I liked him. He took care of me. Are you going to pay me on time? So in an ideal world, when you buy an ATM location, nothing changes. The cash loader is the same because you're third party contracting it.

If you're going to start loading it or your employees are and the cash loader changes, then ideally you ask the guy who owned it previously to tell him, hey, I'm bringing in these other guys. So if you see this guy coming in, he's just helping us out or whatever. At the end of the day, like 99% of ATM contracts, if there are contracts that you're purchasing, don't say anything about assignment. So if it's silent on assignment, it can be reassigned. And so, yeah, that's typically the way that you want to structure it. Okay. And did you have any hangups with this first deal as far as assigning things? No, we still have those three locations today. It's been almost three years. We did have an issue with one of them where one of them demanded a higher commission. We ended up giving it to

him. And then I think one of them, the transaction volume fell off for whatever reason. There are lots of risks that are like completely out of your control when it comes to a neighborhood changing or gentrifying or they open up construction on the street and for six months like nobody goes to that store anymore. There's all this kind of stuff. Okay. So what did it look like post acquisition? How long did you wait to change anything or to raise fees? And then what did the next three years look compared to your projections. How long did we wait to raise fees? I don't remember exactly. I know that for the first like two or three months, we definitely didn't touch anything on those three locations. When we bought the next batch, which was like another 20 machines, I remember we had two or three really low stakes

locations that did like maybe 15 or 20 transactions a month. And we started raising prices on those first. And we basically just made more money. It was like 15, 20 percent increase in net. And we were like, Yeah, let's just start doing this on all of the machines. Okay. And then we started to learn the issues with like negotiating with the store owners. And so we bought those three. Then another like I said, it was about 20, maybe 18 or 20 from the same guy. And then we bought another like 15 or 20 from him again.

In total, that route today is down on transaction volume, but is up in net. So it's down in transaction volume for a few reasons. One is like those kind of out of your control. factors that I just mentioned a minute ago. And another is like we lost locations to places just like shutting down, going out of business, selling to somebody else, the guy just kicking us out, whatever, right? I don't know the numbers off top of my head, but I do know that if it was supposed to do like 5,800 or 6K a month in cash flow when we bought it, it does like 6,500 to 7K a month today on like 20% less transaction volume. And that's all just a function of us raising prices.

Okay. So take me back again to the timeline in number of units. So three years ago, you bought three. Then you added 20 from the same guy. And then take me from there until today. After we bought the next two batches from that guy, we basically stopped for like maybe four or five months. We had gotten to a level where it was like slightly overwhelming. We were loading everything ourselves. And so we were like driving around, fixing machines, loading cash. And it was just we were like, okay, let's take a breather and get a customer. to this. And then I think like later, maybe like maybe early that summer or mid that summer. I remember the timing exactly. We had this broker reach out to us and we bought like one of the best deals that we've ever purchased to date. It was tiny. It was 30K. But it was doing like 26 or 2700 a month. Wow. Yeah. And all of the surcharges were like a one X multiple. Yeah, less than one X. But on top of that, all of the surcharges were like 225 and 250. There were no contracts. And the guy was like really old.

And so that's why he sold it for so cheap because he was like no contracts. I'm running out of time. So he just basically dumped it. And we raised all the prices to 350. And that route still today makes like 4,500 a month that we bought for 30K. So that was like the home run of home runs, unfortunately for 30K instead of 3 million. But that was the next one we did. And then we just like over time.

started buying like small routes here and there as they popped up. And then last year we did our two biggest acquisitions that added like the vast majority of 900 machines that I mentioned on the beginning of this call. And we brought in LPs for that, raised most of the money on Twitter. And so I don't know if that answers your question, but that was like the trajectory. So you're at 900 today, give or take? Yeah, correct. There's a caveat there that like a big portion of that is processing.

locations and I don't know if you're familiar with the difference between locate okay so there's kind of two camps in ATMs one is locations that you own and load and then the other is locations that you process so that means when you process a location it means somebody else owns the machine somebody else loads the cash in it they keep all of the surcharge but there's an additional form of revenue on ATM transactions that's called interchange Interchange is basically this payment that's made from the networks to the processors. So like Visa and MasterCard or Pulse or any of these networks pay the processor for basically allowing the transaction to happen. And interchange, like gross interchange that's paid out is anywhere between 25 and 40 cents per transaction.

And it depends on what kind of card was used. The type of store you're in, like there's like a bunch of different factors. There's nobody really, it's a black box that I think is by design made that way. It's really confusing. You talk to people who are like 20 years into the business and they don't fully understand interchange. That's a lot like processing. Same issue.

Correct. Correct. But regardless, there's that like 25 to 40 cents of gross interchange that's paid out. And then the processor keeps some and gives the rest to the ATM operator. And as you get bigger and bigger, you negotiate more and more of that back. And it becomes this like middleman above a middleman. And so processing portfolios are basically just smaller ATM operators under you. You give them 80% of the interchange and you keep 20% of it.

And so it's like fully passive. Machine breaks, it's on them. Machine runs out of money. It's on them. So out of those north of 900 machines that we have, I want to say like around 50, it's basically 50-50 own and load and processing. Okay. And what does that represent in monthly or annual?

revenue and or EBTA. So our revenue across all the entities is around 180K a month. It's a little bit complicated. Like I mentioned to you on the before we hopped on the call, it's like a little bit complicated talking about the flows of the revenue because we have our entity and then we have PVs that we brought investors in for. But our EBITA margins have been in the range of 40% on our wholly owned entity since day one, I'd say that's like relatively standard in the ATM business because, you know, the way that you really think about it is like if there's a $3 surcharge, a dollar goes to the, a dollar goes to the cash loader, a dollar goes to the store owner and you keep a dollar. And our surcharge is higher than that.

And I'd say most ATM operators that are smart about their portfolio have raised prices beyond that. And so like when you just look at a gross margin perspective, that's the math there. Okay. So wow, those are amazing margins. And it makes sense, like you said, because it's very cut and dry. Like it's you don't have to have a ton of overhead. Do you even have like an office or a warehouse that you need? We do.

We do have an office and we have a storage unit. We have two storage units actually. And then we have one office admin that basically runs like all of the back office stuff. Okay. Now have you acquired. any locations by cold calling or on your own or if you acquired them all. Acquired 95% of them. Yes, I have. In the beginning, I, as I was driving around and loading machines myself, I would like stop into a place and say, hey, how can you guys have a machine? I probably placed

like maybe 15 or 20 locations in the three years I've been in the business. Not surprisingly at all, all of the best locations that I've placed have come from Twitter. It's just like inbound. And like we got like one of one of the most incredible actually the best location in our entire portfolio came from Twitter like six months ago. But driving around is just like a bad R a lot. ATMs are a very mature market. They've been around a long time. It's incredibly unlikely that you're going to walk into a place that doesn't have an ATM and has never had one and placed one in there. And it's going to be an amazing location.

It would have already had an ATM in there. So that's not to say you can't go out and grind and pound the pavement and build a lot. the route you can, but it's going to take a very long time. And you were saying earlier, you're pretty, as long as you stick to Los Angeles, you're pretty supply constrained, right? You can't just scale as much and as fast as you want. Yes. Again, it's a really, or not again, we haven't talked about this, but it really depends on how

you're cash loading the machines. So if you yourself are loading them or your employees are loading them, then your route density matters a lot, right? So if you have 20 machines in like a 10 square block radius, it's going to be a lot easier to load that than something that's like scattered all over the place. You'd be spending all of your time driving. And so in that sense, you're limited in scalability. But if you third party load it and you have other guys loading your machines, then it's infinitely scalable. It doesn't really matter.

But you still have to find deals to acquire in that case, right? Yeah. Yeah, correct. Okay. Are you more interested in scaling to a different geological? or like if you had to choose between a different geography or just stay where you're at, you can't scale at all anymore. What would you prefer? I think it really depends on your goals. Like, I think if you want to build a big ATM business, you really have no choice. But to go

outside of your local geography, there's just not enough deals. There's not enough people selling. And even if there are a lot of people selling, most of the deals are going to be pretty bad, overpriced seller. You can't really trust. et cetera, et cetera. Okay. I'm curious about the deal that you got from Twitter that is your best location. I'm sure you don't want to spill all the beans there, but more generally, I'm very curious about locations that you just wish you could copy and paste, right?

Or locations that are surprisingly or unexpectedly high volume or very profitable and what those look like and why you think they're like that. I don't mind talking about it at all. I actually tweeted about it when it happened because it was a hilarious story. I got a call from this guy and he said he needed an ATM. Like what kind of location is it? He said it's a dispensary. He's like, okay, how'd you get my number?

He said, oh, the rabbi from my synagogue referred to you. What? What are you talking about? He's like, I don't know. I'm like, okay, well, when do you need the ATM? And he's like, now. What do you mean now? He's like today, like right now.

Like you've never gotten a call that had any of these elements in it. And this one had all of them. Yeah. And so I look up the address. I happen to have an ATM. ATM in my garage. I was like, you know, how many transactions does it do? He's like probably about a thousand a month. And without your audience does any context on that right now, but that is anything that does a thousand transactions a month is like anything above that is probably one in five thousand

ATM locations or one in 10,000. So I was like, okay, fine. I went in my garage. I programmed it. I went. I'm Russian also. This guy ended up being Russian. We really hit it off. And so anyway, that that ended up being the best account that we have. When we first installed a machine, it did do a thousand transactions a month in the first month. And in the second month, it did like 1,400 and then 1,600. And today it does like almost 2,000. Wow.

So then Santan, the second part of your question, what are the best locations and, you know, the ones that I would like to copy paste? I'd say that. And then you also mentioned surprising. Yeah, I will say that like the best ATM locations are not really surprising. It's going to be casinos and strip clubs and cannabis dispensaries, right? Probably the most competitive as well, I imagine. Correct, right?

So, like, strip clubs completely dominated casinos mostly owned by the casino owners themselves or just outsourced to these, like, mammoth companies. So really hard to get into any of those slash just impossible. Dispensaries, more doable. But dispensaries have, like, on a timer. Because I don't know if you know anything about the safer banking act and them trying to push regular payments into the cannabis industry. But like, as long as cannabis is federally illegal,

Visa and MasterCard is not going to process payments. So you can't just walk in and swipe a credit card at a dispensary. And it's very likely that's going to change at some point. And so if you put an ATM in a strip club, that's never going to change. People are going to be using cash at strip clubs probably forever, but it's not the case with the dispensary. So, yeah, I guess that was my way.

trying to answer all the pieces of that question. When you grew so fast at that dispensary, were you growing in step with his growth, or were you taking more of his market quote, if that makes sense? Yeah, no, it was a few things, and this is just me guessing, but I'm almost certain that I'm right about this just from having experience in the business. But usually what happens is when a store owner loads the machine themselves, which was the case there, and I'll explain what happened there in a second. when a store owner loads the machine themselves, usually they do a pretty bad job.

They're just focused on running their business. The ATM runs out. They don't have enough 20s on hand. They scramble to get the cash. They end up loading it a day later. When people keep showing up to your store and the ATM is constantly out of service, they just know that like, oh, I need to go, I need to go get cash before I go to this place because the ATM is never worth.

Or I'm going to go to a different place. Yeah, exactly. So I suspect that's what was happening there, especially because it was so high volume. So a location like this, you need to have, like you just can't have less than 10K in there at all times, basically. And really, you should have like 20 or 30K in there. And so unless you run an ATM business and you have an armored truck that'll get you the cash or you have an ATM program with your bank, it's really hard to get your hands on that many 20s. And so we took over. We put in a brand new machine.

We kept it loaded all the time. Slowly people started to know that it was just always in service. And the volume grew that way. So, yeah. I mean, just you having a workable ATM, a reliable ATM grew his business. Because I bet you there were a lot of people that started coming regularly because they knew that they could count on that ATM. It's possible.

I'd like to think so. I don't know if that's exactly true. But maybe, yeah. How much cash outstanding do you have at any given time? And how do you manage that? So again, as I alluded to earlier, there are a few different ways of loading ATMs and really there's three ways. The first is you load with your own cash.

The second is you outsource to a third party, which is usually what's called a vaulter or a cash loader. So that's just another local ATM guy that's going to put their cash into the machine. And the third way is an armored truck. And the armored truck borrows the money from the bank. you pay interest on the cash from the bank. You pay the armored truck per load. So every time they come to each store,

they charge you 80, 90 bucks or whatever, plus the interest on the cash. In scenarios two and three, the cash is not yours. So the money, you don't have the vault combo to the machine. You don't have any access to the cash in the machine. And when somebody comes and withdraws $200, the cash is pulled from the customer's account and sent to the vaulter.

And the searcher are just split between, you and the Volter. Total cash sitting across all of our machines, and let's remove the processing because that's like machines we don't even own. So it's kind of cheating to even talk about that number. Total cash sitting in all of the machines at any given time, I would guess, is in like the $5 million range. I'm just like throwing out a number.

I don't really look at that number because it doesn't mean anything to me. Let's just say it's probably in the range of $4 to $6 million. Okay. But a big portion of that. is vaulters. Cash orders elsewhere. To get into that business, you have to have that much on hand. No, absolutely not.

For 900 machines. Yeah. Is there a location that you placed that did just surprise you, like a gas station that just had crazy volume? Yes. Not a gas station. Generally, what happens is like you place an ATM or actually, it's mostly that we've purchased ATMs where like we looked at the transaction history.

and it said 250 transactions a month or 300 transactions a month, which is like a solid, good location. And you're like, oh, this must be like a fairly successful liquor store or whatever. And you show up and you're like, how could this place possibly do any business? Right. We have some locations here in L.A. where like, first of all, I can't even tell you what the business sells, right? Like you walk in and it's just like they'll have mops and diapers and then a generator.

And like, you know what I mean? You're just like, how are you selling all of these things? at once. They can't tell you what they sell. They're not allowed. Exactly. And there's just like stuff everywhere and the machine is buried under like a pile of stuff that they sell. So I'd say that's the most surprising. But mostly it's like pretty straightforward, right? Like really busy liquor store in L.A. is going to do really well. Really busy dispensary is going to do really well. Some like Chinese restaurant in the kind of shady part of town, probably not going to do well. So if it has a lottery machine inside, typically it does well because those are cash only.

So people will come withdraw from the ATM, put it into a scratch ticket vending machine or whatever. So I'd say it's pretty straightforward. It's rare that you're a surprise. Is risk or liability a big issue for you or you just send a check to the insurance company every month? And have you ever had a stolen or damaged machine? Yeah, machines get stolen. quite frequently if you have a lot of them. If you have, if you, it's just like a law of numbers, right?

If you have 10 of them, it's probably going to be pretty unlikely that one of them gets stolen. But if you have a big route, you're going to have like three, four machines a year stolen at least. We don't have insurance on the machines and the cash inside of the machines.

It's like prohibitively expensive. It's just dumb to have it. It's better to self-insure. We have cash in transit insurance for when our employees drive around with our cash. Knock on wood, nothing has ever happened there, but we're covered. And then we just have, like, general liability if a machine like tips over and crushes someone's leg or something. Interesting. So just a cost of doing business, something you have to eat. Yeah. Yeah. Interesting.

We did. My brother did really early on pull into like a McDonald's parking lot to use the bathroom and accidentally left the backpack in the car and he came out and the window was smashed. So we, I think he had like four or five grand in there. So that was like our first, that was our first help. What is like surprisingly difficult or hairy or complex about this business that you didn't foresee going into it? I think managing relationships with the store owners is by far the hardest part, especially when you acquire. It's my job for portions of the route where I was the one that was introduced. Like so especially the stuff early on, now it's just like our office admin handles all of it. And he's just kind of like the friendly guy that picks up the phone.

I mean, there's tons of locations. I would say a vast majority of our portfolio. I've never met the owner, never spoken to them, never been to the store, right? So there's a decent amount of that where it's like it's actually not a headache when you're just like, oh, this guy owns seven gas stations. He could not care less about the ATM, right? It's mostly the guys that just have one store. They're in there all day.

But yeah, I would say that's by far the most difficult part of the business. You're managing like so many different cultures. People have different backgrounds, ethnicity. Everyone has different preferences of like the way that they like to do business. They have different ideas around pricing and what matters and what doesn't. So you have to have very thick skin and you have to be really good at just like killing people with kindness. Like, you know, somebody's like swearing at you and yelling at you and you're like, hey, man, come on. Like, I'm just work with me here.

It's like you got to just like continue. Like you can't ever. It's a hardened demographic, right? Yeah. Just by nature of the job. Yeah. Yeah, for sure. And you can't blame these people. Like a lot of them are in really rough areas. You know, they're constantly being harassed by their customers. And so, yeah. What would you tell someone that's starry-eyed that wants to buy an ATM business? What's like an optimistic take for them and then like a hard, cold dose of reality take for them? Optimistic take. You're talking about buying one? Yeah. If someone wanted to do what you're doing in their market, yeah.

Yeah. Optimistic take. take is that a lot of ATM operators are extremely unsophisticated. They might sell you their portfolio at a massive discount to what it's actually worth. They have very likely underpriced the surcharges and there's a ton of upside there for you to come in raise prices. It's very scalable. So if you're good at the qualities that are required to run this business, you could go and just keep buying more of them.

And the returns are really good if you pay. the right multiples and, you know, run the business correctly. That's the optimistic take. The other take, slap in the face or whatever you called it, is it's not passive income. There's like tons of like ATM courses selling like passive income from ATMs. It's like, no, it's not. It's actually really difficult for all the reasons that I talked about. Unless you're willing to get your hands dirty and like drive around and learn the business

and do like all like that first like six to eight months for us. was really important. And so, yeah, I would say be prepared to do a lot of like really blue collar work. You don't get to just like click a button and sit in an office and print money. Yeah. It seems like it's a good business for someone who's thick-skinned, who's like high emotional intelligence, a good people person and someone who's like systems oriented, who likes and can enjoy building and following a good system.

Yeah. I would agree with that. What's next for you? Like, where would you like to be in five years in your business? I would like to keep growing. I'd say the bottleneck for us is, it was really a shortage of good deals out there. And when I say a good deal, priced correctly or underpriced, upside with the surcharges, a seller that you can trust is really important on the bigger deals.

That's much less important if you're buying like five or six, I'd say. But if you buy 400 machines and that guy has been running the business for 18 years, and he's a bit of a jerk. And then you buy it. And he's just like, oh, you know, like, see you later. Like have fun. And then they call, like these merchants call. And they're like, where is Bob?

And you're like, I don't know where Bob is. I'm Mitchell. Like, please work with me. Like, you need to be able to lean on the seller in these bigger acquisitions to help you with the transition. So finding like a really good person to buy from. And then fairly priced, lots of good, like contract, high contract. contract coverage on the route and good language in the contract's really hard to come by.

And so there's like a shortage of good deals to deploy large amounts of capital. So like when we find those, I pounce on them and I'd like to do as many of those as possible. And then the stuff that I mentioned to you in the beginning were like the route that we bought for 30K or whatever that was doing 3K a month, stuff like that occasionally pops up. That was the best example of that. there's stuff that's like slightly worse, but still amazing where it's like 15 months or 18 months. When stuff like that pops up, we just like write the check immediately, like almost do zero

diligence because it's just like you can't fake ATM processing. You get them directly from the processor. And so, yeah, I think in five years, I'd like to, I'd like to have as many of them as possible. It just remains to be seen how many of those deals that will get done. Why do you think it is? Obviously, it's not a passive business, but I would say it's more. passive than like a plumbing, HVAC, or electrical business. Why are the multiples so similar? We're looking at like 2 to 4X for something like this and something that's like a really

hard to run business. Why do you think that is? I'd be lying to you if I told you I know why. I really don't know. I think that ATM multiples are where they are because there's a lot of fear around the future of ATMs. There's a lot of fear around having to drive around with cash and load them, it seems incredibly dangerous. And I think that if somebody listed an ATM route for six or seven years, it would just nobody would ever buy it. So I don't know. I guess that's what I would say. That's interesting because I think I would agree. I'm seeing a lot of parallels in the mobile home park space. Like the prices are good because people think that mobile home parks are written with crime. They're not. Like some are, but some neighborhoods or apartment complexes are

written with crime, right? So there's that false perception. There's just the ugly perception that people just can't get over. There's like a pride, there's an ego thing that people can't get over. I just love little industries like that that are overlooked, right? Maybe that's the only reason that the multiples are so good is because it's overlooked. And we can talk about it all day, but as long as people have pride, ego, and fear, like these innate human tendencies, then they're probably both going to stay good industries for the foreseeable future. So we talked about this a little bit. This is primarily an ideas podcast.

Are you seeing any business ideas that are tangential to your space that look interesting or exciting to you? I had in the beginning some having somewhat of a tech background. Like I had some tech ideas around rebuilding software for ATMs because like anybody who has ever had even one ATM and has opened up their processors like mobile app or web app will tell you it's like a peek into like the late 90s. these internet. It's hard to explain how bad it is. And so in the beginning, I was like, man, somebody could just put out like a nice piece of software that actually works and scoop up,

like, all of the processing. And the more I learned about the business, the more I learned that it's just like, that's a pipe dream and you're solving a problem that your customers don't actually, your potential customers don't actually care about, right? Like, because as bad as the UI is on these processor apps, they will print the reports that you need. They will show you how much money is in the machine. And that's all that these ATM operators care about. And it's hard enough to get somebody to move under your processing by telling them, hey, I'm going to give you five extra cents per transaction or, okay, maybe that's a bit extreme.

Let's say three cents extra per transaction and you do 100,000 transactions a month. That's three grand a month for you, like just move over. And I have like probably a list of 15 guys that I've been working on for the last like 18 months trying to get them to move over. And it's a relationship building thing where like you keep following up. Hey, how's it going? You talk about business. Hey, have you thought about moving over? And you just keep following up like once every three months.

Nobody has moved. Like I haven't really been able to get anyone to move. And every time like, hey, you know, like in the last six months, you've lost like 14 grand by not moving over. Like you could have just. And so when I see that, I realize like it's just really hard to sell anything to ATM people. And so I don't have many ideas that are tangential to ATMs where it's just like. I think the more mature market, the harder it is to sell into it. But then once you do, the stickier they are.

So it's like, is that a battle you want to fight or not? And it sounds like from someone in the industry, it's not. So what other business ideas are you seeing out there that might not be related to ATMs at all? Because we started geeking out before I hit record. And I want to hear more about that. Yeah, sure. So I've recently gotten completely obsessed with a open source software called comfy UI. And for anyone who doesn't know what that is, basically if you've heard of stable diffusion or flux or any of these image generation models,

comfy UI is this GUI that was built that allows users to play and manipulate these models in all sorts of different ways. And so I think like the best way to think about it is think of a Photoshop that allows you to play with these models, right? So, you know, like Photoshop, you put a photo in there and there's like the clone tool, if you've ever heard of that. So you like circle, let's say like this screen that you see right now, you'd like circle this window shade and then you'd click somewhere else and start painting and it would

start painting whatever's here. Right. So there's like code on the back end that is doing that, but Photoshop abstracted away all of that like nerdy technology on the back end and gave the user like a user friendly tool called the clone tool. Same exact thing is happening here with comfy UI. And so you can build these things called workflows that do all sorts of things for you. And I think that there's like, I was, I mean, I was telling you, I was like excited to tell you about this because I think like a guy like you is just going to watch like a 20 minute video on ComfUI and instantly just have a hundred different ideas. I think like there's for every single vertical, whether it's B2B or B2C, there's like some application out there that can be built with this thing. And so I'm just

like spending all of my free time learning the software because I think like once you understand the tools, it's going to be really easy to just put out microsass for basically any industry with this thing. Yeah, you were talking about virtual staging, right? And I'm thinking since the cost to generate these images is so much lower than it's ever been, is there a world where you could go scrape zillow or realtor.com, you could scrape existing listings and then have the workflow, auto, you know, stage, virtually stage some of the rooms, send it to the realtor, as like a cold, a very customized cold email that costs real money. 100%.

Right? But it's like the equivalent of you go get your school pictures and they have a big watermark on them and then you can buy them. You send them with a big watermark and then instead of paying 200 for them like they're used to doing, they could pay 20 for them. Is that bingo? An opportunity? Correct.

Yeah. Interesting. That's exactly right. And copy paste in any real estate niche. Like it doesn't have to be residential. Yeah. It does have to be realistic. state either. What about like cosmetic dentists, for example, or like veneer clinics? They could scrape images

of you smiling, put new teeth on you, cold email. Hey, this is what it would look like if you had new teeth, right? So it's just like, you just like literally take that idea, copy paste it into any number of industries. Yeah, I'm picturing like scraping because every state has like a dental board website where they list all the cosmetic dentists or dentist assuming there's headshots there. Scrape, this would be creepy, but creepy is good in this case. Scraping the head shots, putting like this big, bright, beautiful smile on them, which they probably already have because they're cosmetic dentist, but it's different. They'll recognize that's not their smile, cold emailing it to them and saying,

hey, like, you could white, like, you could white label this through us to your end customers. You could charge whatever you want for it. Or it could just, you know, we'll charge you for it and it could be free to your end customers. Yep. Or like tattoo shops. Like right now tattoo shops, a good friend of mine, I was asking him, he has a bunch tattoos. I'm like, when you go and you're exploring like whatever tattoo you want to buy, do they show you images of it on your body? He's like, no, of course not. They just show you like

a template. Of course not. Dude, it's possible. They could. Yeah. I know. Look, that's too good an idea. Of course not. That makes too much sense. Yeah. So, yeah, I'm super excited about this thing. There's companies nowadays that all they do is paint brick homes white. That's all they do. And I could just blanket it. Yeah, you could just blanket a neighborhood. I mean, in Texas, literally every home is like a maroon red brick. You could just blanket a neighborhood or these companies could with a picture of their home, white, with like black trim, really modern and snazy looking. And then it's like it plants a seed in their head.

It's like. It's their home. It's personalized. Yeah. Yeah. Yeah. Oh, man. We could talk.

We could talk for hours just about this. Like you could do it with pools. You could do it. Like, here's what this pool would look like in your backyard and only send. it to people with million dollar homes. Wedding planners. Kitchens. Oh man.

Wedding planners. Landscape architects, painters. Like it's literally, it's endless. Yeah. And like the other thing is there are more and more tools coming out every single day for comfy UI because it's open source.

And so like these nodes are all just built by independent developers. They have nothing to do with comfy UI. And you can yourself go and build custom nodes as well. And so, yeah, I just think like, it's only going to get better and better. There's just going to be more and more capability within the program. And so, yeah.

What are some good resources for people to learn about how to do what you're talking about? So I've gone like down such a rabbit hole now. But I would say that Reddit is a good place to start. So if you go to the subreddit R slash comfy UI, it's quite small for now. I think there's like 30 or 35,000 members in it. But there's a ton on YouTube. There's a guy called, his channel is called Latent Vision. He's really good.

guy named Mateo from Italy. And he's also developed a bunch of custom nodes. And so he has videos explaining each custom node. And then he also has like a beginner video just explaining what comfy UI is. The one caveat is that it does take a lot of compute, like it takes a lot of GPU to run complex workflows. And so if you try to do this on like a basic Chromebook or something, it's just going to crash. And so if you really want to maximize the utility out of this thing, you have to get a virtual. machine and just be using like a really strong graphics card that's like sitting in a server

warehouse somewhere. What are some good companies that offer that? So we use a company called comfy deploy, but it's not very cheap. They have like a lot of additional features that they offer that I was willing to pay for a while I'm learning this thing. But it's like a thousand dollars a month. So that one's not cheap. But if you know anything about VMs, like you can use like I'm pretty sure you could just use like digital ocean or thin link client is another one that I use basically like you just look up any just go and look up GPU VM. Okay, virtual machine. Yeah, exactly. And it's cheap. Like you can get access to like a really nice graphics card and all of the compute power that you need probably for like 50 bucks a month. Wow. Okay. And you don't have to have much technical knowledge.

You don't have to be a software engineer to learn this. No. I will say, you have to be decent with computers. Or at least know how to use chat GPT quite well. Or yeah, exactly. You have to be resourceful. You have to be good at Googling stuff and looking it up on Reddit and following instructions. But I will say the learning curve is steep. Like I'm not going to sit here and tell you that like comfy UI super easy to use.

I think that's going to change over time. It'll become easier. But yeah, as of now, you don't need to be a coder, but you need to have some computer savvy. Yeah. imagine like this like the deadly trio skill set of knowing how to scrape leads knowing how to do cold outreach to those leads and then knowing how to do comfy UI and just kind of putting those together yes endless opportunities I'm going to dive down a rabbit hole as soon as I hang up here I'm sure you will I'm sure you will okay that's awesome well Mitchell thank you for your time

where can people find you if they want to find you I'm on Twitter as the ATM guy My handle is Mitchell underscore Sorkin. My email is Mitchell at Sorkin ATM. And that's pretty much it. Sorkanatm.com? Yeah. All right. How do we feel about the ATM biz?

Please leave me a rating, review. And just thanks for listening. Even if you don't, you know what? Thanks for being here. I appreciate you.