How to Start an ATM Business from Scratch.

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 talk business ideas that have nothing to do with the ATM industry. One in the AI sphere that I think you'll 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'll rewind like a little bit. Spent Most of my 20s building Internet businesses, a polar opposite world. And at some point in 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 and we'll come back to you for a larger acquisition afterwards.

That'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 was 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 the multiple at the time. Ended up 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, then another 40, and then 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 we did not want to buy a landscaping or an H vac or a plumbing company and have to manage like 30 or 40 people. We just want 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 kind of like we'd rather 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, and the guy said, oh, I want to focus all my time on my ATM business. And so that kind of like, piqued our interest that we started going down that rabbit hole.

That sounds like Quinn Miller, right? You. I'm sure you follow him.

Yeah, yeah, he's a good friend of mine. That sounds like his story.

You mean 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. I mean, Quinn is a beast at cold calling, so, like, his. His whole growth strategy is completely different. Than ours in a lot of ways. Like, I'm jealous that he 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 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 1500. And really it's just like more admin work. But it's not like the labor intensity doesn't 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 City are you in again?

I'm in L A. L A.

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 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 thing is, ATMs are incredibly predictable in terms of looking at their historic performance and then projecting that, you know, 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 ATMs performance. So there are some details there where it's like you know, 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 risky, riskier to buy one location that does a thousand transactions than 10 locations that do a hundred 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 high value real estate and so they're constantly trying to get better deals. And so I, I kind of like veered off the path there.

But yeah, I do that all the time. I have ADHD, so it's going to happen like 12 times this interview. So 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, 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. You know, they will see you raise the ATM fee from 3 to 350 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.

And they're wrong.

Nonsense.

But it doesn't matter because it's 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.

So. Yeah, well, like in any business deal, even if contractually you can do that, you, you still want to maintain the good, healthy relationship, right?

Yeah, of course.

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 a higher 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 them, 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 hang ups 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 like 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 like 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 is like another 20 machines. I remember we, we had like 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% increase in net. And we were like, okay, let's just start doing this on all of the machines.

Okay.

And then we bought. 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 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 6500-7k a month today on 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. We were like, okay, let's take a breather and get accustomed to this. And then I think, like, maybe early that summer or mid that summer, I don't 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 2700amonth.

Wow.

Yeah. And all of the surcharges, like a 1x multiple. Yeah, less than 1x. 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 I sold it for so cheap because he was like, no contracts. I'm running out of time. So he just. He basically dumped it and we raised all the prices to 350. And that route still today makes like 4500amonth 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 small routes here and there as they popped up. And then, like last year, we did our two biggest acquisitions that added the vast majority of the 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 kind of like the trajectory.

All right, if you are watching this clip right now, you are watching this on YouTube. And if you are a fan of podcasts, please check out my podcast while you're at it. Because sometimes I have podcasts that aren't on YouTube and sometimes I have YouTubes that aren't on podcast podcast. So check out the Kerner office on any podcast platform. Or you can go to tkopod.com Also, my newsletter will throw that in there. There are no ads. It's weekly. It's different content than my podcast and YouTube newsletter.chrisj kerner.com you'll see the link and we'll see you there. 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. And interchange is basically this payment that's made from the networks to the processors. So 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. There's like a bunch of different factors. Nobody really. It's kind of 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.

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, above a middleman.

And.

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, okay, 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 EBITDA?

So our revenue across all the entities is around 180k a month. It's a little bit 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 that we brought investors in for. But our EBITDA margins have been in the range of 40% on our wholly owned entity since day one. I'd say that that's like relatively standard in the ATM business because the way that you really think about it is like, if there's a $3 surcharge, a dollar goes to the 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 on 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 have 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 come 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, like we, 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 kind of like a bad ROI. 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 place 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 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 it 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 gonna be a lot easier to load that than something that's like scattered all over the place. You'll 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 geography 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. 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 I'm 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. It's like, okay, how'd you get my number? He said, oh, the rabbi from my synagogue referred 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.

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 in my garage. I was like, how many transactions does it do? He's like, probably about a thousand a month. And like, you know, your audience doesn't have any context on that right now, but that is anything that does a thousand transactions a month is like anything above that is probably 1 in 5,000 ATM locations or 1 in 10,000. So I was like, okay, fine. I went in my garage, I programmed it. I'm Russian also. This guy ended up being Russian. We really hit it off. And so anyway, that's. That ended up being the best account that we have. When we first installed the machine, it did do a thousand transactions a month in the first month. And then the second month it did like 1400 and then 1600. And today it does, like, almost 2000.

Wow.

So then sentence or like the second part of your question. What are the best locations and the ones that I would like to copy paste? I'd say that. And then you also mentioned surprising. Yeah. I will say that 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, like, 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. Like 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 of 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, there's a few things, and this is just me guessing, but I'm almost certain that I'm right about this just from, 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. But when a store owner loads a 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.

Or I'm gonna go to a different place.

Or I'm gonna go to a different place. Yeah, exactly. So I suspect that that's what was happening there, especially because it was so high volume. So like a location like this, you need to have. 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 like 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 it's either 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 surcharge is split between you and the vaulter. 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.

It's not like to get into that business you have to have that much on hand.

No, absolutely not.

For 900 machines.

Yeah. No, no.

Is there a location that you placed that 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, you know, 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 LA where, like, first of all, I can't even tell you what the business sells, right? Like, you walk in, they'll have, like, mops and diapers and, like, a generator and, like, you know what I mean? You're just like, how?