Why First-Time ATM Owners Lose Money Quietly (and the Leaks I’d Watch)

Nobody loses money in the ATM business with a bang. There’s no crash, no margin call. It’s a slow leak: a machine that does four withdrawals a week, a vault that runs dry on a Saturday, a venue split that looked generous in the moment. You don’t notice for months, and by then you’ve told yourself it’s “passive income.” I want to walk through the leaks I’d watch for, because most of them start in your head before they show up in your bank account.

Disclosure: there’s one affiliate link below. If you sign up through it, I may earn a commission at no cost to you. I only share tools I’d use. I’m not a financial advisor and this isn’t financial advice.

A note on where this comes from. I haven’t run an ATM route. I have owned vending machines and written about the pitfalls I ran into there, and the ATM version rhymes almost line for line. The specifics below come from public rules and dealer and processor information I checked this month; the mindset stuff comes from me.

Leak #1: Falling in love with the asset

The first mistake is emotional. You watch a video, you run the surcharge math, and you want to own the thing. So you buy the machine. Now you have $2,500 of steel in your garage and a deadline in your head, which is exactly when people accept a bad location just to get it placed.

The fix is to reverse the order. The location is the business; the machine is inventory. Get a written yes from a venue whose customers actually need cash, then buy hardware sized for it. I’ve written about shiny object syndrome before, and an ATM is a very shiny object. It has a screen and it dispenses money. Of course we want one.

Leak #2: Confusing “passive” with “unattended”

An ATM needs less work than a snack machine: no expiration dates, no restocking chips. But it isn’t hands-off. The cash has to be loaded. Cassettes jam. Receipt paper runs out. Cellular modems drop. When a machine is down, it earns exactly zero, and you might not know unless you’re watching.

The quietest leak I can imagine is a machine that’s been “temporarily out of service” for five days while the owner assumes everything is fine. So I’d treat monitoring as part of the job from day one: turn on every alert the processor offers, and give the bartender or manager my number with one request: text me if the screen says out of service.

Leak #3: Forgetting the cash in the box is capital

Here’s the one that surprised me when I thought it through. Every ATM needs cash inside it, often $2,000 to $3,000 to start at a typical spot and more at busy ones. That money isn’t spent, but it isn’t working anywhere else either. Add a second machine, and you double it. The float scales with every placement.

And the demand for that cash is slowly shrinking. The Federal Reserve’s payment diary found cash was about 13% of consumer payments by count in 2025. That’s not zero, and certain places stay cash-heavy, but it tells me to be picky. I’d only put my money in a box where customers clearly still reach for cash.

I’d also keep a separate reserve for the boring stuff: a repair, a replacement part, a slow month. For me that would sit in a high-yield savings account like Marcus, apart from both my personal money and the vault cash. Federal bank guidance even notes that a dedicated account for ATM funds makes the money flow clearer to your bank, so separation isn’t just tidiness.

Leak #4: Buying someone else’s story

There are companies that sell ATMs “already placed” at a location, sometimes with a monthly management fee. Some are legitimate. Some are selling a weak location with a strong pitch. The story usually includes a monthly income figure that sounds reasonable enough to believe.

My rule: I’d only pay for history I can see. Actual monthly transaction reports for that terminal, the placement agreement, and the processing contract. If the seller won’t share them, the answer is no. Same rule for buying an existing route. Anyone confident in their numbers is happy to show them.

Leak #5: Being too generous to win the spot

This one comes from a good place. You want the location, so you offer the owner a big share of every surcharge. It feels like a partnership. Then you run the real math: after the venue’s cut, wireless, and processing, a quiet location might leave you with less than a hundred dollars a month on a few thousand in capital.

Here’s an illustrative example, a made-up person. Sam offers a laundromat owner 40% of the surcharge to beat a competitor. The machine averages about 60 withdrawals a month at $3. That’s $180 in surcharges, $72 to the owner, $20 for wireless, leaving roughly $88 before repairs. On $2,500 of machine and $2,000 of cash, that’s a lot of risk for $88. A 20% split would have left him about $124. Still modest, but the gap is his entire margin for a bad month.

I’d set a floor before walking in: the minimum I need to net per month for the capital I’m tying up. If the only way to win the spot is to go below it, the spot isn’t the win.

Two technical leaks that feel like bad luck

These aren’t mindset problems, but they catch people who skip the homework:

• The used machine that can’t run. PCI required a newer encryption key format (TR-31 key blocks) for ATMs starting January 1, 2025. Older PIN pads that can’t support it can be refused by processors. A $900 used ATM that can’t connect is $900 of scrap. I’d get written confirmation from the processor for the exact machine before paying.
• The contract you didn’t read. Processing agreements can run years and renew automatically. Read the cancellation clause before you sign, not when you want out.

For the practical side of this, with the tables and the math laid out line by line, read ATM Business Mistakes to Avoid: 12 That Cost New Operators over on vendbuddy.io.

The fix underneath all of them

Almost every leak above gets smaller when you have more options. If you know twenty possible venues instead of three, you don’t accept a bad one, you don’t overpay on the split, and you can move a weak machine without panic. If I were starting, I’d build that list first. VendBuddy (full disclosure: VendBuddy is my company) (vendbuddy.io/app) pulls bars, laundromats and similar businesses in any ZIP with the owner’s contact, and you can pay for one pack of credits instead of subscribing.

I wrote once about the ugly truth about vending, and the truth for ATMs is similar: the model works in the right spots, and the owner who’s honest about the leaks early is the one still running machines two years later.

If you’re considering this, here’s a journal prompt for tonight: What’s the smallest monthly net that would make this worth my capital and my Saturday nights? Write the number down before you ever talk to a venue. It’ll make every decision after it easier.