Think about the last time you paid to take your own money out of a machine. You probably winced, pressed “accept,” and forgot about it by the time you got back to your table. That tiny moment of friction, and the fact that almost everyone pushes through it, is one of the cleanest lessons about pricing I know. It applies to ATMs, and it applies to almost any small business you’ll ever own.
This post is about how ATM operators think about the fee they charge. But really it’s about pricing power: when you have it, when you don’t, and how to change a price without guessing.
People pay close to $5 to get their own money
Start with what the customer actually experiences. Bankrate’s 2025 fee study found the average surcharge at bank-owned ATMs hit a record $3.22, the fourth straight year it went up. On top of that, the customer’s own bank often charges an out-of-network fee, averaging about $1.64. Put together, a typical out-of-network withdrawal costs around $4.86.
Independent ATM owners, the people who put machines in bars and gas stations, usually set their fee somewhere between $2.50 and $3.50, higher at clubs and events. And people pay it. Not because they love it, but because the alternative is walking three blocks at midnight, or not buying the thing they came to buy.
That’s the lesson in one line: you’re not pricing the product, you’re pricing the alternative. The cash is identical at every ATM. What differs is how far away the next option is and how much the customer wants to be done.
The same fee is cheap in one place and expensive in another
A $3.75 fee inside a bar at 11 p.m. barely registers. The same $3.75 fee at a laundromat where regulars come every Sunday, with a gas station ATM charging $2.50 visible across the street, is a donation to the gas station.
Here’s an illustrative operator to make that real (a made-up composite, not a real person). Call her Dee. She raises the fee on the machine in a late-night pool hall from $3.00 to $3.50 and, over the next three months, withdrawals barely move, so she’s earning roughly 17% more per month from the same box. Feeling good, she tries a similar raise at a laundromat, from $2.75 to $3.25, and within two months about a quarter of her withdrawals are gone. She rolls it back.
Same person, same season, opposite results. Nothing about Dee changed. The only thing that changed was how close the customer’s alternative was.
I think about this a lot outside of ATMs. It’s the same reason a bottle of water costs more at the stadium than at the grocery store, and the same reason a vending machine in a busy break room can carry higher prices than one next to a corner store. I wrote about that side of it in why the right vending locations make all the difference. Location isn’t just about traffic. It’s about what else the customer can do.
The one formula that keeps you from guessing
Every price change is a bet that volume will move less than the price did. There’s a simple way to check the bet before you make it:
Volume you can afford to lose after a raise = 1 − (old price ÷ new price)
So if you go from $2.50 to $3.00, you can lose up to 1 − 2.50/3.00, about 17% of your customers, and still earn the same. From $3.00 to $3.50, about 14%. From $3.00 to $4.00, 25%.
And cuts are the mirror image, but steeper:
Volume you must gain after a cut = (old price ÷ new price) − 1
Cutting from $2.50 to $2.00 means you need 25% more customers just to stand still. Here’s why that’s so hard for an ATM: most people never see the fee until they’re already standing at the machine. A lower price doesn’t pull in new customers who never knew about it. It just gives money back to the ones who were coming anyway.
| Price move | What it takes to break even |
|---|---|
| $2.50 → $3.00 | Can lose up to ~17% of volume |
| $3.00 → $3.50 | Can lose up to ~14% |
| $3.00 → $4.00 | Can lose up to 25% |
| $3.00 → $2.50 | Must gain 20% |
| $2.50 → $2.00 | Must gain 25% |
That asymmetry is why I’m skeptical of “compete on price” as a small-business strategy unless your customers are actually comparison shopping. Most of the time, they’re not. They’re choosing between you and inconvenience.
How I’d test a price without fooling myself
Testing is where most people get it wrong, including in businesses that have nothing to do with ATMs. A few rules I’d follow:
- Change one thing at one location. If you raise prices everywhere in January and sales dip, was it the price or was it January?
- Give it 60 to 90 days. Regulars adjust slowly. The first two weeks almost always look fine.
- Compare to a control. Look at your other locations over the same weeks. If everything dipped, the price probably didn’t do it.
- Judge on dollars, not headcount. Fewer transactions at a higher price can be a win. For an ATM, it even means fewer trips to reload cash.
For ATMs there’s also a rules side. The fee has to be shown on the screen before the customer commits, so they can cancel. The old requirement to also stick a fee notice on the machine was removed by Congress in 2012 and dropped from the regulation in 2013. And state rules vary, so an operator should check before going aggressive.
The part that’s harder to put in a spreadsheet
I’ll be honest about the thing that bothers me a little. ATM fees land hardest on people who use cash because they have to, not because they want to. A $3.50 fee on a $20 withdrawal is 17.5%. For someone living paycheck to paycheck, that’s real money.
I don’t think that makes running an ATM wrong. The machine is a real service: it puts cash where there wasn’t any, and someone has to buy the machine, fill it, fix it, and carry the risk. But I think it’s worth pricing with your eyes open. The operators I’d respect price higher where people are paying for late-night convenience, and stay reasonable at the laundromat where the same families come every week. That’s not just ethics. As Dee found out, it’s also usually the more profitable choice. The rules of making money I believe in, which I wrote about in what are the rules of making money, start with giving real value for what you charge.
For the practical side of this, with the tables and the math laid out line by line, read ATM Surcharge Fee: How Much to Charge (Break-Even Math, 2026) over on vendbuddy.io.
Where this leaves you
If you’re thinking about an ATM route, the fee matters, but it’s a tuning knob, maybe 10% to 25% on a good machine. It can’t fix a machine in the wrong spot. The bigger decision is where the machine goes, which comes back to the same idea: find the places where customers have no easy alternative.
If you want to see what those places look like near you, VendBuddy (full disclosure: VendBuddy is my company) (vendbuddy.io/app) will list bars, laundromats, convenience stores and venues in a ZIP code along with who owns them, and credits come in one-time packs if you don’t want a subscription.
And if you’re not in the ATM business at all, try this instead. Pick one thing you sell, whether it’s a service, a product, or your own hourly rate, and ask: what’s my customer’s real alternative, and how far away is it? Then run the break-even formula on a 15% raise. You might be leaving money on the table for the same reason Dee almost did at that laundromat, just in the other direction: you never checked. If you’re still deciding whether to build something of your own in the first place, this is your sign to start your own business is where I’d start.
