← All resources

Self-Checkout vs. Hiring a Cashier: The Real Math for a Small Stand

Self-Checkout vs. Hiring a Cashier: The Real Math for a Small Stand

The register is the one thing that ties a person to a stand. Not the produce, not the fridge, not the display — the moment of getting paid. As long as payment has to pass through a human, someone has to be there for every hour you want to be open. That's the real cost of a checkout, and it's the one most "self-checkout ROI" articles skip right past.

Search "is self-checkout worth it for a small business" and you'll find calculators built for a supermarket buying a $4,000 kiosk. That math has almost nothing to do with a farm stand, a coffee cart, a market stall, or a self-serve fridge. This guide runs the numbers that do fit — and is honest about where staying staffed still wins.

First, what a cashier actually costs

A checkout attendant's wage is the visible number, so start there. Here's what the U.S. Bureau of Labor Statistics reported for cashiers in its May 2025 wage survey:

What a cashier earns per hour (U.S., May 2025)Median wage: 15.81 $/hr; Lowest 10% earn under: 11.79 $/hr; Top 10% earn over: 19.43 $/hrWhat a cashier earns per hour (U.S., May 2025)15.81 $/hrMedian wage11.79 $/hrLowest 10% earn under19.43 $/hrTop 10% earn over
Source: BLS Occupational Employment & Wage Statistics, May 2025

The wage isn't the whole cost. On top of it, an employer pays the 7.65% employer share of Social Security and Medicare, plus state unemployment insurance and workers' compensation — realistically another 10–15% before you've counted the time spent scheduling, training, and covering the shift nobody shows up for.

But even the fully-loaded wage understates the problem for a small operation, because the real constraint isn't dollars per hour — it's that the register sets your hours. Your stand is open exactly as long as a person is standing at it. Want to sell eggs at 6 a.m. before the farmers arrive, or catch the dog-walkers who pass your fridge at 9 p.m.? You can't, unless you (or someone you pay) is there.

The monthly number

Turn that hourly wage into a monthly commitment and the decision gets concrete. Below is the base-wage cost of covering the register at the BLS median of $15.81/hour, for three realistic small-stand schedules — next to the seat cost of running an unattended till on a phone or tablet you already own.

Monthly cost to staff the registerHired help, 36 hrs/wk (6 days): 2467 $/mo; Hired help, 24 hrs/wk: 1644 $/mo; Hired help, 16 hrs/wk (weekends): 1096 $/mo; Unattended till (free Stand plan): 0 $/moMonthly cost to staff the registerHired help, 36hrs/wk (6 days)2467 $/moHired help, 24hrs/wk1644 $/moHired help, 16hrs/wk (weekends)1096 $/moUnattended till(free Stand plan)0 $/mo
Base wages at the BLS median of $15.81/hr. Payroll taxes and workers' comp add ~10–15%. Per-sale card fees apply to card sales either way, so they're left out of this comparison.

A note on that last bar so it isn't misread: $0 is the seat cost, not the whole cost of taking money. A per-sale card processing fee applies to card payments no matter how you take them — through a hired cashier's reader or through an unattended till — so it cancels out of this comparison. What the chart shows is the part that doesn't cancel: the labor or subscription needed just to keep the checkout open. On TallyTill's free Stand plan, that part is nothing, and it runs on a device you already have. (Paid plans, roughly $15–$35 per device per month, buy lower per-sale fees — check current pricing before you count on a number.)

Why the standard ROI math never closes for a small stand

Here's the part the kiosk-vendor calculators get structurally wrong for our world.

A self-checkout machine — the freestanding kind with a scanner, a scale, and a bagging area — runs $1,500 to $5,000 per unit, plus roughly $1,000 a year in maintenance (more if it handles cash). Their ROI pitch is: the machine costs $4,000, it offsets 40 cashier-hours a week, so it pays for itself in about a year.

That logic depends entirely on high volume. It only works if you have 40 cashier-hours a week worth of customers to offset in the first place. A weekend farm stand, an office fridge, or a market stall doesn't. If your stand offsets 16 hours a week, a $4,000 machine takes years to pay back — and by then the "savings" are swamped by maintenance. Run the small-volume numbers through the vendor's own formula and the honest answer is: don't buy the kiosk.

The mistake isn't the arithmetic. It's the hardware assumption. The number that actually changes the math is the one you can drive to zero: the equipment. A self-checkout that runs on a phone or tablet you already own has no capital cost to pay back, so there's no payback period to wait through. The break-even is immediate, at any volume — because you didn't spend anything to break even on.

That's the whole reason the small-stand economics work when the supermarket economics are shaky: you're not comparing a $4,000 machine against a cashier. You're comparing a device in your pocket against a cashier.

"But doesn't unattended checkout mean people just steal?"

It's a fair worry, and the research behind it is real: in big-box stores, self-checkout lanes see meaningfully higher shrinkage than staffed lanes — some studies put it around 3–4%, and most shoppers admit it's easier to walk out without paying.

But look at where that number comes from. It's the anonymous-crowd context: a stranger in a store of a thousand strangers, at a machine watched by no one they'll ever see again. A small stand is almost the opposite environment — regulars, a small community, often a camera, and an owner who knows what the shelf held this morning.

More to the point, the honest comparison for most of our readers isn't self-checkout versus a staffed lane. It's self-checkout versus the cash box you're using now. And there, a recorded digital checkout wins outright:

  • A cash honor box tells you how much money is in the box. It never tells you what left the shelf, or when, or whether the amount matches.
  • A self-checkout logs every sale with a timestamp — and, if you switch it on, a photo. You can see exactly what sold, reconcile it against what's missing, and spot a leak instead of guessing at one.

You don't eliminate the risk of a bad actor. You trade "I hope the money in the box is right" for "here's a record of every sale." For a stand that's been running on trust and a coffee can, that's not more exposure — it's a lot less. (We dig into where the losses actually come from in our guide on honor-system stand theft.)

What a till has to do that a person used to

If you're going to take the human out of the loop, the checkout has to quietly cover the jobs that person was doing. Here's how each one maps:

The four jobs a register does — and how an unattended till covers eachAnswer “how much is this?” — Prices, photos, and a search box on the screen replace the question a customer would have asked a person.; Ring up the item — The customer scans the barcode — or, for something with no label, snaps a photo and on-device AI names it. No barcode and no camera match? They search the catalog and tap it.; Take the payment — Tap-to-pay, card, or a QR code, right on the same device. No cash drawer to count or leave sitting out.; Record the sale — Every sale is logged with a timestamp so you see what sold and what to restock — the part a cash box can never give you.The four jobs a register does — and how an unattended till covers each1Answer “how much is this?”Prices, photos, and a search box on the screen replace the questiona customer would have asked a person.2Ring up the itemThe customer scans the barcode — or, for something with no label,snaps a photo and on-device AI names it. No barcode and no cameramatch? They search the catalog and tap it.3Take the paymentTap-to-pay, card, or a QR code, right on the same device. No cashdrawer to count or leave sitting out.4Record the saleEvery sale is logged with a timestamp so you see what sold and whatto restock — the part a cash box can never give you.

One thing worth being precise about, because plenty of self-checkout marketing overpromises it: the camera does not watch a counter and automatically ring things up as people set them down. The customer scans a barcode, or takes a photo so the AI can identify an unlabeled item. It's a deliberate tap, not a passive camera watching the room. That distinction matters when you're deciding whether the tool will actually work at your stand.

When staying staffed still wins

An honest guide has to say where the cashier is worth it. Keep a person at the counter when:

  • Every sale needs a conversation. If you're upselling, sampling, giving advice, or building relationships that drive the sale, that human interaction is the product. Automating it throws away the thing that works.
  • You legally must check something at the point of sale — age verification for regulated goods, for example — and can't reliably move that check elsewhere.
  • Your volume already justifies the staff and the line is the problem. If you're busy enough that one register is a bottleneck, the answer might be a second self-checkout lane alongside your staffed one, not replacing the person.

The pattern that fails is trying to automate away a cashier whose real job was talking to customers. The pattern that works is automating the checkout for the hours, days, or shelves where there was never going to be a person anyway — the early morning, the late evening, the weekday when the stand would otherwise just be closed.

The bottom line

For a supermarket weighing a $4,000 kiosk, "is self-checkout worth it?" is a genuinely hard question. For a small stand, cart, or fridge, it's a different question with a much cleaner answer, because you can take the expensive part — the hardware — off the table entirely.

The comparison isn't a machine against a cashier. It's the phone in your pocket against paying someone $1,000 to $2,500+ a month to stand at a counter, or against tying yourself to that counter for every hour you want to be open. When the checkout costs you nothing to run and no one has to be there for it, "open" stops being a function of who's available.

You can set that up today and take your first sale on hardware you already own. Create your free TallyTill account — the Stand plan has no monthly seat cost and doesn't ask for a card to start. Point a tablet at your shelf and see what it costs to keep the register open when the answer is nobody.

Wage figures: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025. Kiosk cost and shrinkage figures as cited above.

Reconnecting…
One moment — we're restoring your session.