The coinbox only knows quarters
Every price a coin machine can charge is a multiple of the smallest coin it accepts, and for most stores that is a quarter. So when the cost arithmetic says $4.60, the real choices are $4.50 and $4.75, and one of those is 2.2 percent under the price you wanted while the other is 3.3 percent over. Across a store doing a few hundred turns a day that gap is a four-figure sum a year, and it is invisible because nobody ever writes down the price they could not charge.
The awkward property is that the gap is not stable. It depends entirely on where your ideal price happens to land relative to a quarter, so it can be nearly zero one year and near the maximum the next, purely because your gas bill moved. The table at the bottom of the page shows that by walking the target price a few cents either way.
What a card actually costs per turn
Two fees, and they behave completely differently. The percentage scales with the vend and is roughly neutral across machine sizes. The fixed per-transaction amount does not scale at all, and it is brutal against small vends: thirty cents against a ten dollar wash is three percent, and against a two dollar dryer vend it is fifteen. That is why the field asking how many turns a customer pays for in one transaction matters so much. A system where a customer tops up a balance once and starts four machines spreads that fixed fee four ways; a system that charges per machine start does not spread it at all.
Against those fees sit two things coin costs that nobody puts on a statement: the hours spent emptying, counting, rolling and banking, and the fraction of the counter readings that never turns into a deposit. Both are entered here as your own measurements rather than assumed, because both vary enormously between a store somebody visits daily and one visited weekly.
Neither side is the answer
The page prints the difference for the numbers you supply and stops there, deliberately. Which way it comes out turns on the card share, the per-transaction fee and how many turns share a transaction, and all three vary so much between stores that any general claim about coin or card is close to meaningless. Run it twice at the ends of whatever range you believe for the card share; if it comes out the same sign both times, you have an answer, and if it does not, you have learned that the card share is the thing to find out.
What the arithmetic does not contain
It holds the turn count fixed, and that is the biggest thing it cannot know. Removing the trip to the change machine plausibly increases turns; raising the price plausibly reduces them; being unable to take payment when the network is down definitely reduces them. None of those are in here. What is in here is the accounting piece — the pricing gap, the fees, the handling and the losses — which is one input to the decision and not the decision.
Questions people ask
What does rounding a vend price to a quarter cost?
Anything between nothing and half a step per turn, depending on where the price you want falls. If the arithmetic says $4.60 and you round down to $4.50, you are ten cents a turn under; at 420 turns a day that is about $1,280 a month. The page prints it for your own target and then walks the target a few cents either way, because the gap moves every time your costs move.
Is a card system cheaper than coin for a laundromat?
It depends on the card share, the per-transaction fee and how many turns share one transaction, so the page computes the difference for your numbers rather than answering it in general. Against the fees sit the coin handling hours and the share of the counter readings that never reaches the bank, both of which you enter from your own measurements. Run it at both ends of your card-share guess and see whether the sign changes.
Why does the per-transaction card fee matter so much?
Because it does not scale with the vend. Thirty cents on a ten dollar wash is three percent; the same thirty cents on a two dollar dryer vend is fifteen. Anything that lets one transaction pay for several turns — a balance top-up, a card that starts four machines — divides that fee across them, and it is usually the single largest lever in the comparison.
Should I round the vend price up or down?
The page will show you both but it cannot decide it, because the answer depends on how many turns you lose at the higher price, which is a question about your customers and your competition. What it does give you is the size of the gap in both directions, so the trade is explicit: rounding up is worth a specific amount per turn only if it costs fewer turns than that.
Does the calculator assume cards increase business?
No, and that is deliberate. The turn count is held fixed on both sides, so the figures cover only the pricing gap, the processing fees, the coin handling and the coin losses. Removing the trip to the change machine may well increase turns and a network outage will certainly reduce them, but neither is something this page can size for your store, so neither is in the arithmetic.