Payment fees are quoted as a percentage plus a fixed amount, and the fixed amount is where the interesting behaviour lives. Because it does not shrink with the basket, the effective rate a store pays is not the rate on its contract: it is a function of what customers spend per order. This study takes an example store's 137.947 orders, prices every payment with the gateway rule that was effective on its own day, and sorts the result into basket bands. The rate falls from 3,45% on baskets under €40 to 2,64% above €150, on identical contracts, purely because of the fixed leg.
- Under €40 the blended rate is 3,45%; between €40 and €80 it is 2,97%; from €80 to €150 it is 2,78%; above €150 it is 2,64%.
- Nothing about the rate card changes across those bands. Only the basket does.
- That is 81 basis points between the cheapest and dearest band, which is real money at scale.
- It is also an argument for basket size that has nothing to do with shipping thresholds or merchandising.
Why a fixed fee behaves like a percentage
A fixed fee of twenty-five cents is 1,25% of a €20 order and 0,05% of a €500 one. So a store selling small baskets is paying a materially higher effective rate than a store on the identical contract selling large ones, and neither has done anything differently.
This is why the only fee number worth acting on is the blended fee rate computed from your own payouts: total fees over total charged volume, which already contains your basket profile, your gateway mix and every split payment charged twice.
| Basket | Orders | Average charged | Blended fee rate |
|---|---|---|---|
| Under €40 | 29.946 | €30 | 3,45% |
| €40 to €80 | 30.873 | €62 | 2,97% |
| €80 to €150 | 44.879 | €108 | 2,78% |
| Over €150 | 32.043 | €235 | 2,64% |
What 81 basis points is worth
On a store doing a million euros of charged volume a year, the gap between the cheapest and dearest band is €8.100. Nobody negotiates that with a gateway; it is won or lost by what customers put in the basket, which is a merchandising question rather than a payments one.
It also compounds with everything else that is charged per order. The parcel, the picking and the packaging behave the same way, so a small-basket store pays more on four costs at once, and each of them is invisible in a percentage-of-revenue view.
The levers this actually points at
Raising the average basket is the obvious one, and the free shipping threshold calculator is the tool for pricing whether an offer to do so pays for itself. Bundles, minimum order values and pack sizes all move the same number.
The second lever is payment mix, because methods differ in their fixed component as well as their percentage. A method with a low percentage and a high fixed fee is expensive precisely where this study says you are already expensive, and a Shopify payment fees calculator will price your own mix against a proposed change before you make it.
Why the effect is bigger than the table looks
The bands above are averages inside wide ranges, so they understate the extremes. An order of EUR 15 pays a far higher effective rate than the 3,45% of its whole band, because the fixed leg is a larger share of a smaller amount and the band contains everything up to EUR 40.
Split payments compound it. An order paid with a gift card and a card is two transactions, so the fixed fee is charged twice on two smaller amounts, and those orders sit disproportionately in the lower bands where the fixed leg already hurts most.
The trap in reading this backwards
A high fee rate on small baskets is not an argument against selling small baskets. Those orders may be profitable, they may bring first-time customers, and a store that stopped taking them would lose the volume as well as the rate.
The argument is narrower and more useful: know that small orders cost more to process, price them accordingly, and stop assuming a single blended rate describes every part of the business equally. A statement that prices each order at its own gateway rule already knows this; a spreadsheet applying one average rate does not.
About this data
137.947 orders from an example store's own order book, generated rather than taken from any merchant, with each payment priced by the fee rule effective on the day it was processed rather than by a single average. Your own bands will differ with your gateway mix and your fixed fee, which is exactly why the exercise is worth repeating on your own payouts; the rules behind every figure here are in how every figure is computed.