How to read the effective rate on this site
The effective rate is everything a provider takes over three years, divided by everything the shop takes on cards. It carries three choices you did not make: a 36-month horizon, an assumed average sale, and interchange left outside the total.
Every plan in the provider list that publishes enough to compute carries one number in that column. It is not a score, and it is not a rating. It is arithmetic, and arithmetic depends on what you feed it. What follows is what I feed it, and where that treatment is unfair to somebody.
What goes into the number
Seven items go in. The card-present percentage, the per-transaction fee, the monthly fee, terminal rental, a published PCI fee, the provider's minimum monthly charge, and hardware bought outright spread over 36 months. The sum of those over three years is divided by three years of card sales. "What you keep" is 100% minus the result.
The arithmetic is the same for every provider, which is the only reason the column can be sorted at all. What card payments cost walks through the same formula with a shopkeeper's numbers instead of a definition.
What stays out
Five items stay out. Interchange on interchange-plus deals, chargebacks, refunds, foreign and business card surcharges, and tax. Each is left out for the same reason: the provider does not publish an amount I could add.
That exclusion is not neutral. A plan whose published rate covers only domestic consumer cards looks cheaper than one quoting a single rate for everything. SumUp's UK Payments Plus plan publishes 0.99% for domestic consumer cards and a separate 1.69% for non-EU or corporate cards (checked 2026-08-25). The column shows 1.07% at £25,000 a month, which is the 0.99% world, not yours if half your customers pay on a business card.
Why three years and not one?
Because a card terminal is bought or rented on that kind of horizon, and a one-month view would make every up-front cost disappear. The choice has a cost of its own: a 36-month window flatters providers that sell hardware outright and penalizes providers that rent it.
Square's UK Reader costs £19.00 to buy, and the plan's effective rate falls from 1.78% at £2,000 a month to 1.75% at £100,000 a month (checked 2026-08-25). The £19.00 is spread thinner as sales grow, then stops mattering. A rental never stops. Tyl by NatWest charges 1.39% plus £0.05 on both its Tap to Pay plan and its PAX A80 countertop plan (checked 2026-08-25). The only difference between the two is £13.99 a month for the machine.
| Tyl by NatWest plan | Machine cost | Effective rate at £2,000 a month |
|---|---|---|
| Tap to Pay | none | 1.64% |
| PAX A80 countertop | £13.99 a month | 2.34% |
^ Tyl by NatWest card machine pricing and fees page, both plans at 1.39% + £0.05 card present (checked 2026-08-25).
Seven tenths of a percentage point is the whole of the difference, and all of it is the rental. It is that large because £2,000 a month is a thin base to spread £13.99 across. At month 37 the bought reader has cost nothing further, and the rental has charged £13.99 again. Stretch the horizon and the ranking moves in the renter's disfavor; shorten it to twelve months and it moves the other way.
How much does the assumed average sale change the answer?
A great deal, on any plan with a per-transaction fee. The site assumes an average sale of $25.00 in the US and £20.00 in the UK. That is the only figure on this site that no provider published. I chose it, and it decides how heavily a fixed fee per sale lands.
Square's US Free plan is 2.60% plus $0.15 in person (checked 2026-08-25). Here is the same plan, the same monthly card sales, and two different assumptions about the size of a sale.
| Average sale | Card sales a month | Transactions | Cost a month | Effective rate |
|---|---|---|---|---|
| $25.00 | $25,000 | 1,000 | $800 | 3.20% |
| $10.00 | $25,000 | 2,500 | $1,025 | 4.10% |
^ Square US Free plan, 2.60% + $0.15 card present, from the Square pricing page (checked 2026-08-25). The $25.00 row is the figure the column prints. The $10.00 row is the same published rate under a different assumption, worked below.
Over 36 months the first line comes to $28,800 and the second to $36,900, a difference of $8,100 on identical published pricing (checked 2026-08-25). Nothing about Square changed. Only my assumption about the till changed.
Work out your own effective rate with your own average sale. A coffee counter and a furniture shop are not the same business on a plan with a fixed fee per sale. If your average sale is well above $25.00, the column overstates what a per-transaction plan costs you. If it is well below, the column understates it. The three numbers behind a card rate explains why the fixed fee behaves so differently from the percentage.
Why a percentage-only plan is treated as having no flat fee
Because that is the provider's own claim, and I would rather be wrong in the provider's words than wrong in mine. SumUp's UK Tap to Pay plan publishes 1.69% per transaction, states no per-item fee, and needs no hardware. The column therefore reads 1.69% at every volume from £2,000 to £100,000 a month (checked 2026-08-25). Epos Now Payments reads 1.70% across the same range for the same reason. Dojo's Essential plan on Tap to Pay reads 1.20% at £2,000 and £5,000 a month (checked 2026-08-25). Those are the only levels inside the £100,000-a-year band Dojo publishes it for.
A flat line across every volume is a claim, not a measurement. If one of those providers does charge a few pence per sale and does not say so, the column is too low. That error grows as your average sale shrinks. Method records which fields each provider left unstated, and the row's unconfirmed list in the data is the honest version of that gap.
What the column is biased toward, in one paragraph
It is biased toward providers that publish everything, sell hardware outright, quote one rate for all card types, and serve merchants whose average sale is near mine. It is biased against renters, against plans with a large fixed fee per sale, and against anyone who has to price a business-card-heavy till. It says nothing at all about plans priced as a markup on interchange, because no total exists for those. How to choose between two card payment providers covers the questions the column cannot answer for you.
Where the answer changes
- If your average sale is far from $25.00 in the US or £20.00 in the UK, rerun the arithmetic with your own figure before you compare two plans, because the fixed fee moves the ranking and the percentage does not.
- If you would keep a terminal beyond three years, a plan with rented hardware is worse than the column suggests, and a plan with a bought reader is better than it suggests.
- If your card takings move with the season, do not decide yet. Square's US Plus plan reads 4.08% at $5,000 a month and 3.20% at $50,000 a month (checked 2026-08-25), and a monthly fee is punishing in a quiet January.
Who this article is not for
This is not for anyone who wants a shortlist. The effective rate is one number produced by one formula, with three assumptions written on the Method page. It will not tell you whether a provider settles quickly, answers the phone, or supports your till. If you have already read the column and want a verdict, this article will only make the number less certain, which is the point of it.
Sources
- Square pricing page, https://squareup.com/us/en/pricing, checked 2026-08-25
- Square UK pricing page, https://squareup.com/gb/en/pricing, checked 2026-08-25
- Tyl by NatWest card machine pricing and fees, https://www.tylbynatwest.com/card-machines/pricing-fees, checked 2026-08-25
- SumUp UK card readers page, https://www.sumup.com/en-gb/card-readers/, checked 2026-08-25
- Dojo pricing page, https://dojo.tech/pricing/, checked 2026-08-25
- Epos Now payments page, https://www.eposnow.com/uk/payments/, checked 2026-08-25
How this article was made. Written by Kaz, an independent publisher. All figures come from the companies' own published documentation; each figure carries the date it was last checked. I have not used these products, and nothing here is a personal test result. The arithmetic is shown so you can recalculate with your own numbers. Found an error? Use the contact form; corrections are logged on the Editorial policy page.
Related articles
- The three numbers behind a card rate
- What a 'get a quote' button is for
- What the headline rate leaves out
Written by Kaz (pen name). Figures are copied from providers' published pricing pages on the dates shown. We have not used these products. Corrections: contact form.