What does a payment processor actually charge you, once everything is counted?
A published rate is one of three cost models, and mixing them up is the most expensive mistake in this category. A blended flat rate quotes one number for everything: SumUp charges 1.69% on in-person transactions, and that number does not move.
Interchange-plus splits the cost into the card scheme's own fee plus the processor's margin: Adyen charges roughly €0.11 plus the scheme fee, which is cheaper at volume and impossible to predict at low volume. Per-method pricing gives each payment method its own line: Mollie charges from €0.29 for an iDEAL transaction and from €0.29 + 1.8% for a European card.
The crossover point matters more than the headline. A business processing €50,000 a month at SumUp's 1.69% pays about €845 in fees, where an interchange-plus provider would land well below that. Below roughly €10,000 a month the arithmetic reverses, because interchange-plus carries the overhead of an actual sales process and SumUp does not.
The costs that never appear in a comparison are chargebacks, currency conversion and the settlement delay. A provider that settles in two days rather than five is worth real money to a business with thin working capital, and no rate card mentions it.
Why do European local payment methods decide the conversion rate?
Payment preference in Europe is national, not continental. A Dutch checkout without iDEAL loses sales outright, because iDEAL is how the Netherlands pays online. Belgium expects Bancontact, Germany expects SEPA Direct Debit and increasingly invoice-based options, Austria expects EPS, and Switzerland expects TWINT.
This is the practical reason European merchants shortlist European processors. Mollie supports the full local set alongside cards, wallets and buy-now-pay-later, and adds new methods as markets adopt them. Payrexx is built around TWINT and the Swiss QR-bill, which no international processor treats as a first-class method. Adyen supports more than 250 payment methods across 175 currencies with local acquiring licences, which is the same argument at enterprise scale.
A card-only checkout in Europe is not a neutral choice. It is a decision to lose the share of customers in each market who do not reach for a card, and that share is much larger than most merchants assume before they measure it.
How much does the jurisdiction of a payment provider really change?
More than for almost any other software category, because a payment processor is regulated as a financial institution and holds both personal data and money. Two separate questions follow: which authority supervises the entity, and which law governs the data.
Adyen holds a full banking licence from De Nederlandsche Bank and is supervised as a bank under European banking rules. Klarna operates as Klarna Bank AB under a Swedish banking licence. Mollie is authorised by De Nederlandsche Bank as an electronic money institution. All three are supervised inside the EU, and all three process European transaction data in the EU.
GoCardless is established in the United Kingdom and authorised by the FCA; Payrexx is established in Switzerland. Both countries hold an adequacy decision from the European Commission, so transfers are lawful without standard contractual clauses — but an adequacy decision is a political instrument that can be reviewed, and it is not the same as intra-EEA processing. For a merchant that matters at the DPA level rather than at the checkout.
The contrast that makes European buyers look here in the first place is the US CLOUD Act, which can compel a US-established processor to disclose data it holds abroad. That is a live exposure for Stripe and PayPal that does not exist for a Dutch, German or Swedish processor.
When is a specialist processor better than a general one?
When the payment shape is the business model. GoCardless processes bank-to-bank Direct Debit and SEPA and nothing else — no cards, no wallets, no buy-now-pay-later. For a SaaS company, a membership organisation or any business billing the same customers every month, that narrowness is the point: Direct Debit costs less than card processing and does not fail when a card expires, which is where most subscription churn quietly comes from.
Klarna is the other specialist. Klarna is not a cheaper way to take a payment; at 2.5% to 5.99% it is considerably more expensive than a card. What Klarna sells is a larger basket and a higher conversion rate on mid-range retail, with the credit and fraud risk moved off the merchant's books. That trade only works if the uplift is measured rather than assumed.
The cost of a specialist is that it is never the only provider. GoCardless has to be paired with a card processor for one-off payments, and Klarna sits alongside a normal checkout rather than replacing it. Two providers means two reconciliations, and that overhead is real.
Where do payment integrations usually go wrong?
The first failure is treating settlement as instant. Direct Debit clears in three to five working days, and a business that plans cash flow around authorisation rather than settlement discovers the gap at the worst moment. GoCardless's Instant Bank Pay closes it for one-off payments, not for the recurring mandate.
The second failure is skipping Strong Customer Authentication planning. PSD2 requires SCA on European card payments, and the exemptions — low value, trusted beneficiary, transaction risk analysis — are where conversion is won or lost. Mollie applies exemptions automatically through risk analysis; a hand-rolled integration usually does not, and challenges every transaction.
The third failure is a single point of failure. A processor outage is a full stop on revenue, and merchants who have lived through one keep a second provider configured even if it carries no volume.
The fourth failure is outgrowing the pricing model without noticing. A flat rate that was obviously right at €3,000 a month is quietly expensive at €50,000, and nobody inside the business is paid to notice the month it flips.
