How to choose a payment processor: Bank Index scores Adyen, Square and Fiserv on the needs of an online store

Bank Index lists 9,987 payment and e-money companies, and an online store needs one of them, perhaps two. With a field that crowded, how to choose a payment processor stops being a hunt for options, because almost any store in Europe or North America already has dozens of licensed firms willing to take its card payments. It becomes a filtering job instead. Thousands of licensed firms must shrink to the few that fit one business, its sector and the countries its customers pay from.

Bank Index is the financial directory that BankStore publishes, and it grades processors on the same business needs it uses for banks, from the fees a merchant pays to the strength of the firm holding the takings. Each need gets its own mark. The hand-checked profiles add the evidence behind every mark, with a link to the processor’s own page.

Three of the fullest processor profiles got a close reading this week, Adyen, Square and Fiserv, along with the register-only entries a European store will also meet. No processor wins every line on its profile. The line that settles the choice differs for a marketplace, a restaurant and a subscription app, and I’d want any merchant to see that before a sales call starts.

How to choose a payment processor when the list runs to thousands

Start from the business, not from the brand. A processor that suits a restaurant with three card readers can be a poor match for a software company billing customers in forty countries every month.

Separate scores for each need make that split visible. In the Bank Index finder, a merchant selects the needs that count and the whole list re-sorts by those alone. The page sets a simple bar for a match: a provider "fits a need when it scores 5 or higher on it."

For most online stores, five lines carry the decision. Price comes first, in dollars per sale. After it come the online and subscription features, the speed of going live, the list of refused industries and the record of what merchants say once the contract is signed.

Strength matters too, I’d argue, though in a different way than at a bank, because a processor may hold thousands of dollars of a store’s takings at any moment. A processor holds a store’s takings for days at a time, and for longer when it keeps a reserve against chargebacks, so the company in the middle needs to be solid enough to pay out on schedule.

Is the price published, or negotiated?

Fees split processors into two camps, and I like that the profiles say plainly which camp each one sits in. Published pricing earns a decent mark from the index. A price agreed behind closed doors earns a low one, whatever it turns out to be.

On the Adyen profile the fee mark is 7.0, for "No setup fee and no monthly fee, a fixed USD 0.13 processing fee and interchange plus plus on cards." The same entry explains the ceiling: "The all-in cost depends on interchange the merchant cannot see in advance."

Square’s flat published rates earn it a 6.0. The entry lists "2.6% plus 15 cents in person and 3.3% plus 30 cents online on the free plan." On a 50-dollar online order, that rate comes to about 1.95 dollars. A flat rate is easy to budget around, though it can cost more than interchange-plus pricing once card volumes grow into the millions of dollars a year.

Fiserv lands at 3.5 on price, and the explanation is blunt. "No acquiring, Clover or platform price list is published on the site," the entry says. Contracts priced one by one, it adds, "made up 80 percent of 2025 revenue."

I don’t read that 3.5 as a claim that Fiserv is expensive. A store can’t compare its price with anyone else’s until a salesperson has quoted one, and the score reflects what a buyer can see before signing.

Online selling: the widest gap between processors

How to choose a payment gateway depends on whether a store sells once, sells by subscription or runs a marketplace that pays out to other sellers, and the online line on each profile is where processors differ most.

Adyen leads the three on this need with an 8.5, for unified commerce "across online and point of sale, marketplace payouts and local methods in more than 60 shopper markets through one integration." A store selling into many countries buys one technical project instead of a dozen.

Fiserv earns a 6.5 on the strength of its sheer reach, with acquiring and processing "across about 6 million merchant locations," the Clover point-of-sale system and a commerce product for larger merchants. Its entry also records that "There are no multi-currency collection accounts."

Square falls to 5.0 on online selling. Its tools suit shops and restaurants that also sell on the web, and the profile says it "has no dedicated recurring-billing or SaaS product beyond basic invoicing." A subscription business would notice the gap at its first renewal date.

Going live: self-serve signup or sales call

Onboarding speed decides whether a new store takes its first payment this week or next quarter. The profiles separate a self-serve signup from a sales process with a project plan attached.

Square tops this need among the three with an 8.0. Its banking side "activates instantly online once a Square account is approved, with no branch visit and no minimum deposit," according to the entry.

Adyen drops to 4.5 on onboarding, and its profile gives an honest reason. "Onboarding runs through a sales conversation instead of a self-serve signup, and Adyen publishes no time to go live." That model suits a large merchant with an integration team, and I’ve no quarrel with it. A shop that wants to take its first dollars by Friday will struggle with it.

Fiserv matches Adyen’s 4.5 for a similar reason. Much of its Clover business is sold through independent sales organisations, and the company publishes no application time or document list for a merchant to prepare against.

Payment processor reviews and what they leave out

Payment processor reviews are a tricky source, and the directory treats them with care. Merchants rarely post when their dollars arrive on time. They post when funds are held or an account is closed, so the review pool leans towards disputes.

Adyen sits at 2.5 for satisfaction, and the reasoning is spelled out: "Trustpilot 1.3 out of 5 from 442 reviews, a small review base for a company of this size." A few hundred unhappy merchants say little about a processor of Adyen’s scale.

Square does better at 5.5, on a Trustpilot average of 3.7 across thousands of reviews. The profile describes "mixed accounts of customer service and account management," which is roughly what a self-serve platform’s users would predict.

Fiserv’s European presence on Trustpilot is rated "1.8 out of 5, in its Poor band and below the 2.0 line." Credit card processor reviews of that kind skew towards complaints. The newest ones still deserve a look before signing, because they show how a processor behaves once a merchant needs help.

Will the processor take this sort of business?

Some industries are refused outright, and the high-risk line on each profile says which ones. For a store selling supplements, adult products or gaming items, this line decides the shortlist before price does.

Square scores lowest here, at 1.5, since its terms "prohibit cannabis and CBD products, adult entertainment and gambling or betting businesses." Adyen reaches 2.5, because no onboarding path for those operators appears on its published pages.

With a 3.5, Fiserv leaves the door a little wider. It publishes no prohibited-activities policy, and the entry says "Risk appetite is set contract by contract, with collateral demanded from merchants judged risky." That leaves room for a deal, at the price of a reserve held back from every payout.

For a store in a borderline category, I’d ask for the acceptable-use policy in writing. Learning about a refusal before integration costs a few emails and no dollars at all. Learning about it later costs weeks of frozen takings.

Strength of the firm holding the takings

A processor sits between a store and its money for days at a time. The strength line measures that middle firm. The evidence behind each mark varies a great deal from one profile to the next.

Adyen reaches 8.0 on strength as "A licensed European bank listed on Euronext Amsterdam and included in the AEX," with net income close to a billion euros in 2024. Fiserv also reaches 8.0 as an S&P 500 member, though its entry adds that "It carries no bank capital ratio or deposit insurance."

Square’s parent, Block, earns it a 6.5. The company is listed on the NYSE with total assets of nearly forty billion dollars, and the profile records pending state anti-money-laundering settlement talks disclosed in November 2024.

Strength offers no FDIC-style protection for a merchant’s balance. Takings in transit sit under the processor’s licence, safeguarded or held in trust depending on the country, and they are a different kind of asset from a deposit at an insured bank.

What do register-only processor profiles mean?

Many processors a European store will meet have shorter profiles, built from registers rather than from their own pages. Checkout Ltd, PayPal UK, Elavon and Global Payments all fall into this group for now.

Their lines repeat one sentence, saying the firm "moves money and issues accounts but takes no deposits and makes no loans." That sentence describes the licence type rather than the product, and the marks of four to six that follow are the defaults for firms of that type.

The strength line on these profiles is unusually frank. The Checkout Ltd entry calls its evidence "the weakest evidence of strength," built from how long the company has traded and how many people it employs. Then it adds a line I’d like to see on more financial websites: "A long history is not a balance sheet."

That candour helps me more than a confident number would. Our note on checking a processor’s licence explains how to confirm a register entry, and the directory of e-money and payment institutions shows which firms have full profiles and which only the register line so far.

Using the finder to build a shortlist

The finder converts all of this into a shortlist for a single store. A subscription software company billing in dollars and euros might pick online businesses, fees and onboarding, then limit the region to where its customers pay from.

One rule in the method keeps thin profiles honest. When a provider has no score on a need the merchant selected, that gap "counts as half of that provider’s average on the selected needs it does have," so nobody climbs the list on a line they never documented.

A restaurant group taking most of its dollars at the table would tick different boxes, with onboarding and fees at the top and online features lower down. The same directory then produces a different shortlist, and that is the whole point of scoring needs one at a time.

The profiles for Adyen, Square and Fiserv show every line discussed here, with each source a click away.

Limits of a processor score

The directory states the nature of its marks in its own method: "Scores are editorial judgments based on those facts and on the BankStore team’s aggregated experience." It also promises that the marks rest on public facts and use nothing from clients or partners.

No score can price a contract on its own. Interchange, scheme fees, chargeback costs and reserves depend on a store’s sector, its card mix and its history, and only a written quote puts dollars against each of them. A payments consultant or the store’s accountant can compare two quotes line by line far better than any public profile, and an hour of that work often pays for itself within a month of card sales.

Policies on risky sectors move as well, and every profile carries its date. The processor’s own terms on the day of signing are the version that binds the contract, so I’d save a copy of that page with the date on it.

Choosing by need, not by name

Set the three full profiles next to each other and the trade-offs are plain. Adyen leads on online selling and strength but onboards through sales. Square goes live fastest and publishes its price, yet refuses several industries and offers little for subscriptions.

Fiserv brings scale and strength to the table. Its prices, though, stay invisible until a salesperson names them, and for a small store that can mean weeks of back and forth before the first payment clears.

A store selling online in one country can narrow its list in an afternoon, starting from price and speed. For a subscription business selling worldwide, I’d start from the online line and treat price as the second question, since a cheap processor that cannot bill renewals in forty currencies costs more in lost sales than it saves in fees.

The full ranking sorts all 9,987 payment firms by whichever needs a merchant selects. The right processor is the one that scores well on those needs, and it is often not the most famous name on the list.