Ask a high-street bank for "merchant services" and you will be quoted a rate and a contract. What that phrase actually covers is a bundle of four separate things, and knowing where the seams are is most of the skill in comparing providers.
There is the acquiring relationship, which is the licensed institution that receives the money from the card schemes on your behalf. There is the capture method, the gateway or terminal that reads the card. There is settlement, the movement of the money into your account, on a timetable and in a currency. And there is the wrapper: risk monitoring, dispute handling, compliance obligations and whoever answers the phone.
Sold as one product, this is convenient. It becomes inconvenient the first time one component stops fitting, because bundled products rarely let you change one part without renegotiating all of it.
Why some businesses get declined, and why it is not personal
The single most common experience that sends a business looking for alternatives is a decline it does not understand, or an offboarding letter citing nothing more specific than commercial reasons.
It is worth understanding what is happening, because it is rarely a judgement about your particular business. Large acquirers run sector risk appetites: whole merchant category codes are scored, and the scoring reflects the aggregate behaviour of that category rather than yours. If your code carries elevated chargeback rates or regulatory exposure across the industry, you inherit the category's reputation on the day you apply.
The economics follow from scale. A provider onboarding thousands of merchants a month cannot underwrite each one individually and stay profitable, so it underwrites the category and accepts that it will decline good businesses inside bad codes. That is a rational trade for them and an infuriating one for you, particularly when nobody will explain it.
Which is why the useful question when comparing providers is not "will you accept me" but "how do you decide, and will you tell me why if the answer is no".
The nine things worth comparing
Headline rate is the least informative number available, because it is the one every provider optimises for display. These are the axes that actually differ.
Sector appetite. Does the provider underwrite your business or your category code? Ask directly whether your sector is on a restricted list, and ask before you apply rather than after.
How a decline is explained. Some providers will tell you the reason and what would change it. Others will not, sometimes for legitimate legal reasons. Knowing which you are dealing with is worth more than an optimistic quote.
Onboarding time, honestly stated. Every provider quotes a best case. The useful question is what makes it slower, because the answer is almost always incomplete documentation, and a provider who tells you precisely what complete means before you start is saving you weeks.
Routing. A single fixed processing route means a declined transaction is simply declined. Where more than one acquiring path is available, a transaction that fails on one can be retried on another, which shows up as an approval rate rather than a feature. Ask what happens to a soft decline.
Reach and settlement currency. If you sell abroad, ask which currencies you can settle in rather than which you can accept. Accepting a euro payment and being paid in sterling means a conversion you did not choose, on a rate you did not negotiate, on every single transaction.
Pricing transparency. Ask for every fee category in writing before you sign: processing, gateway, monthly minimums, PCI charges, reserves, chargeback fees, and any sector loading. A provider who names them all is comparable. A provider who quotes one blended number and changes the subject is not.
Who you can actually call. A named contact matters least while everything works and most on the day settlement stops. Ask whether you get one, and whether they are still there after go-live.
Settlement timing. Ask how long after a sale the money actually reaches you, and whether that changes for the first few months. New merchants are commonly settled more slowly while a risk profile builds, which is a reasonable thing for a provider to do and an unreasonable thing to leave unmentioned. A business that plans cash flow on the eventual timetable rather than the opening one can find its first quarter considerably tighter than the spreadsheet promised. Ask for both numbers, and ask what moves you from one to the other.
What happens when a dispute arrives. Ask who assembles the evidence, whether you get early warning of a chargeback, and what the fee is for losing one.
What it costs to leave
Worth establishing before you sign the next agreement, not after.
Four things determine how easily you can move: the minimum term, the notice period, any early termination charge, and whether your terminal sits on a separate lease.
The terminal lease is the one that catches people. It is frequently a distinct finance agreement with its own term, arranged through a third party, and ending your processing contract does not necessarily end it. Businesses discover this when they have already switched and the direct debit for hardware they no longer use keeps going out.
None of this makes leaving impossible. It changes the timing, which is precisely why it belongs in the comparison you do at the start.
When the incumbent is the right answer
Often, and this is worth saying plainly rather than burying.
For a mainstream, single-market retailer with a bank that is happy to serve it, a traditional acquiring relationship is frequently perfectly good, and the pricing has improved considerably over the last decade. If your card mix is domestic consumer debit, your sector raises no eyebrows, and your existing provider answers the phone, the honest advice is to stay and use the eight questions above at your next renewal instead.
The businesses for whom this comparison matters are the ones outside that profile: an unusual sector, several markets and currencies, high average transaction values, a growth curve that keeps tripping risk thresholds, or a history that takes a conversation to explain. Bundled products are built around an average customer, and the further you sit from that average, the more the bundle costs you in things that are not price.
Where this group sits
Stated plainly so you can weigh it: the group underwrites businesses on their merits rather than by category alone, routes across more than one acquiring path (a 97% average approval rate across the acquiring network), reaches 180+ countries and 40+ settlement currencies on that network, and quotes per business with every fee category named before you commit rather than publishing a rate card. Onboarding runs to as little as 48 hours, subject to all relevant documentation. You get a named specialist rather than a queue.
That is a fit for some businesses and not others, which is the entire point of the section above it.