A subscription business almost never loses customers loudly. Nobody writes an angry email when a card quietly fails on the third month. The charge does not go through, the account lapses, and the person is gone. By the time anyone looks at the numbers, the churn is already on the chart. Moreover, it looks like a product problem rather than what it actually is: a billing problem.
Recurring billing decides whether that revenue arrives on schedule. It is also the part banks look at most closely. A merchant charges the same card every month. Therefore it carries risks a one-off shop does not.
Why banks treat recurring billing as its own category
A single purchase is a closed story. Money moves, goods ship, and in a month
the transaction is forgotten. A subscription stays open indefinitely, which means
the bank carries an open-ended obligation on its books. The service may degrade, the company may disappear, the customer may forget what they signed up for. In every case, the disputes land on the acquirer.
That is why the model sits in the middle of the risk scale rather than at the bottom. It is not a restricted industry. Still, it is not as plain as a corner shop either.
Businesses in this band usually need a dedicated merchant account rather than a standard retail one. Besides, the paperwork is heavier, so planning early saves weeks.
Three things drive that classification:
- The charge happens without the customer present. Nobody types a card number on month four. So the issuer has fewer signals to confirm the payment is wanted.
- Disputes arrive late. A customer who notices a charge in month six disputes month six. Sometimes months three through five as well.
- Cancellation is a payments problem, not a support problem.
If the cancel button is hard to find, people dispute instead. Card schemes read
that as a merchant defect, whatever the terms of service say.
What you actually need to start
The setup itself is unremarkable, which surprises people who expect months of
negotiation. Connection takes from five days, and most of that time goes on
document review rather than technical work. What matters is that the documents
arrive complete the first time, because every missing paper restarts the clock.
You will be asked for company registration and ownership structure. Then a working site with visible terms. Finally, a clear description of what the customer buys and how often. That last point is where applications stall most often. A landing page promises «access» without saying what renews, when, and for how much. As a result, it reads as a risk signal rather than marketing. Our
connection process walks through
the stages in order.
Technically you need three things working together. First, a gateway that stores the card securely and charges it on schedule. Then a billing layer that knows who owes what and when. Finally, a settlement account that receives the money. Smaller
companies often run all three through one provider, while larger ones split them
as volume grows. Either way the processing side must support scheduled charges natively. An afterthought bolted onto one-off payments will not do.
Where the first month usually goes wrong
Two mistakes repeat. The first is treating the opening charge and the renewals as the same event. The opening charge has a present customer who can confirm anything. Renewals do not. The second is launching without a cancellation flow,
because the team assumes nobody cancels in month one. Somebody always does, and
if they cannot, the dispute costs more than the subscription was worth.
What breaks at scale
Everything above works until the portfolio gets large enough for statistics to
show up. Then three failure modes appear, and each one leaks money quietly.
Cards expire and get reissued. A customer who never cancelled
stops paying because the card in your system no longer exists. Scheme-level card update services exist precisely for this. They recover a meaningful share of those subscribers without anyone noticing a problem.
Charges fail for recoverable reasons. Insufficient funds on the first of the month is not a lost customer. It is a customer who gets paid on the fifth. A flat retry on the next day converts far worse than a schedule built around pay cycles. Yet most setups ship with the default.
Disputes concentrate in renewals. Because the customer was not
present, the merchant carries the burden of proof. Tie the subscription terms, the consent record, and the usage log to each charge. That turns an automatic loss into a winnable case.
None of these appear in a revenue chart as a line item. They show up as a slow drift. Typically it gets explained away as seasonality, until someone measures the failed-payment rate directly.
Choosing a provider: what to ask before you sign
Pricing is where most comparisons start, and for a subscription business the
headline rate matters less than its stability. Rates for middle-risk models start from 1.8%. However, the question to ask is what moves that number: volume, dispute ratio, your vertical. Also ask how much notice you get before it moves.
Our published pricing sets out the
structure rather than a single figure, because the figure depends on what you
sell.
Beyond the rate, four answers tell you whether a provider has run subscriptions
before:
- Settlement timing. When does money reach your account, and
does the schedule change during a dispute? - Retry control. Can you set your own retry logic, or are you
stuck with a fixed interval? - Card updates. Is the scheme update service included, and
does it run automatically? - Reporting on failures. Can you see why a charge failed,
split by reason, or only that it did?
If you also pay money out — to affiliates, partners, or creators — check whether
that runs through the same contract. Companies that bolt payouts on later reconcile two systems by hand. In addition, reconciliation errors cost more than the fees they avoided.
Revenue sometimes sits in a dedicated account before it moves on. In that case a business IBAN keeps the chain together.
What to do before you launch
Write down the renewal terms in plain language and show them where the customer
pays, not three clicks away. Build the cancellation path before the first charge
goes out. Decide what happens on a failed payment: how many retries, how far apart, what the customer sees. Write that down too. After all, the default is almost always wrong for your billing date. Then measure the failed-payment rate
from week one, so you know what normal looks like before something breaks.
Subscription models scale well because the hard part is set up once. However, a mistake in the setup scales with you too. The account looks slightly different for a subscription business. We covered that in a separate piece.
Connection from 5 days. Fees from 1.8% — transparent terms, no hidden charges.
Leave a request or book a consultation. We will put together the right setup for your niche and risk profile.