When a charge fails, something has to decide whether to try again, when, and how many times. In most setups that decision was made once by whoever installed the billing system, using the default. Ever since, it has quietly cost money.
Retry logic is worth an hour of attention. It operates on every failed payment you will ever have. Besides, the difference between a thoughtless schedule and a sensible one is measured in recovered revenue.
Not every decline deserves a retry
The first thing a retry system needs is the ability to tell soft declines from hard ones. A soft decline means try later: insufficient funds, a temporary block, a limit reached this month. A hard decline means stop: the account is closed, the card is reported stolen, the issuer has refused permanently.
Retrying a hard decline achieves nothing and does active harm. Issuers track merchants who hammer dead cards and treat their traffic accordingly. The reason code is in the response. Using it requires a provider that passes it through rather than collapsing everything into «failed». Ours appear in the processing reports by reason.
Timing: follow the money, not the calendar
The useful question is not how often to retry but when money is likely to be there. For consumers, that means salary dates. Most people are paid at the end or the middle of the month. An account that was empty on the first is rarely empty on the sixth.
A schedule that works for most consumer subscriptions looks like this. First retry after two or three days, second around the next likely pay date. Third about a week after that, then stop. Three attempts spread over two weeks beat five attempts in five days. That holds for recovery rate and for how the issuer sees you.
Business customers are different. Company cards fail for different reasons: a spending limit, an expired corporate card, an employee who left. Therefore retrying rarely helps without a human conversation. For B2B, fewer retries and an earlier message to a named contact work better.
Small details that change the result
Vary the hour. Retrying at the same minute each time meets the same account state. Shifting by a few hours catches deposits that landed overnight.
Avoid weekends for the first attempt. Transfers settle on business days, so a Saturday retry often tests the same balance as Friday.
Do not retry after a dispute. If the customer has already gone to the bank, another attempt reads as pressure and makes the dispute harder to defend.
Cap the total. Beyond three or four attempts, recovery rates fall to near zero while costs and risk keep accruing.
Where retries collide with other systems
Two layers retrying independently is a classic source of double charges. If your billing platform retries and the payment provider also retries, the customer can be charged twice. Moreover, you will hear about it from them rather than from a report.
Decide which layer owns retries, turn the other off, and document it somewhere the next engineer will find. The same goes for card update services. The update check should run before the first attempt, not between retries. Otherwise you spend attempts on a card you know is dead.
Where the recovered money lands matters too. If settlement and payouts sit in different places, reconciling partial recoveries becomes manual. A business IBAN keeps that in one chain.
What the customer should see
Retries happen in the background. Yet the customer experience around them decides whether a recoverable failure becomes a cancellation.
Say what happened in plain language and give a date. The payment did not go through, access continues until the tenth. Here is a link to update the card. Avoid the word «declined» where possible, since customers read it as an accusation. Mention the last four digits so people know which card to fix. That matters most for those paying several services from different cards.
Offering a second route helps more than another email. A customer whose card keeps failing may happily pay by transfer. So alternative payment methods turn a lost subscription into a changed one.
Common questions on this are in our FAQ.
How to tell whether yours works
Measure the share of failed payments that eventually succeed, split by decline reason. If insufficient-funds recovery is below a third, the timing is wrong. If recovery after expired cards is low, the problem is not retries at all but the missing card update step.
Review it twice a year rather than never. The pattern shifts when your customer base changes country mix or when you move billing dates. Nobody notices until someone looks.
Account setup takes from 5 days; the steps are on how it works.
Rates for middle-risk models start from 1.8% on the pricing page.
None of this requires new software. Retry rules, timing and messages are settings rather than features. So a team that spends one afternoon on them usually sees the effect within two billing cycles. A team waiting for a redesign waits indefinitely.
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.