Dunning is the unglamorous name for everything that happens after a charge fails. That means the retries, the emails, the grace period and the decision to close the subscription. Most teams configure it once during setup, using whatever defaults the billing system shipped with. Afterwards nobody looks at it again.
Those defaults are usually wrong for your business. They were chosen to be safe for everyone rather than effective for anyone. Tuning them is one of the few projects where a week of work produces a permanent revenue increase.
Contents
- Retries: timing beats frequency
- Messages: fewer, clearer, earlier
- The grace period is a commercial decision
- What to do before any of this
- Who owns this
- Measuring whether it works
Retries: timing beats frequency
The instinct is to retry quickly and often. In practice the opposite works better. The most common reason for a soft decline is an empty account. Besides, an empty account does not refill in six hours.
Build the schedule around when people get paid rather than around convenient intervals. Try a first retry after two or three days and a second around the time salaries land. A third about a week later covers most recoverable cases. Four attempts in four consecutive days covers almost none of them. Moreover, the issuing bank starts treating your merchant ID as noisy.
Two refinements matter. Vary the time of day. A charge at three in the morning meets a different account balance than one at noon. And stop retrying permanent declines — a closed account will not open because you asked five times. That distinction requires decline reasons, which is why they appear in our processing reports.
Messages: fewer, clearer, earlier
The email sequence usually looks like three identical messages with increasing urgency. It works better as two different messages with different jobs.
The first one is informational and goes out early. Your payment did not go through, access continues until a date. Here is a one-click link to fix it. No alarm, no threat, no account jargon. The second one goes out shortly before access ends and says exactly what will happen and when.
After that, stop. A third and fourth reminder recover almost nobody. Instead they generate spam complaints, which cost more than the subscription. If the person has not acted after two clear messages, the subscription is over. The right move is a clean cancellation with the door left open.
The grace period is a commercial decision
How long you keep access alive after a failed payment is not a technical setting. It merely lives in a technical system. Cutting access immediately protects revenue and loses customers who would have paid on the fifth. Keeping it for fourteen days recovers more of them. However, it gives away two weeks of service to people who were leaving anyway.
The answer depends on marginal cost. A SaaS product with near-zero delivery cost can afford a long grace period. A service with real per-customer costs cannot. Decide it as a business question and write it down. Revisit it once a year rather than when someone complains.
What to do before any of this
Dunning recovers payments that failed. Preventing the failure is cheaper. That is why card update services run first, while dunning picks up what they could not fix. Some teams implement dunning without card updates. Then they email customers about a problem that could have been solved silently.
The same logic applies to authentication. If renewals fail because a confirmation request goes unanswered, no retry schedule will help. The fix sits in how the charge is submitted, not in what happens afterwards.
The steps on our side, from application to the first charge, are set out on how it works.
Who owns this
Dunning falls between departments, which is why it stays on defaults for years. Finance sees the revenue gap, support sees the angry emails, product owns the billing screens. Meanwhile, nobody owns the retry schedule.
Give it to one person with the authority to change the schedule and the message text. Then review it quarterly. The review takes twenty minutes: look at the recovery rate, look at complaints, change one thing. Teams that do this twice a year end up well ahead. Others discuss it constantly and change nothing.
Which payment methods you support affects this too. A customer who can switch from card to direct debit solves the problem permanently.
Questions that come up often are answered in our FAQ.
Measuring whether it works
One number matters: the share of failed payments that eventually succeed. Everything else — open rates, click rates, how many emails you sent — describes activity rather than result.
Track it monthly and split it by failure reason. After all, recovery from insufficient funds and from an expired card are different problems. A base recovering more than half of its soft declines is doing well. One recovering under a quarter has something misconfigured, usually the retry timing.
The reporting that makes this visible is part of the merchant account rather than an add-on. The account opens from 5 days.
Commercial terms, from 1.8% for middle-risk models, are on the pricing page.
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.