Turn a soft decline into an approval.
When a bank softly declines, it is rarely saying never — it is saying not like that, not right now. A blind retry just annoys the issuer and risks the relationship. A smart retry answers the hesitation: the right route, the right moment, fresh credentials. And it never, ever touches a genuine decline.
What happens the moment a bank hesitates
The whole decision runs in the time it takes the checkout to settle. The shopper sees one outcome; behind it, four things have already happened.
Catch the soft decline
The issuer returns a "not now" code. We catch it instantly and confirm the card and funds are genuinely fine.
DetectDecide if it is worth it
Soft and recoverable? It qualifies. Hard — fraud or a reported card? The loop stops here and the decline stands.
QualifyChoose route and moment
We pick the path the issuer is most likely to approve and the moment it is most likely to clear — sometimes now, sometimes shortly after.
RouteRetry once, then stop
We try again with fresh credentials. If it clears, the sale is saved; if it does not, we stop — no hammering, no fee churn.
RecoverNot "try again" — try better
A naive system fires the identical request at the identical path and hopes. A smart retry changes the variables that caused the hesitation in the first place, so the second attempt is materially more likely to clear than the first.
- The right route. We send the attempt down the path that issuer approves most often, not the one that just failed.
- The right moment. A temporary outage or a low balance clears with time — so we wait exactly as long as the reason needs.
- Network-token credentials. The retry carries a token kept current at the network, sidestepping expired or reissued cards.
- Issuer-aware limits. We respect each bank's appetite for retries, so recovery never strains the relationship.
Blind retries lose; smart retries recover
Blind retry
The same request, the same route, fired again and again. It annoys the issuer, can flag your traffic as abusive, racks up authorization fees and rarely changes the outcome — because nothing about the attempt changed.
Smart retry
One considered attempt that changes what mattered: a better route, the right moment, fresh credentials, within each issuer's tolerance. It recovers the sale without straining the bank — and it knows when to stop.
Every recovery shows its work
Each line is a payment that first came back as a soft decline, then cleared on a smarter attempt — with the reason, the action and the amount in plain view. The hard declines sit at the bottom, deliberately untouched, so you can see what we recovered and what we left alone.
- Reason in, result out. Read the original decline and the move that recovered it on the same row.
- Counted once, where it cleared. A recovered payment is booked at the moment it actually succeeds, never double-counted.
- Hard declines stay held. Suspected fraud and reported cards never enter the retry loop — full stop.
Disciplined recovery, measured
How the retry engine behaves
Do you ever retry a hard decline?
Never. Suspected fraud, a stolen or lost card, or any code that signals a firm "no" exits the loop the instant it is classified. The retry engine only ever touches soft, recoverable declines — answering a hesitation, not overruling a refusal.
Won't repeated attempts upset the issuers?
That is exactly what blind retries do, and it is what we avoid. We retry within each issuer's known tolerance, change the route and credentials so the attempt is meaningfully different, and stop after a single considered try. Recovery should strengthen the issuer relationship, not strain it.
How do you decide whether to wait or retry now?
The decline reason sets the timing. An issuer outage or a momentarily low balance clears with a short wait, so we hold and try again inside the recovery window; a routing or credential issue is better solved immediately. We match the moment to the cause rather than using a fixed delay.
Recover the sales a soft decline nearly took.
Bring your soft-decline volume to a revenue review and we will model how many of those maybes the retry engine would turn into approvals — and what they are worth.