Does routing replace fallback?
No, they work in sequence. Routing chooses the first provider for a payment. Fallback decides what happens when that attempt fails: retry, reroute or stop, based on the return code. An operator needs both, and fallback across providers also requires credentials that each provider can use.
In detail
No, they work in sequence. Routing chooses the first provider for a payment. Fallback decides what happens when that attempt fails: retry, reroute or stop, based on the return code. An operator needs both, and fallback across providers also requires credentials that each provider can use.
Two decisions in sequence
Routing answers the first question: which provider gets this payment? It uses what is known before the attempt, such as card type, country, amount and merchant preferences. Fallback answers the second: the attempt failed, what now? It uses the return code to decide between retrying on the same provider, rerouting to another, or stopping.
Why both are needed
- Routing without fallback sends a payment to the best provider and gives up if it fails.
- Fallback without routing retries blindly, including on hard declines that should stop the flow.
- Together, they pick the best first attempt and recover the recoverable failures.
The credential condition
Fallback across providers only works if the second provider can process the payment: a stored card must be a token the second provider accepts, which is why network tokens and a universal vault matter. A wallet payment needs a processor that supports the wallet. Operators should check, for each fallback path, that the credentials and configurations exist on both sides, and test the path with real traffic rather than discovering a gap during an outage.