Use case summary
Connect a second provider for the traffic that matters most, and set cascading rules for what happens when the first one fails. Corefy monitors provider and account availability and reroutes transactions to a healthy route when an issue appears. Result: when one provider has an incident, transactions continue through another, and your checkout keeps taking payments.
- Payment Manager
- Payment Ops
What a provider incident does to a single-provider setup
With one payment provider, every transaction depends on that provider being available. When it has an incident, payments fail or hang regardless of the customer's card, bank, or basket.
Incidents are not limited to full outages. A provider slows down under load, an acquirer's risk engine starts declining clean traffic, a merchant account reaches its daily volume cap in the middle of a sale. Each of these shows up as a drop in approval rate before anyone identifies the cause.
Teams with a second provider connected often still switch traffic by hand. Someone notices the drop, confirms the problem is not in their own system, and changes the routing. The transactions that failed during that time are usually not recovered.
How to set up failover across providers
Failover is a set of routing rules decided in advance: which provider takes the traffic when the primary one cannot, and under which conditions.
- 1
Connect a second provider for your critical traffic
Choose one that supports the same methods and currencies as your primary provider for the flows where downtime costs you most, then link your MID credentials.
- 2
Define the cascade order
Set provider priority and retry logic per payment method, market, or merchant account. A single cascade chain can include up to 20 merchant accounts.
- 3
Let availability monitoring move the traffic
Corefy continuously monitors provider and account availability and reroutes transactions to healthy paths when issues occur, without anyone changing the routing by hand.
- 4
Handle volume caps alongside outages
Turnover caps pause a route once it reaches its daily volume limit and resume it when the limit resets, so contractual limits at one acquirer do not stop processing.
- 5
Retry only what can be recovered
Cascading retries soft declines through another provider, up to five attempts by default, with caps per transaction and per card. Mastercard advice codes and Visa retry categories are respected, so hard declines are not sent around the cascade.
- 6
Cover payouts the same way
Payout routing supports cascading as well, so a payout that fails at one provider can be sent through a backup provider up to the attempt limit you set.
What you get
When one provider has a problem, the business keeps taking payments while your team investigates.
Revenue protected during provider incidents
Transactions that would have failed during an outage go through a second provider, so a peak sale or a product launch can continue while one provider is having problems.
Rerouting that happens without a person
Traffic moves to a healthy route automatically, so your team can investigate an incident without also redirecting transactions by hand while it is still going on.
Approval rates that hold under load
Volume caps, slow responses, and soft declines at one acquirer are handled by the next route in the cascade, which keeps your overall approval rate steadier during busy periods.
A working alternative at every renewal
A second live provider reduces your dependence on one company's pricing, risk policy, and roadmap, and gives you a real option when contract terms come up for review.
From one active integration to a stable multi-provider stack at 79% approval
Grand Cru Concepts, a software publisher running high-volume consumer products across the US and Canada, started with a single active payment integration and payouts tied to whichever provider had processed the deposit. It now runs its whole stack across several providers from one interface, with routing that keeps providers separated by geography, so adding a market-specific provider does not affect the routes serving other regions. As of May 2026, more than a year after go-live, all active providers were running stably; the approval rate across the stack stood at 79%, and no declines were attributed to downtime.