Single-provider sending means a single point of failure. When an outage, rate limit or reputation event hits the provider, the campaign stops. Multi-provider sending routes across Resend, Amazon SES and Mailgun so a problem on any one of them is rerouted automatically and the campaign keeps moving.
What can go wrong on a single provider
- Outage that lasts hours during your peak send.
- Rate caps you discover only when you hit them.
- Policy changes that block a category overnight.
- Reputation events on shared IPs you do not control.
How failover works
Each message is queued with a preferred provider. If that provider returns an error class that warrants failover (5xx, rate limit, defer), the message is rerouted to the next eligible provider with the right authentication for the sending domain. The recipient sees one email, sent normally.
What it requires
- DKIM signing keys for every domain on every provider you might use.
- A unified suppression list across providers (one unsubscribe is forever).
- Centralised reporting so you can see what went out and what did not.
- Cost rules: cheapest provider first, fallback when it fails.
Why this is normally a managed service job
Running three providers in production with consistent auth, suppression and reporting is non-trivial. Fortitude Send does this for you across Resend, SES and Mailgun. See /deliverability.
Frequently asked questions
Doesn't multi-provider cost more?
Per message, no. You pay each provider for what it actually sends. Total spend is usually lower because you can route to the cheapest provider first.
Will recipients see two emails?
No. Failover routes the same message to the next provider; it does not send twice.
Does each provider need its own warm-up?
The domain needs to be warm on each provider that will carry meaningful volume, yes. A managed pool keeps warming in the background.