What’s a rate limit and how do ESPs manage it?

You send a campaign. Some messages bounce back with a 4xx code instead of delivering. A few minutes later, the same addresses succeed. That's a rate limit in action.

A rate limit is a cap that recipient servers place on how much mail they'll accept from any single IP or domain within a time window. Exceed it and you get temporary deferrals, not permanent rejections. The server is saying "not right now" rather than "never." Most sending systems will retry automatically until the limit relaxes.

Rate limits come in a few forms. Connection limits cap simultaneous open connections from your IP. Message-per-connection limits restrict how many messages you can send per connection before reconnecting. And hourly or daily volume caps put a ceiling on total mail accepted from your sending address. Major mailbox providers adjust these based on your sender reputation. Established senders with clean history get generous limits. New or questionable IPs get tight ones.

ESPs manage this behind the scenes. They have dedicated IP pools with established reputations, they track per-provider limits in real time, and they queue and retry messages automatically when limits are hit. If you're using a well-configured ESP, rate limits are mostly invisible to you unless you're pushing very high volumes or your reputation dips. If you're running your own infrastructure, you'll need to configure per-destination concurrency and rate delays in your MTA manually. Watch for deferral patterns in your logs. Persistent throttling from a specific provider usually means your sending volume is too aggressive for your current reputation level.