How do ESPs differ in bounce classification?

You migrated to a new ESP and your hard bounce rate is suddenly different from what you saw on the old platform. Did your list quality change? Maybe not. Bounce classification is not standardized across ESPs, and the same SMTP error code can be categorized differently depending on which platform you're using.

The most common divergence is around "mailbox full" errors (typically 452 or a 550 variant). Some ESPs classify these as soft bounces and keep retrying indefinitely. Others upgrade them to hard bounces after a few failures. Neither approach is objectively wrong, but if you're comparing metrics across platforms or migrating historical suppression data, you need to know which classification logic your old and new ESPs each use.

Suppression thresholds also vary. One ESP might suppress an address after a single hard bounce. Another might wait for two consecutive soft bounces before adding it to suppressions. That means an address your old ESP considered active might already be suppressed in your new one, or vice versa. Check your new ESP's documentation for their bounce handling rules before you try to interpret post-migration metrics.

For practical purposes, the key things to verify during migration: what SMTP codes does your new ESP classify as hard bounces versus soft? At what threshold does it suppress an address? And does it export suppression reasons in a format that maps cleanly to your old platform's categories?

If you're seeing unexpected changes in your hard bounce rates after migration, this classification difference is often the first thing to investigate before concluding your list quality has changed.