How do ESPs differ in bounce classification?

Bounce classification isn't standardized across ESPs. The same SMTP error might be categorized differently by different platforms.

Common variations:

Hard vs soft threshold: Some ESPs classify certain errors as hard bounces; others call them soft. "Mailbox full" is soft to some, hard after N occurrences to others.

Suppression rules: Different policies for when to suppress. Some suppress after one hard bounce; others after multiple soft bounces over time.

Category granularity: Some ESPs have many bounce categories (unknown user, domain not found, policy block, etc.). Others simplify to just hard/soft.

Block vs bounce: Reputation-based rejections may be classified as bounces or as a separate "blocked" category.

Why this matters during migration:

An address suppressed as "hard bounce" at old ESP might not match new ESP's criteria

Import all suppressions regardless of category to be safe

Don't assume categories translate directly

Review new ESP's documentation on classification logic

Post-migration considerations:

Historical bounce rates may not compare directly between ESPs

Suppression list size may differ due to different policies

Set baseline metrics fresh at new ESP rather than comparing to old

Understand the new ESP's logic to interpret ongoing metrics correctly.