Hi @duncanc, thanks for prompt feedback.
Yes I could be wrong but my understanding is that some services like SES sends a “special” message, not a bounce, when a message is delivered to a client and he manually marks it as SPAM (https://docs.aws.amazon.com/ses/latest/DeveloperGuide/e-faq.html#e-faq-cm).
Messages are basically composed of:
Subject: complaint about message from 54.XXX.X.XX
(edited complete sample)
Content-Type: text/plain; charset=us-ascii
This is an email abuse report for an email message from amazonses.com on Sun, 15 Sep 2019 09:01:[47 +0000](https://xxxxxdomain/admin/47%20+0000)
Received-Date: Sun, 15 Sep 2019 09:01:[47 +0000](https://xxxxxdomain/admin/47%20+0000)
Authentication-Results: authentication result string is not available
And include the body of your message.
On our manual process of bounces my method lately has been:
- remove all possible false positives from a large grep rule. Ignore and delete bounce.
- unsubscribe user if he issued a complaint.
- some other direct rules for permanent errors… (domain not existant, relay not permited, etc). Blacklist email, unsubscribe
- let the rest count a bounce
My intention is to change the second one for a “courtesy email to unsubscribe” list instead of a direct unsubscribe. I could do it from within AWS itself as I could possibly call the API from there (instead of receiving an email) but that requires a bit of setup in AWS.