I have been sending English emails using smtp settings for quite a long time with no problem. They are spf/dkim signed with dmarc implementation as well and have always tested great as not spam. I have now added Greek alternatives and I am facing this issue. All Greek emails go to spam. The spam testers return the following error.
"MPART_ALT_DIFF_COUNT HTML and text parts are different Make sure your TEXT version looks similar as the text inserted in your HTML version"
The html text I have is identical to the plain text: https://pasteboard.co/JtUD15C.png
I have made plenty of tests and the error only starts showing after the 250 character (around 250, I have not counted exact number). If the text is shorter than this, the email is spam free with a 10/10 as usual.
Here are my suspicions and what I have tried to counter this. The html source when I manually type Greeks letter from my keyboard turns into this: https://pasteboard.co/JtUE18M.png
Is this why spam filters think the plain text is different from the html text? I tried adding encoding <meta charset="UTF-8">
in the html text with no success.
Every time the 250 character "limit" is crossed, all Greek emails end up in the spam folder.
What else can try to avoid MPART_ALT_DIFF_COUNT HTML
There is supposedly a CHARSET_FARAWAY rule that you can utilize to include foreign language inclusion in the spam filter.
Along with a CHARSET_FARAWAY_HEADER that tells that even the header of the message is written in a foreign language. I guess this could help.