5 ms·
And for what it's worth, what I've set up lately includes SPF, DKIM, and DMARC.
by wpietri 12y ago
And for what it's worth, what I've set up lately includes SPF, DKIM, and DMARC.
- kazinator 12y agoSPF only protects your domain from being spoofed by others, and DMARC helps you get reports about who might be doing that. You have a different problem here in that you are somehow being identified as spammy by this big organization's filters. The only way to combat that is to discover why. It could simply be that you are co-located with spammers: your outgoing SMTP is emanating from a spammy IP block, and you are collateral damage. One important consideration for "deliverability" is who you choose for passing on your SMTP.
- wpietri 12y agoReliable sources inform me that some big providers are using SPF, DKIM, and possibly DMARC as a sign of non-spammyness. I would love to discover any other reasons why my mail is getting filed with spam. For what it's worth, searching various blocklists for my and neighboring IPs doesn't turn up anything.
- kazinator 12y agoAh, your message contains a clue that you're using your own (hopefully) static IP to send SMTP mail, rather than using a forwarding host from your ISP. Speaking of looking up in databases, you can check multiple block lists in one step for your IP using: http://www.anti-abuse.org/multi-rbl-check/ http://www.anti-abuse.org/multi-rbl-check/ If that's not the issue, there could be other problems. For instance, does your SMTP-sending IP address have consistent and reliable reverse DNS? Does the reverse DNS match the forward DNS? Some SMTP receivers perform DNS checks. You need clean forward and reverse DNS for reliable mailing. As a minor point, probably it is best if, in addition to clean DNS, the HELO string used by your server is a fully qualified domain name identical to a reverse lookup. You might want to have a "plan B": that is, have some SMTP forwarding host that you can fall back on if things go wrong with your direct mailing.
- kazinator 12y agoIf so, those big providers are wrong. SPF, and those other mechanisms, are about authentication; but spam is not defined as e-mail which is not authentic, but rather as unsolicited, bulk mail. That is to say, any scoundrel with a domain can publish an SPF record. For example owner of "cheappharmaonline.com" can create an SPF record which says that the SMTP servers of "dirty-spammers.com" are allowed to send messages with "cheappharmaonline.com" as the SMTP sender. So when a machine from "dirty-spammers.com" connects to to receivers and sends "MAIL From: whatever@cheappharmaonline.com", that connection will cheerfully pass SPF checks. What the check indicates is that the mail is not a forgery: yes, it really is a genuine spam from the owner of cheappharmaonline.com, and not a competitor forging that domain. Receivers must implement SPF checks in addition to existing anti-spam measures, not instead of them. Anyone who thinks that messages which pass SPF are "non spammy" will eventually learn otherwise.
- kazinator 12y agoSPF could backfire if it is incorrectly configured. If your SPF record authorizes domain X to deliver mail on your behalf, but the actual path is so that it's coming from domain Y, then your genuine mails will fail SPF checks. When I rolled out SPF, I checked that things are working by sending mails to an account with a mail provider which checks SPF.
- kazinator 12y agoAlso, I would create an account with "major email provider" and send test mails to that account. Then go into that account's spam folder and see if any headers have been added to these mails that might hold clues to why they were classified that way.