Wouldn’t a valid DomainKeys indicate the mail
is not spam.
Domain Keys and DKIM try to provide some assurance that the sender is
who they claim to be. That's it -- no assertion is made about the
contents, just the identity of the sender.
Don't despair because this is very important -- it's only after you can
be certain that you can tell which messages are really from
BigSender.com that you can start to build a reputation for them. Once
you know BigSender.com always uses DK/DKIM and they are responsible
email senders, you can think about automatically
quarantining/discarding anything claiming to be from BigSender.com that
fails those checks. And you can do it without wasting a lot of CPU
cycles on complicated spam detection techniques...
Blacklists are shortcuts used to block spam. If a sender/host is listed
on one you (the receiver) trust, you may choose not to accept the
message or to discard it. But not appearing on a blacklist just means
your message can't be pre-judged -- most recievers will just start
examining the incoming message to see if it can be classified as spam
or not based on other factors.
So as a sender, even though you've implemented DK you still have to pay
attention to how your message will be viewed by spam filters. Do your
messages include clickable links to domains other than the sending
address in the From: header? That may increase the likelihood of your
message being scored as spam. Using a lot of "sales-y" words with
exclamation points? You've probably increased your score a little more.
If that score gets too high you could wind up in the spam folder even
if you pass SPF, Sender-ID, DK and DKIM checks every time.
Sorry if this feels remedial, but it seems like there are folks on the
list who will benefit from a reminder of what sender authentication
techniques will and won't do for you. Maybe one of these days I'll post
this and other
tidbits somewhere for reference...