Creative Data Concepts Limited, Inc. SPAM Filtering Policy Explained

Since many of our customers have questions or concerns about how our SPAM filtering system works, we have put together this document to help explain our SPAM filtering process and policies. If you have questions beyond what this document covers, please let us know as you’re probably not the only one with those questions. We can then add to this document as needed for clarification.

One last thing before we get down to the details. We know that you are busy. We’re rather busy, too. This document is meant to take no more than 10-15 minutes to read and assimilate, whereas a phone call to our office would take at least as much time and probably not cover as much ground. We only ask that you print this and read it at your convenience before you decide to give us a call.

Our process for filtering SPAM boils down to this:

1. Our mail server (SMTP) receives an email.
2. It sets up an internal SPAM score for each email, which starts at 0.
3. It examines the FROM address, what mail server that corresponds to, and what other servers the email message has passed through along the way. Several lists are checked to see if the sending servers have been SPAM-friendly in the past, whether spammers can still send SPAM through them, etc. The SPAM score is marked up accordingly for each appearance in one of these DNS blacklists.
4. The text of the message is then scanned for keywords or phrases that would indicate SPAM. The SPAM score is marked up accordingly.
5. The FROM address is traced to see if that domain has an anti-SPAM policy, and has an ABUSE account to which SPAM complaints can be sent. If not, the SPAM score is marked up accordingly.
6. Several more esoteric tests are performed that may mark the score up or down.
7. If the score is over a certain amount, the message is either sent to the SPAM folder or deleted depending upon the total SPAM score.

The nice thing about this system is that an email may fail a test or two but not get marked as SPAM, depending on the severity of the problem. The other nice thing is that as each test fails, notes are added to the email headers to tell you what the tests the message failed. If you are using our WebMail system, there is a link at the top of each email page that says “Headers”.

In Outlook and Outlook Express, you open the mail and go to the “View” menu and choose “Options…” and at the bottom of the new window there is a box labeled “Internet Headers”. You’ll be looking for headers like “X-SPAM-Tests-Failed”, where there will be a list of the tests that failed.

For example, take the following:

X-SPAM-Tests-Failed: SORBS-SPAM, MXRATE-ALLOW

The problem here is that the mail server the sender uses is on a SPAM list, “SORBS-SPAM”. You can also tell this by finding the sender header (“X-Declude-Sender: user@company.com [192.168.51.78]”), taking that IP address at the end (“192.168.51.78”), and looking it up to see what lists it is on at:

http://www.dnsstuff.com/

That particular list, SORBS-SPAM, is stating that this particular sender’s mail server has sent out SPAM in the past, and probably still does, so any email from that sender would probably also get marked as SPAM. Remember, there’s no hard line saying “this is SPAM” or “this is not SPAM”, the filter just makes its best educated guess.

In this case, the sender needs to have their mail administrators get themselves off the SPAM list(s) so that their mail will stop getting marked as SPAM. As time goes on and more companies adopt more stringent SPAM controls, they’ll be able to send to fewer and fewer people without getting marked as SPAM, so it’s in their best interest to get on their case now.

There are some times when the sender is not going to have the option of getting on their mail administrator’s case about fixing their system. For example, Hotmail has been spammer heaven for quite some time and they don’t have an account that handles SPAM complaints. Getting Microsoft to fix their servers and stem the tide of SPAM coming from their users is probably not going to happen any time soon. But, were we to make a blanket statement and say “all mail from Hotmail can get through”, then all of our customers who are paying us to not have to worry about SPAM would not be getting what they paid for. In this scenario, you may want to recommend to the sender that they sign up for another free email service (such as Google’s Gmail) that does not act as a haven for spammers, and use them instead. We strongly urge businesses owners to use purchase a domain name and secure email service for that domain from a reputable provider who takes a hard line stance against SPAM.

You are also welcome to have the sender call us and we can give them pricing on hosting their email services with us, thus eliminating their SPAM problems to all of their clients.

So, long story short (too late!), there’s not much that we can do here to make sure that mail servers managed by other providers are configured correctly and are not on SPAM lists. The SPAM lists aren’t ours, nor do we have any control over who is on them, nor do we want to ignore them and start letting through mountains of SPAM. It is not our policy to bend the rules, such as to allow all email from Hotmail or ComCast or whomever, as we see that as defeating the purpose of providing a SPAM filtering service in the first place. We just try to spur the victims, and in this case we mean the people who are trying to send email to you, to inform their ISPs and encourage them to fix the configuration issues on their mail servers.

As a last resort, you can configure your Outlook (or other email program) to download email from your SPAM box. This way you can see all your mail, even SPAM, without having to go through our WebMail interface. Let’s say your email address is bob@bobswidgets.com. You would just need to add a new account to Outlook that also checks bob-SPAM@bobswidgets.com, and you would download everything in your SPAM folder every time you checked your mail. (The password for your SPAM folder is the same as the password for your primary account.) Of course, then you’d have to deal with all of the SPAM on your own. Should you then see that a message was erroneously marked as SPAM, and if you were feeling particularly feisty, you could send the original sender a quick note telling them to light a fire under their ISP.