Spam Filters

How Spam Filters Operate

Spam filters look at a long list of  criteria to decide whether or not your email is junk/spam. The list of  content that is considered spam criteria is constantly being adapted.  Spam filters learn more about what junk content looks like every time  someone clicks ”this is junk/report as spam”. Some spam filters even  sync with each other to share what they’ve discovered. There is no  universal formula, but there are a few guidelines to help you avoid  common design mistakes that often send email marketing directly to junk  folders.

The Spam Filtering Process

When you send  an email campaign to your subscribers, your messages have to get through  their ISP’s spam filters, then their email application’s spam filters.  It's surprisingly easy for an innocent and legitimate email to be  mistaken as spam (this occurrence is called a "false-positive"). It  therefore helps a great deal to understand how a spam filter works. The  following is a brief overview of what to look out for when sending your  campaign to your subscriber list.

The Subject Line

The  spam filter is not the only hurdle you’ll encounter, your subject line  is how your recipients will judge whether or not to open or delete your  message, and so it HAS to be relevant. Try to make your subject line  concise so the recipient knows: who sent it, and what it refers to. Make  it intriguing, but relevant to your readers. Also it should assure them  of where the message came from, and what it contains. For instance, if  we sent a Total Send newsletter with the subject line, "FREE GUIDE  INSIDE! INCREASE OPEN RATES NOW!!!!!" we'd be making a huge mistake.  Something like, "Your guide to improving open rates" would be much  better. Avoid spam words and phrases, like FREE, act now, limited time,  insurance, casino, coupons, click now, open immediately, etc. Also avoid  using expletives and redundant phrases. Don't try to be creative with  numbers/symbols in place of letters, like: V1AGRA, @CT N0W, CH01CE, etc.

Who is the message is addressed "To:"

Checking  your mail seems very impersonal when you receive a letter that is  addressed to, "CURRENT RESIDENT" or "TO WHOM IT MAY CONCERN". It's  obvious that it was generated by a computer.  Likewise, don't just  address your email newsletter to the email address. Merge your  subscribers’ first and last names into the “to:” field using Total  Send’s personalisation links.

The content of message

Spam  filters look for "spam triggers" both on “the surface”, and in the code  of your email's content: Don't just send a single large graphic. This  is considered spammer behaviour. Your HTML email needs a good balance of  graphics and text. If you're sending a simple piece to your recipients,  and all it takes is one image to get your message across, you can still  include some text at the bottom of your newsletter, such as an  "unsubscribe" link, and your physical mailing address, company  disclaimer etc. Always include a plain-text alternative with your HTML  email. Spammers aren’t interested in doing so. It may seem like a lot of  extra effort to write both versions of your email, but it is an  imperative step. Also don’t make the plain-text version too bare. 

Spam  filters compare your plain-text alternative to your HTML email. It’s  considered “spammy” if 90% of the message is in HTML, and 10% is in  plain-text. Spam filters punish lazy email marketers. A powerful and  easy to use tool to do this is Premailer, which can be found at this  link: Code  your HTML email the right way. If your HTML email is coded badly,  you'll look like a spammer. Broken images, missing tags, and  non-web-safe colours are some of the things they look for. Never use  Microsoft Word to generate your code! Word creates horrific HTML code  and it’s almost a guarantee to work against you. Learn how to code HTML  properly. Or pay someone to do it for you. It is worth the investment of  time or money.

Avoid these common mistakes:

•     Using spam-like phrases such as “Click here!” or “Once in a lifetime opportunity!”

•     Using! Lots! Of!  Exclamation marks!!!!!!


•     Colouring fonts bright RED or GREEN

•     Coding sloppy HTML, usually from converting a Word file to HTML, or using Publisher.

•      Creating an HTML email that contains nothing but one large image,  with little or no text (spam filters can’t read images, so they assume  you’re a spammer that’s trying to trick them).

•     Using “Test” in the subject line

•      Sending a test campaign to many recipients within the same company  (that company’s  email firewall assumes it’s a spam attack)

•     Designing HTML email in Microsoft Word and exporting the code to HTML (That is coded badly, and spam filters hate it.)

The Sender's Email Address

Some  spam filters require that the sender of their mail must be  "white-listed" or they are viewed as a stranger. And with being labelled  a stranger, the spam filter will automatically categorise your email as  spam/junk, or it will become unreasonably strict when judging your  email's content. When people subscribe to your newsletter, it’s a good  idea to ask them to "add this email address to your contacts list, to  ensure proper delivery." Placing this request on your subscription  confirmation screen, and all welcome emails will help in the battle  against the spam filters. Some people place this at the top of almost  every newsletter they send. When sending your mail and deciding your  “sender” information, it’s best if you're not using anonymous details,  like a free email account (Yahoo, Hotmail, etc.). Having a  public email domain on your reply-to address will knock your spam score  up further. For the best results - setup a real email address with your  company’s domain name. Then get your users to whitelist it. Also, never  change it. This ensures it will remain widely whitelisted and will  prevent the need to whitelist another one.

The Domain Name

Some  spam filters check to make sure that an email claiming to be  originating from a particular domain name did in fact originate from  that domain name. This is called "authentication" it's slowly becoming a  normal procedure. Emails that are not "authenticated" are either  classified as "junk/spam" or are flagged as "suspicious."

Community-based reporting

Whenever  an email recipient clicks the "this is junk" button for a particular  message they received, that "complaint" is sent to the ISP. If enough of  the ISP’s users report an email from you as spam, then the ISP may  block all your future email to their servers.

Do not let Spam Filters deflate your efforts

Most  of the trusted spam filters out there aren't "black or white" with  their algorithms. They're usually quite "realistic" about emails, and  they use a well-rounded combination of criteria. SpamAssassin is an  excellent example that you can learn from. It assigns "severity points"  for each "rule" broken by messages. For example, using "CLICK HERE!" you  could get 0.5 points added (for each occurrence in the message), while  using bright red fonts might get you 0.1 points added,  and including the word, "V1AGRA" in the subject line might get you 4.0  points added. It will count up the total score, and if it exceeds a  certain amount (which is set by the person who installed it on the  server), the email is labelled as "spam." When you write your email  newsletter, don’t be paranoid about using phrases like "Click here" a  few times in your newsletter. As long as the rest of your newsletter  checks  out, and you don't break too many of the rules mentioned above,  spam filters will let your emails through.