Lost: 200 emails gone missing. Last seen on the 25.10.2003
Sound familiar? Ever wondered what happens to those unopened emails that you see in your stats from your latest email campaign/newsletter? Yep, you guessed it-their mail servers are doing it for them.
This is what is referred to as a "False Positive". It means that although the recipient has opted in to receive your newsletter, for some reason or another, the mail server's filters have deemed it to be Spam and, in most cases deleted it, before the recipient has had the chance to give their verdict on whether it was worth opening or not.
In order to protect themselves (and presumably us) against the onslaught of unwanted emails, the servers have implemented filters, which are designed to stop Spam in its tracks. Of course there are bound to be some casualties - and in this case it is your legitimate correspondence.
There are four main things that happen when an email is filtered - depending upon which type of filter is being used.
So, let's have a look at the types of filters currently in use:
Community: This is based on the community joining together and installing the filters, then reporting any occurrences of Spam. False Positive occurrences should be low, but in reality they aren't. This is due to a couple of reasons: The first being the weed theory. What one person regards as a weed, another may regard as a flower. We all have different opinions. The other is that many people are now reporting once wanted email as Spam instead of unsubscribing.
Blacklisting: ISP's generally use their own proprietary lists of known Spammers, created by themselves. Third party blacklisting companies also publish databases of Spam senders. False Positive's are extremely high, as these third party blacklists have no accountability. In fact anyone can report an IP Address as a Spammer and do not have to qualify their reasons.
Address recognition: Sender's email address is in the recipient's address book or the server's/gateway's acceptable list. Email senders must earn their position within these lists.
Trapping: The filter company plants email addresses all through the Internet to attract Spam Harvesters. Once the false addresses receive Spam, the source of the Spammer is identified; the sender is then blocked immediately across all users of the filter. False positives are extremely low.
Challenge-Response: This white listing concept is based on all emails being Spam and the sender having to prove otherwise. Before an email is delivered, the ISP or PC that uses the filter challenges that a human sent the email. The sender has to then manually click the link and then they are added to the subscriber's permissible list. Any auto-responses such as Double Opt in responses do not work well with this type of filter.
Rule-Based: This is the most commonly used type of filter on a server. It catches Spam by rejecting/accepting messages based on predetermined rules of what is acceptable.
Bonded Sender: Email markers gain status as a non-Spammer through these bonder senders and are therefore permitted through the filters.
Volume Blocking: Spammers are known to send high volume emails without thought to speed, volume or list cleanliness. The ISP sees the large amount as well as the other problems and terminates the connection.
Probability: Using rules, a method "learns" the users definitions of Spam and decides what may be regarded as Spam. False Positives increase as rule-based Spam is reduced.
Remember though, not only are you dealing with rule-based filters as well as the other types of filters (and most servers use a combination of filters), you also need to get past the personalised filters (i.e. the recipient themselves), so be sure to test your campaign thoroughly before sending it.