Why do I have so much SPAM? And where did my email go?

If you haven't found yourself sifting through your inbox, wondering why you have so many Stupid Pointless and Annoying Messages (SPAM really stands for Simultaneously Published Advertising Message), then you are already making use of SPAM filtering in one form or another.

The truth of the matter is that 92.3% of all emails worldwide sent during the first quarter of 2008 were SPAM.  That's a ratio of about 12 SPAM messages to ever 1 legitimate message.  For home users this can be overwhelming, but for most businesses, this is unacceptable.  SPAM causes a loss in productivity and efficiency.

Over the past two weeks, I have received over 200 emails, putting the amount of SPAM emails I should have received at over 2,400.  In actuality, I have seen 3 SPAM emails over the past 2 weeks.  The reason for this is due to the above mentioned SPAM filtering.

The problem with emails going from point A to point B is that they travel across the internet to get to their destination.  And just because they are telling B that they came from A, does this really mean they did?  Just like you probably carry a passport when visiting different countries, an email will carry some documentation to deliver upon arrival, informing you of its origins.  This is all well and good when the documentation is official, but can it be faked?  For as annoying as the people behind SPAM are, they are often smart (and/or well funded) individuals and are constantly figuring out new techniques to sneak their way into your inbox.  Messages can carry 'fake passports' telling you that they originate from 'this' location, when they really originate from 'that' location.

Your Defense
SPAM filtering.  However, like any remedy, SPAM filtering has side effects.  Most filtering operates by grading your message and deciding if the message passes the test.  It does this by looking for patterns in the messages or for known strings of text such as website links or telephone numbers.  The filtering software generally assigns the message a score, and when it has finished grading the message, decides whether to send it on or not based on its score.  Some filtering software might penalize a message based on it's originating county, or if the words "cheap" and "Viagra" appear in the same sentence; a message might receive an automatic fail if it contains a phone number which is known to link to the guy who sells real Rolex watches for $10 (and if you buy 3, he will also give you the phone number to his brother, who happens to be the prince of Nigeria and wants to give you about $40 million).  Depending on the size of a message and the processing power of the computer that is grading it, grading will take less than one tenth of a second.  (Or more simply yet, more than 10 messages can be graded every second)

The Side Effects
These are called 'false positives'.  This is when the filterer marks a legitimate message as SPAM.  Unfortunately, there is nothing you can really do about it the first time around.  If you requested a SPAM mailbox be set up when configuring your service, then the message is recoverable, otherwise the sender would have gotten a message back saying something like "This message appears to be SPAM and was therefore unable to be delivered."  Now, just like a message can receive an automatic fail because of who it's from or what it says, a message can also receive an automatic pass just because of who it's from.  The system knows what messages to give an automatic pass to based upon if the sender appears on the whitelist.

SPAM isn't going anywhere any time soon.  The problem is that SPAM works and the people behind it make lots of money doing it.  23,300 new SPAM related websites are launched every day (an average of one site every few seconds).  The following is a breakdown of countries contributions to SPAM from the first quarter of 2008:

(Source: Sophos)

Posted April 21, 2008 | Read all posts by Allan Bruford