spamtraps

This is the approach SpamTrap uses. There are several flavours of implementation and it's worthwhile noting the differences (and, you won't be surprised to hear us argue why we believe that SpamTrap's technology is superior), but first to describe in general how Spam Traps work...

Spam messages are not sent in their hundreds or thousands, but in their hundreds of thousands and often millions. They are sent to lists containing vast numbers of addressees which Spammers build using a variety of tools. Spammers also exchange lists amongst themselves – which explains why you might initially receive the odd spam message from what is obviously the same source and then suddenly find yourself bombarded by spam messages from a wide variety of sources.

This helpful behaviour of spammers also exposes a weakness of theirs to those, like us, who would prefer not to receive their unsolicited messages: the fact that the same message is sent to a very large number of recipients. This enables organisations like SpamTrap to seed spammers' lists with email addresses that aren't 'real' in the sense that they don't belong to real people and are not used for genuine email messaging. How we seed effectively is a trade secret, but the important thing here is that any email messages that arrive in one of our Spam Traps must be either a mistake (a genuine message that someone has misaddressed) or spam. If the same message arrives in more than one Spam Trap, it isn't a mistake: it is spam.

These identified spam messages can then be automatically rejected should they arrive at the mailbox of anyone using a service such as SpamTrap. Of course, spammers attempt to circumvent such an approach by slightly altering each transmitted spam message so that messages arriving at Spam Traps are never identical to those that arrive at genuine mailboxes (or at other Spam Traps). Changes inserted by spammers can include basic alterations like including a salutation ("Dear John...") tailored for each recipient to more sophisticated approaches involving inserting long meaningless strings of random characters at the end of the spam mail message. But the technology that SpamTrap uses looks past these attempts to deceive and generates a 'fuzzy fingerprint' for each identified spam message. A fingerprint is then generated for each email message in our subscribers' mailboxes and, if the fingerprints match, a spam message is detected.