The IDC correctly predicted that over 40 billion of the nearly 97 billion emails sent worldwide in 2007 were spam. 2008 is expected to be the first year that spam email volumes are expected to exceed person-to-person email volumes sent worldwide.
“But why?” I hear you ask. Well, it's the same reason that South Africans have to deal with hawkers selling cellphone chargers, sunglasses and chickens made out of brightly coloured plastic bags at every traffic light - enough people buy to make it worthwhile. One survey claims that 11% of all Internet users have made a purchase through unsolicited email. Essentially, unless our consumer behaviour changes, spam is here to stay.
The good news is that the amount of spam has plateaued. The bad news is that its plateaued at about 85 to 95% of all email sent and received. Backbone providers are under pressure from this onslaught. Spam represents a significant cost to both them and the Internet community as a whole. Its costs companies real money and people real time.
Quite apart from the straight bandwidth costs associated with spam, there are a number of other financial risks ranging from wasted storage space to declining productivity, particularly among IT staff.
Lots of spam means lots of spam management. For most administrators, the headache is not just trying to stem the tide of spam, but making sure the technologies and policies they have put in place don't impede their ability to manage the ever-growing list of new IT issues such as email retention requirements, archiving and discovery criteria, all elements of the big hairy beast that is known as compliance.
The IT industry and spammers have been locked in an arms race for three decades. We still don't seem to have any idea of who is winning or if this struggle can be won at all.
One of the issues with anti-spam technology is that it is largely performed at the main ingress point, where the message is about to enter the corporate network - by which time the cost of carriage has already been billed. Another issue has been the almost ubiquitous use of content scoring mechanisms to identify spam which spits out lots of false positives and fills quarantine folders to the brim with legitimate mail. All of these place a higher burden of administration and cost on organisations.
So, what is one to do? Change. Adapt the attitude to spam. It's not about the content anymore. While Bayesian scoring worked well initially, spammers have learned to trick the system. Spam filtering is not about where the message originates.
The only way to fight a technology war is with technology. Until users wise up to the ever-increasing cunning of the spammers, companies will need to rely on software that can do the thinking for them.
