One of the services we provide to our clients is spam filtering. The goal is to stop the spam BEFORE it gets to the mail server so it doesn't wind up on the users' desktop, laptop, iPad, smart phone, etc.... Each month when we do the reporting and roll up the numbers, it's amazing how much volume there is. Overall, around 80% of all email traffic is flagged as spam. For some of our clients, this means blocking out tens of thousands of messages a month. I looked at our own domain yesterday, and in February, over 10,000 messages were blocked or quarantined, including 348 emails containing viruses.
One of the frustrating things is that with all the sophisticated algorithms involved with the spam security filters we put in place, some spam still gets through. Just yesterday we heard from two clients who reported receiving a spam message that to any human reviewing the email, the disposition should be obvious, but to a computer scanning thousands of messages with respect to certain algorithms, a few get through. In both cases, we saw "Breaking News" emails where for one user, the server logs showed that one email got through and seven were blocked in the past week, and for the other users, one got through, and 65 were blocked/quarantined. In this case, we can see that the filters ARE working, but they are not 100%. For a message with carefully crafted language, the initial emails typically get through, and it isn't until the volume of identical messages is detected that the rest get properly dispositioned as spam.
It's annoying for all of us that these spam messages just keep coming to us, but at least with good filtering, the vast majority are stopped.