SPAMfighter performance review

Last year I started to use SPAMfighter to get rid of the growing amount of unwanted mails (read more). After an update that went wrong all the stats had been deleted, and I started from scratch. Now, almost five months later, I would like to share my experience with the service with you.

SPAMfighter statistics

The first thing that springs to mind is the fact that I have “just” about 60% spam. I have seen reports that about 80% of all emails on the Internet are spam. Not so for me (which is nice)!

Not so nice is the fact that SPAMfighter catches only 57% of the spam that hits me (i.e. 35% of all mails), and that I still have to see and block the remaining 43%. Which surprises me a bit. After all, SPAMfighter could increase their member base from 3.6 million in July 2007 to almost 4.7 million in March 2008. This bigger community should be able to catch spam better.

I guess most of the community members are consumers who do not bother to report incoming spam. One of the main problems is that there is no incentive for anyone to actively block spam, except for the somewhat cloudy, intangible benefits for the whole community. There is no “top SPAMfighter” award, no top lists on their web site, no feedback loop, no “Thank You!” mails. So, members might wonder why they should block spam at all? Where is the benefit for them? The service seems to be able to catch spam even without their help, so why bother?

Also, I am quite surprised that I still receive a lot of mails that are obvious spam (e.g. carrying certain adult keywords in the subject line). This spam could and should be easily detected by the systems, but apparently goes through undetected. I just wonder why? I would strongly suggest to SPAMfighter to also introduce keyword based filters that can be populated by the user individually, or -if the user is too lazy- by optionally using default lists provided by the community. The knowledge on the keywords has to be somewhere in the SPAMfighter systems; it’s just a matter of applying this knowledge to create more powerful filters.

Comments are closed.