1 FEB 2011 1
This is the third and final part of our series on how email spam filters work. If you missed the first two posts, you should read them here and here and then get back to this one.
Today I will brief you on how community-based filters work and domain/IP authentication and filtering.
These spam filters are a great application of the crowdsourcing concept. Rather than building up a set of rules on what spam looks like and then updating and tweaking them as spammers get smarter, major webmail providers have implemented the so-called community based filters. Whenever a user receives an email and clicks on the ‘Mark as spam’ button, a complaint is filed on the provider’s servers. Once a lot of complaints have been registered for a certain sender, that address is blocked.
If you scale up your email campaigns to millions of recipients, you will run into this issue, no matter how ‘white-hat’ and legitimate your messages are. There will always be some black sheep in the crowd who would rather click the ‘Spam’ button than go through an unsubscribe process or people who simply forgot they have a running subscription with your service.
As discussed in December last year, this could be a valid reason to outsource your email campaigns to some specialized service provider. If you insist on taking matters in your own hands, you should hire someone to communicate with all major ISPs on a constant basis and have your IPs white-listed as soon as spam complaints kick in.
Some major spam filters will synchronize with ‘Blacklists’, services that keep an eye on IPs that ‘misbehave.’ If your IP address is in such a list, ISP’s are likely not to allow your email to pass through.
Such practices could affect you even if you don’t send out spam. If a blacklist service decides to list an entire IP range and you are on it, then your emails won’t get through either. Fortunately, this can be taken care of—most of these services will de-list you upon request and only list you back if more complaints occur.
Major and trusted ESP’s have teamed up with spam filters in an attempt to check whether an email coming from a domain really originated from there. While there is no unitary authentication protocol, Yahoo and Microsoft have created their own authentication protocols—DomainKeys and Sender Policy Framework.
Don’t be paranoid. As long as you play by the book, don’t spam your list and follow the basic guidelines outlined in this three-part article, your emails shouldn’t be flagged. However, if a majority of your mails don’t get to the inbox, check the blacklisting service(s) and/or the ISPs to see if that's the problem.