In electrical engineering, the signal to noise ratio is a measure of the real intelligence on a modulated waveform compared to some level of signal background noise. There’s electrical noise (spurious background pulses and pops) to some degree on every signal, some amount of bad that you must accept with the good. It’s not possible to eliminate every bit of noise, but over the years engineers have developed techniques to filter it out as best we can.

Likewise, those of us who use email frequently have come to accept some level of signal to noise in our inboxes. We’ve gotten used to the idea that junk email is a fact of life, and that we can’t eliminate it altogether. The trick is to keep the noise level down so that email is still a useful tool for the real communication signals we want to listen in on. Fortunately, there’s a tool to help me boost my email signal to noise: the Donet Spam Filter.

What is “Junk Email” Anyway?

First, let me define “junk email” in my own personal context. For me, junk email is any email I don’t need or want to read. It might be unsolicited commercial bulk email, a.k.a. spam. It might be a mailing list that I once subscribed to that I cannot, for whatever reason, un-subscribe from. It could be emails I regularly receive from a friend that is a FW: of a FW: of a FW: of a real funny or touching story someone I never heard of sent to another person I never heard of. To me, any item that clutters my inbox, or my mind, without adding value is junk mail.

Reduce Email Noise With Donet’s Filtering Tools

Nearly every day I start my Daily Digital Routine soon after the alarm goes off. Near the top of my to-do list is opening up my email to see what gems have arrived through the night. This task had gotten progressively more onerous over the years as my volume of junk email increased.

Alek blogged about this earlier in his post Manage Your Donet Spam Filter. His post was an excellent walk-through of our filtering tool in general, and I won’t re-hash his points here. Instead, I’ll focus on how I used our spam filtering to reduce my inbox noise by eliminating 1,500 pieces of junk email in the month of June, 2009.

The first thing I did with my filtering setup is register each email address I use with the tool. I receive a great deal of email to david and dmezera, both “@donet.com”, so I registered each email address and duplicated the following settings for each:

1. High filter sensitivity (block more spam).
   With the sensitivity set to “high”, our spam filters will aggressively tag emails as junk more often. You run the risk of “false postives” doing this, but in my experience the benefits of aggressive filtering usually outweighs the downside of missing a few emails. Besides, if you have spam collecting enabled (as described in the next bullet) you can always look for those emails that were incorrectly tagged as junk.
2. Collect spam mail set to “Yes”, delivered to a “Spam” folder in my inbox.
   This effectively quarantines my junk email in a folder I can poke through if I want to take the time to look for messages that shouldn’t have gotten blocked.
3. Set “Tag as spam in message subject” to “*** SPAM ***”.
   This setting re-writes the subject line and prefixes it with “*** SPAM ***”. This really isn’t necessary in my case, but could be useful for someone who was letting their Outlook mail client do the quarantining based on text found in the subject line.

The second thing I did to tune my email filter was to patiently examine the junk email “From:” addresses as these messages flowed in.  I say “patiently”, because my first inclination is to simply delete junk email en masse and move on to my real work for the day.  Trust me, the payoff is worth it if you take the time to crack open each junk email and use it to build patterns to block similar unwanted emails in the future.

For example, look at the “From:” in the email message shown below.  I’m not sure how I ended up with this particular email, but at the moment I have no desire to get or read emails like this:

I look for two things in each piece of junk email: 1) the “From:” address, and 2) the “To:” address.  Both pieces of information are critical.  In this email above, the “From:” address is “noreply@thefederalmarketplace.com”, and the “To:” address is my “david” email address (@donet.com).  Until I decide otherwise, I don’t want any email from anyone at “thefederalmarketplace.com” — so I created two pattern rules:

*@thefederalmarketplace.com
*.thefederalmarketplace.com

The “*” is a wildcard pattern matching character, and will match against zero or many characters in that position in the pattern string. So, the first pattern blocks any email @thefederalmarketplace.com, regardless of what’s to the left of the “@” symbol.  The second blocks any email coming from any host in the federalmarketplace.com domain name.  The second rule is useful for blocking junk email sent out with a slightly modified “From:” addresses like “noreply@mail.thefederalmarketplace.com” — where the “*” in the pattern matches against the characters “noreply@mail”.

As another example, consider the email below:

Before I tuned my email filter, I got a lot of these “Better than BOGO” emails.  I mean, A LOT.  Here I decided to go with a more aggressive junk email pattern to match against:

noreply@*.info

I don’t know anyone with an address in the “.info” top level domain, and even if I did it’s unlikely that their email address would have “noreply” in front of the “@” symbol, so I think this pattern is safe to use.

There’s one additional thing to note here: the “To:” address on this email is “dmezera”, not “david”.   If I were to accidentally put the “noreply@*.info” filtering rule in my settings for “david”, it would have no impact on email arriving for “dmezera” and I’d continue to receive these junk emails.  The way I handle this situation is to use my “david” account as the settings master, and add all my rules there — even if the rules I’m adding don’t specifically apply to that email address.   After all my rules have been added to “david”, I block copy the rules and paste them into each of my major secondary email addresses.  The end result is that my main email addresses all operate with the same settings — a Good Thing.

Finally, I added some rules to immediately pitch email that comes with a “From:” address in a domain I simply never expect to receive legitimate email from.  These are, generally (but not always), non-English speaking countries that I currently have no personal or business relationship with.

*.cn
*.hk
*.es
*.co.uk

In effect, the above rules say that I want to block any email that I get with a “From:” address indicating that it originated in China, Hong Kong, Spain, or the UK.  I’m sure this list will grow over time to include most other foreign countries.

Impact

By adding filtering rules in this manner I have been able to block about 1,800 pieces of junk email in the month of June 2009 — emails that I would have otherwise had to sift through and delete.  The benefit of the reduced workload is actually quite huge, and I’ve found my email experience to be more productive than it has been in a long time.  Prior to the addition of my filtering rules the number of pieces in my inbox would jump by 20 or 30 emails during the night, and then by roughly the same amount during each hour in a typical business day.  This morning I had less than five emails in my inbox when I woke up, and none of them met my previously established “junk” criteria.

Safety Tip

Don’t ever use the following as a filtering rule, unless you really, really want to block everything (and I mean, *everything*) from getting into your inbox:

*

The “*” wildcard will pattern match on every string of characters, no matter the length.  This will result in a eerily silent inbox.

Remember, if you set “Collect my spam” to “Yes”, every piece of email that gets blocked will be quarantined in the “Spam” folder for a few days.  Take a look in that folder every now and then (using webmail or your favorite email client), and make sure your rules are blocking the right kinds of messages.  Good luck!