What's the point in spending hours preparing a newsletter, message or report if it's automatically filtered into the junk folder before the recipient even sees it?
Spam threatens to choke the communication channels promising global freedom of expression.
Internet Service Providers (ISPs), corporate server administrators and end users are increasingly using new anti-spam technology to try to stem the relentless tide of junk email flooding the Net.
The problem is: how can we prevent the dolphins from being caught along with the sharks?
The origin of 'spam'
SPAM is a pink canned luncheon meat immortalised in Monty Python's spam-loving Vikings sketch.
In an Internet context, lowercase spam refers to unsolicited commercial or bulk email (such as get-rich-quick schemes, miracle cures, weight loss, Viagra, lotteries, loans, pornography and Nigerian sob stories) and allegedly originated in a MUD/MUSH community.
Of more practical use is the origin of the actual spam mail itself.
Where does all the junk come from?
In the mid-90s, Usenet newsgroups (also called "discussion groups" or "bulletin boards") were the number one source of email addresses for spammers.
Today, the most common origin is web pages, especially if they're listed in a search engine or directory.
Some people have tried foiling address-seeking spambots by inserting the word "UNSPAM" in capitals in the middle of all email addresses on their sites. This stops auto spammers working but enables human beings to work out what to do.
Spammers also harvest addresses from headers of messages you send to friends who forward them to their friends (a good reason for using BCC -- blind carbon copy rather than simple CC which displays all recipients - although some people filter out mail sent using BCC as many spammers also use it).
Other sources include open e-mail discussion lists and web pages that invite you to "insert your address here to be on a 'do not mail' list".
Spammers can simply guess addresses by generating lists of popular names and random words attached to common domains (bob@aol.com, john@hotmail.com).
Once on a spam list, the only way to get off is to change addresses.
If you reply or respond to instructions to "remove", your message will simply confirm your address is valid and you'll get even more junk.
Depending on your email client, you can try tracing junk back to its owner by contacting the server listed in the full message header information (the From address is generally fake - check your Help files to find out how to "reveal full headers").
How to stop spam
Despite legislation against unsolicited commercial email, the volume of junk is increasing alarmingly.
The simplistic oft-cited fix -- "just hit delete" -- is only a bandaid solution and fails to discourage the junk merchants.
Self-regulation and industry codes are difficult to enforce. ISPs face problems if they disconnect service to spammers under some countries' telecommunications laws.
Technical solutions have centred on filtering technology.
Types of filters
Many corporations and ISPs filter incoming mail on or after delivery.
Server-side filtering software typically looks at the headers, subject line and/or contents of the message.
Some filters -- and their users -- are smarter than others.
SpamAssassin is an open-source, collaborative, community anti-spam effort based on filtering rules to analyse email content.
The software gives each message a score based on how many rules it breaks.
Any programmer can suggest rules for new releases of the software which spots, not blocks, spam.
ISPs and server administrators then decide whether to send suspect mail to junk folders, automatically delete mail tagged as spam, or bounce it back to sender.
Unfortunately for email publishers, some of the filter rules are too broad or the threshold is set too low.
Many innocent messages are being lumped in with the guilty.
One of my newsletter readers notified me that his ISP had tagged a recent issue as spam -
SPAM: -------------------- Start SpamAssassin results ----------------------
SPAM: This mail is probably spam. The original message has been altered
SPAM: so you can recognise or block similar unwanted mail in future.
SPAM: See http://spamassassin.org/tag/ for more details.
SPAM:
SPAM: Content analysis details: (13.2 hits, 6 required)
SPAM: Hit! (1.5 points) BODY: Asks you to click below
SPAM: Hit! (0.2 points) BODY: No such thing as a free lunch (1)
SPAM: Hit! (2.6 points) BODY: Instant Access button
SPAM: Hit! (3.5 points) URI: URL of page called "unsubscribe"
SPAM: Hit! (4.1 points) URI: 'remove' URL contains an email address
SPAM: Hit! (2.1 points) BODY: FONT Size +2 and up or 3 and up
SPAM: Hit! (0.0 points) BODY: Includes a URL link to send an email
SPAM: Hit! (-0.8 points) BODY: Image tag with an ID code to identify you
SPAM:
SPAM: -------------------- End of SpamAssassin results ---------------------
Even if you can identify the problems that raised your score, you can do little about it.
Best practice in e-newsletter publishing -- and now law in many countries -- is to enable your readers to unsubscribe easily yet the word "unsubscribe" or similar would lead to your newsletter being blocked by filters such as SpamAssassin.
Just look at the seemingly innocuous words on the constantly changing list of tests SpamAssassin performs on mail messages to determine whether they're spam.
A growing number of email users also filter their own inboxes by setting up rules to delete messages containing certain keywords or punctuation (for instance, the phrase, "This is not spam", which is among the most commonly used phrase in junk email).
Some desktop applications, such as MailShield and MailWasher, pre-set the rules and tag or delete mail without the user having to do anything.
Some client-side filters prioritise mail into different folders.
To ensure they receive wanted mail, email users can add favourite mailing lists to a "white" or "green" list so the mail ends up in a specific folder.
The problem with email filtering
Arbitrary and unpredictable email filtering, as with web content filtering, can backfire and cost time, money and frustration -- the very things filters were designed to save.
Tracking down problems and dealing with aggravated users or administrators wastes resources that could be spent creating content, marketing and improving customer service.
Many filters can be imposed only after you have received unsolicited email from each source, meaning you have to respond manually to every single spammer.
Spammers regularly change their email addresses to bypass filters.
"Qualitative" filters can try to detect junk mail from unknown sources but legitimate mail can be inadvertently trashed in the process ("false positives") while some junk mail still gets through ("false negatives").
Filters are becoming a serious problem with many legitimate opt-in publishers being mistaken for spammers and blacklisted.
Once on a black list, the only way to be unblocked is to contact the ISP and ask to be placed on a white list.
While some e-newsletter publishing tools let you monitor how many messages are opened or bounced, you might never know which readers are being filtered unless you know which email program they're using.
Even then, the recipient might just be busy, uninterested, on leave or not receiving your message.
Some filter programs don't even notify you that your message was blocked while others bounce back mail with "sensitive" content without specifying the rejected words.
On another occasion my own free weekly e-newsletter was blocked on its way to at least two subscribers thanks to Trend SMEX Content Filter detecting "sensitive content" yet the "naughtiest" word I could find was "teenage".
Pity the poor medical publishers who often use "controversial" words such as "breast" or "penis" in their newsletters.
You might be able to track down the reason you were blocked by examining the full header information (ask your subscribers to forward the entire copy of your junked message).
The header should list filter rules that the message violated, if the ISP uses software such as SpamAssassin.
If not filtering, then what?
Some online publishers advocate web services where you submit your newsletters for screening of content or technical problems that might trigger spam filters before mailing.
Of course, spammers could also abuse such a service to get their own mail past the filters.
Others back the idea of an Internet-wide white list along the lines of web site privacy site TRUSTe.
Spam-busting campaigners believe the only effective method of addressing the problem is through simple, direct legislation to:
Meanwhile, ashile the world waits for more countries to introduce tough anti-spam legislation, online publishers must continue to be vigilant.
How to avoid the junk folder
Choose the subject line and words in your emails carefully if you don't want to end up in the junk file.
More people are setting up their email programs to filter incoming mail automatically for classic junk mail keywords and symbols (eg "advertisement", "free", "rich", "porn", "lover,", "opt-in," "e-mail", "money-back guarantee", "teen", "sex", ! # or $, removal instructions, all capitals).
Drop all hype and advertising-speak from your subject lines and messages.
Apart from saving you from the filters, honest concise text is preferred by the impatient, overloaded online reader.
Online publishers wishing to avoid being trashed should send a welcome message to new subscribers or correspondents telling them how your messages will appear (eg "the newsletter will appear from yvette@brizcomm.com.au") and suggesting they add the address to the white or green list of their email program.
Set up test accounts with popular email providers such as AOL, Yahoo and HotMail to see if your messages are getting through.
Beware Microsoft Outlook's "junk e-mail filters" which, when turned on, colour or delete messages considered spam based on a filters.txt file.
I searched for the filters.txt file on my PC to see which words were no-gos and was surprised to see so many seemingly harmless words.
Use a professional list hosting service unless you can afford an experienced in-house support staff.
E-newsletter publishing is becoming a highly specialised and complex field - don't risk your reputation to amateurs.
Ensure your list host monitors other clients to ensure they're publishing to only opt-in recipients -- if opt-out mailers, rental lists and spammers are blacklisted, you might be guilty by association.
Insist on detailed bounce reports for every issue you send out.
You should be able to see which names bounced and the exact reasons (eg "mailbox full", "address does not exist" or the frustrating "unknown").
Just because your bounce report says your mail was delivered, doesn't mean it did.
Many filters stop email after the "not-bounced" signal has been sent.
Monitor your open and clickthrough reports for unusual patterns.
Read all reader feedback, particularly if readers are having trouble subscribing or getting your mail.
Chances are they might have to ask (or scream at) their ISP or techies to add you to a white list.
(Of course, you need to publish quality content if you want your readers to jump up and down when they don't get it.)
Many permission-based publishers include their newsletter or trademark in square brackets in the subject line to avoid being filtered.
Other ethical publishers, including me, have resorted to disguising known filter trigger words by inserting unusual characters or deliberate typos -- for instance, using f^ee instead of free or sp*m instead of spam.
It's just a matter of time before the spammers work around these "tricks".
Bastardising or censoring our language to suit new communications technology would seem an extreme solution -- not to mention bitterly ironic -- but what's the alternative?
Useful links
Coalition Against Unsolicited Commercial Email
http://www.cauce.org/
The Net Abuse FAQ
http://www.cybernothing.org/faqs/net-abuse-faq.html#2.1
How to show full header details for your email program -
http://www.uia.net/emailabuse.htm
Put junk mail in its place (tips on using Outlook) -
http://www.microsoft.com/Office/using/column01.asp
Tips on writing hype-free online content
http://www.brizcomm.com.au/
Yvette pioneered web writing training in Australia in 1998 and has inspired thousands of people through her workshops and presentations from London to LA, Sydney to Seoul.
She has reviewed web sites professionally for print, online and radio for since 1997. Her own e-newsletter and site - www.brizcomm.com.au - won the Smart Communities category in the 2002 Asia Pacific ICT Awards.