Best Practices: Avoid Sending Spam | Predictive Response
Best Practices: Avoid Sending Spam
The process of determining SPAM in lists is an evolving process. You could consider it an “arms race” in a way. As such, as soon as someone finishes their research and publishes their findings – the information is almost always out-of-date!
This module has been written to give you a brief overview of the basic principles so that you can avoid getting labeled as a SPAMMER.
Bayesian Filtering
As an email recipient, you likely have many emails that go into your spam or junk mail folder without you taking action. This may be due to you indicating in your email settings ,or as emails are received, the types of emails that you consider spam. When you open an email – or even before you open the email, most email providers include an option to add the email to your spam or junk mail folder. In this manner your email learns which types of emails to automatically move to your spam or junk mail folder. This method is known as Bayesian Filtering.
The filter doesn’t know what to move to these folders automatically, it has to be trained. To train the filter, the user manually indicates whether a new email is spam or not. For all words in each training email, the filter will adjust the probabilities that each word will appear in spam or legitimate email in its database. For instance, Bayesian spam filters will typically have learned a very high spam probability for the words “Viagra” and “refinance”, but a very low spam probability for words seen only in legitimate emails, such as the names of friends and family members.
After training, the word probabilities (also known as likelihood functions) are used to compute the probability that an email with a particular set of words in it belongs to either category. Each word in the email contributes to the email’s spam probability, or only the most interesting words. This contribution is called the posterior probability and is computed using Bayes’ theorem. Then, the email’s spam probability is computed over all words in the email, and if the total exceeds a certain threshold (say 95%), the filter will mark the email as a spam.
What does this mean to me?
If your email content includes words that can be considered as spam triggers, then they will very likely go directly into your campaign members spam or junk folders. Do some research on the Internet using the keywords “Spam Triggers” to find list of words that you may not want to use in your campaign emails.
Link Manipulation
Link manipulation is a form of “Phishing”. Link manipulation can occur when the link includes a mis-spelled URL or incorrect sub-domain name.
For example, this URL – http://www.yourbank.example.com/ – appears to be a link to the example section of the yourbank website. In fact, the person who clicks the link will be taken to the yourbank section on the example website.
Another example shown here involves the text that displays on a web page – or in a campaign email – and the URL that has been entered.
In the example above, the text in the campaign indicates that when the link is clicked, the user will open the “Genuine” article on wikipedia.org. However, instead when the user clicks the link, they will open a completely different article called “Deception”. Compare the URL in the Link Test field to the URL below.
What does this mean to me?
- Make sure that the URLs you enter are accurate. A simple one letter typo can re-direct the user to a page you did not intend to include.
- When creating campaign emails with links, make sure your test members not only review the content, but click all the links to ensure they are working as you intend.
- Avoid using links or images from domains that are not under your control. If the data changes on the outside resource, you may not know that it has been changed and your campaign emails will not work as you intended.
Spam Traps
Sending emails to a spam trap is like putting out a sign with flashing arrows that says, “I’m spam! Block me now!” What is a spam trap? They are email addresses that ISPs and receivers use specifically to shine the light on spammers.
Spam Traps come in two different varieties:
- Email accounts that are not linked to or accessed by humans. Since no actual person uses the account, it follows that they could not go through opt in process for an email delivery program.
- Those email accounts that have been left along the wayside, abandoned, and no longer used. ISPs will use these accounts and will block you if emails are being sent to them. Their justification is if they aren’t active anymore then they should not still be receiving email.
Repercussions of being sent to a spam trap are quick and severe. Most of the time, ending with ISPs totally blocking the sender’s IP address, thus making email deliverability nearly impossible, until you remove these spam traps from your email list.
Which addresses that the ISPs are using as spam traps are not public knowledge, making it nearly unfeasible to recognize and eliminate these specific spam traps from the email list of marketers.
While the former is unfeasible, it is feasible to recognize if the email lists you are using have more spam traps than Swiss cheese has holes, since most ISPs strictly do not tolerate emails from email marketers to these outlawed addresses. These are easily recognizable when your email deliverability seems to jump off a steep cliff and then through the appearance of your IP address on several blacklists.
What does this mean to me?
The first step to avoiding a buildup of spam traps is to sustain a correct subscriber record. Ways to do this are:
Center your attention on cultivating your house list naturally instead of purchasing an email list from brokers: This is by no means an implication that all email list brokers partake in dubious business practices. With that being said, purchasing a rotten list with data that is not the cream of the crop is the best way inadvertent damage is done to email marketers sending reputations.
Send to active leads/contacts only: Review your lead and contact records on a regular basis to confirm that have actually responded to your campaign emails. Salesforce includes a check box for “Responded” on the lead and contact page layouts – add this field to a lead or contact list report to easily see those leads or contacts that active.
Purchased Lists: If you do purchase lists of leads from a supplier – request that they send an initial email and ask them to forward the results for bounced emails or emails with other deliverability issues. You can then remove these email addresses from your list before uploading into Salesforce.
Scraped Email Addresses
Web email address extractors are useful tools for developing contact lists, sales prospecting and other purposes for which access to information is valuable. An email extractor harvests emails from the internet using standardized protocols. If you purchase a lead list from a supplier, it is very likely that they used a web email extractor tool. When using this tool, it is very likely that email addresses not accessed by “real” people will be included. Examples of these types of emails are sales@domain.com or abuse@domain.com. Emails sent to these addresses are going into shared inboxes – not the the company email or personal address of your campaign member. It is highly likely that these emails are quickly deleted from the inbox and never acted upon.
What does this mean to me?
Confirm the email addresses for your leads and contacts. Avoid sending emails to address with the prefixes listed below. Predictive AM will consider these emails as “invalid email format (20)” and will not send campaign emails to them.
- sales@
- info@
- support@
- abuse@
- postmaster@
- webmaster@
- careers@
- jobs@
- contact@
- pr@
Insight
Use our Insight feature to see what items in your email content may send your email into your recipient’s spam folder. Insight provides Email Previews (view your content in popular email clients and mobile applications) and Spam Analysis (your content tested in over 50 spam filters).
Depending on the Predictive email package you have purchased, Insight may be included or you can purchase a single user license or a site wide license.
For more information on purchasing additional tests, please contact sales@predictiveresponse.com.
(0)
(0)