Email Marketing News and Tips

With the proliferation of spambots and spambot networks, email filters have become part of everyday internet life.  As an email marketer, you realize the value of getting your email into the user’s inbox and past all of the filtering.   Spam Filters of various kinds are located in many different locations in front of the user’s inbox.  As a result of the various levels of filtering, it can be difficult to get the email past all of the filters and into the Inbox.  Hence, the rise of such Inbox delivery checking services.

What are Inbox delivery checking services?

Inbox delivery services provide an email marketer with a statistical analysis of deliveries by using a seed list.  This seed list delivers emails to a set of mailboxes that are checked by a service like Return-Path.  Services, such as what Return-Path offers, allows a marketer to obtain percentage metrics of inbox/spam folder/missing deliveries.  Even these metrics have limitations and, as a marketer, you will need to understand these metrics and how relevent they are to your email marketing campaigns.  When using such a service and after a campaign completes, they will compile these deliveries metrics into a report for you to ‘use’ as determination of how well your campaign succeeded in reaching the inbox.  At least.. that’s the theory.

Guaranteed Delivery?

Unfortunately, with ISPs such as AOL, Yahoo and Hotmail, there can be no guarantees of delivery to the inbox.  Although, with diligence, compelling content, proper list building and list maintenance combined with a dedicated IP address, whitelisting, SPF records, Domain Keys and, if desired, subscriptions with services like Goodmail or Habeas, you can increase the likelihood your email will land in the Inbox (or at least, not in the spam folder).  However, you will also need to understand how the filtering process works to understand the reasons why there can be no inbox guarantees.  Please read our three part series entitled Spam Filter Inner Workings to better understand exactly how spam filters work.  Suffice it to say that many ISPs combine ISP controlled filters along with user controlled filters.  User level filters are usually the filters that cannot be circumvented unless the user chooses to allow it.  Server level filters can possibly be circumvented when combined with the use of whitelisting services like Goodmail or Habeas and a dedicated IP address.

Note: The use of services like Goodmail and Habeas costs you additional per email sent (think of it as the cost of a stamp per email) and may also require an additional monthly subscription fee.  Also note, in order to be approved with Goodmail or Habeas, you must be on a dedicated IP and you must additionally wait through a 90 day to 6 month waiting period.

How Inbox metrics work

Services like Return-Path set up a number of Inboxes (AOL, Yahoo, Hotmail, etc) that Return-Path directly owns and controls.  Return-Path knows the logins and passwords to these seed email addresses.  They set up a script to poll these inboxes that they own at certain intervals.  More than this, they set up the timer based script based that’s triggered when an email campaign is launched.  Thus, there is a definitive start-time and definitive end-time for such services to poll for emails sent to these mailboxes from your specific campaign.  Note that Return-Path may be watching for multiple email campaign content using these same inboxes simultaneously.  

So, if you set your campaign is set to start at noon, they will begin polling for emails at noon (or when the first email arrives).   They will continue to monitor these inboxes for a specified period of time (usually around 4 hours) for the same content.  After the 4 hours has elapsed, they stop watching for emails from this campaign.

During this ‘watch’ window, they will see which of their email addresses received emails, which didn’t and which had emails in the spam/bulk folder.  If your report shows a high percentage of missing emails, that doesn’t always mean that the emails did go missing.  It means that Return-Path, as an example, didn’t receive the email within the 4 hour window.   However, most proper mail servers will continue to attempt delivery of emails for at least 24 hours.  Some even attempt delivery up to 3 days.  Since the statistical window closes in 4 hours, that limits their reports to what is received during that 4 hour period.  So, this can skew your reports.

A small sampling of data

Once Return-Path counts which email arrived in which boxes, they will send a report describing delivery percentages.  So, if they have 10 email boxes set up and 7 emails arrived, one was placed in the spam folder and the rest didn’t arrive they will say there was a 20% missing, 10% spam folder and a 70% Inbox delivery rate.  They may only give you this data numerically rather than in percentages.  Note, however, that this is only a small controlled sampling.  They are giving you these statistical metrics based on email addresses that Return-Path owns.  So, here is the crux of this whole matter.  An inbox delivery service IS NOT giving you statistics of how your job actually performed to REAL Inboxes controlled by live people in your list. 

Again, the statistics that any Inbox monitoring service offers you are not based on email addresses that are your real live subscribed people.  Instead, these are boxes that are used for the sole purpose of monitoring by the inbox monitoring service.  So, any extrapolation to your actual list delivery rates may be totally incorrect.  In fact in the example above, out of those 10 sent and 2 went missing, you will need to understand the reason that 2 emails went missing (i.e., the 4 hour limit).  Those reasons for the failures could be something that Return-Path has done specifically to those 3 Inboxes (i.e., user level filtering).  It could be that the server that accepts mail for those Inboxes was having technical difficulties.  It could even be a problem with Return-Path’s monitoring script not properly recognizing the email and properly attributing it to your campaign.  There are a myriad of reasons for both the failure and the spam box delivery that may have had nothing to do with your email campaign and, thus, provided you with inaccurate statistics.

Determining Actual Metrics

Unfortunately, ISPs don’t allow ESPs (or anyone other than the mailbox owner or the ISP’s admins) to monitor an ISP mailbox. Once an email has been delivered to the ISP’s email gateway server, the path the email takes is hidden.  Only the ISP and the mailbox owner have direct access to this information and ISPs will not disclose it to third parties.  Again, only the ISP and the mailbox owner has the ability to see if an email was delivered directly into an inbox.

As an email marketer, the best statistical information about Inbox deliveries is link tracking and openers.  But, this relies on users to actually open the emails and click the links in the emails you send.  However, once a user has clicked a link in your email, you KNOW that it was delivered to the Inbox… or, at least to a box that the user regularly monitors.

Because there is no direct way to monitor an actual user’s Inbox, there is no direct way to know real delivery metrics.  Indirect methods, however, using Boomerang’s ESP tools such as opens, link tracking and replies allow you to conclude which specific users are responding to your content.  Any mailbox monitoring service metrics that you receive may vary as much as 50% (or more) from your actual delivery statistics.  Also, these mailboxes they are using are not actual users.  They are pseudo users set up by Return-Path (or whomever).  So, as you subscribe to such services, you will need to keep in mind that their metrics will likely not represent your actual delivery statistics.

(continued in part II)