The Anatomy of a Deferred Email MessageRecently, the email sending industry saw some fairly significant issues with delivering email to Yahoo. Senders trying to deliver bulk emails to their Yahoo subscribers found these to be continually \”deferred\”, or temporarily bounced back. It was later discovered that Yahoo had upgraded their filtering systems, which caused this continuous deferring. Since then, they have fine-tuned their systems, and are reporting that the issue has been resolved. Yahoo also noted that any sender still experiencing their mail being deferred is an issue that the sender will need to resolve. I would like to discuss both the mechanics of mailing systems, with respect to deferring, as well as why some ISPs (like Yahoo) choose to defer messages.

MTAs Explained

To the average individual, sending email is all about composing a message and clicking the \”send\” button. The message then is somehow delivered to the recipient. However, there is a whole lot of technology that happens in those few seconds when the message is transmitted. Here\’s a simplified process flow:

    1. Sender composes message to a recipient and clicks \”send\”
    2. Sender\’s MTA (Mail Transfer Agent) or mail server tries to contact the recipient\’s MTA. If the receiving MTA is active and able to accept messages, it opens a connection to the sending MTA
    3. Sending MTA says \”I want to send a message to user@example.com.\” Receiving MTA checks to see if user@example.com exists or is a valid email address for that MTA
    4. If the email address is valid and is working, the receiving MTA delivers accepts the message for delivery.
    5. Recipient receives email message.

Of course there are a multitude of reasons that message may not get delivered. Some of these might include:

  • Recipient\’s mailbox is full
  • Recipients mailbox has been deactivated because of inactivity
  • Recipient\’s mailbox is not valid
  • The receiver may be blocking the sender
  • The individual recipient might be blocking the sender (or the content of the sender)

For these and other reasons, ISPs will \”bounce\” or return a message back to a sender.

Hard and Soft Bounces

When a message is bounced back to the sender, it will have a status code that will tell the sending MTA if the bounce is a hard bounce (permanent) or a soft bounce (temporary). If the bounce is permanent (such as when the email is invalid or undeliverable) the receiving MTA will hard bounce the message, and the sending MTA will not retry that message again. However, if the bounce is temporary, such as when a recipient\’s mailbox is full, or temporarily deactivated, the receiving MTA will soft bounce the message. In turn, the sending MTA will attempt to resend that message again, sometimes many times, until either the sending MTA gives up (after a specified amount of time) or the receiving MTA hard bounces it as a permanent failure.

Deferred Messages

Deferred messages are those that are in a temporary state of neither being delivered nor rejected. Many MTAs will have a deferred process in place where the message is retried over a given course of time, usually 6-8 hours, to attempt a successful delivery. In some cases, messages can be in a deferred state due to technical issues on the receiver end, like the issues that Yahoo experienced as noted above. Other times, as part of their anti-spam efforts, ISPs will sometimes defer messages from a sender that is causing complaints. If the complaint level goes down before the sending MTA retries the message, the ISP may opt to accept the message on the second (or subsequent) attempts. To use the Yahoo example again, once they fixed their technical issues, they still deferred many messages, and will continue to do so, due to their decreased threshold for complaints.

Ultimately, having a deferred system in place is good for the sender as it enables them to retry campaigns multiple times with some of the larger ISPs that can often be overwhelmed with sending/receiving email. However, it also means that you need to be aware of the possibility of such messages being delivered hours after their original intent as well as possible discrepancies with your delivery reports. Therefore, it\’s always a good idea to check such reports 18-24 hours after the initial send for the most complete analysis.

Hopefully this has provided some insight for those senders that have experienced (and may still be having) issues with deferred email delivery.

Until next time,

Drink Responsibly, Drive Responsibly, Email Responsibly.

Jaren Angerbauer
Director of Deliverability
mobileStorm Inc.