Over the last five to eight years, the rules that the largest mailbox providers use to block and filter email have changed dramatically. Large mailbox providers such as Gmail and Microsoft Outlook have taken advantage of new technologies to build a better understanding of how subscriber users interact with types of messages coming from particular senders to make better decisions about whether to deliver each email message. In turn, email marketers must change to ensure that their mail continues to be delivered to their most engaged (and economically valuable) subscribers
Here’s a short primer on how the rules of deliverability have changed:
Deliverability 1.0: Global Reputation
From the early 2000s, mailbox providers made their filtering decisions at a largely “global” level for a given campaign. Campaigns from a particular sender were treated mostly the same way based on a combination of content rules and reputation metrics. The most important factors were things like sending domains (envelope domain, DKIM domain) and the sending IP address of the message.
“classic” negative metrics calculated for these identifiers were complaint rate, “trusted” complaint rate, spam trap hits, and unknown user rates. In some mailbox provider
systems, positive metrics were also calculated for sending identifiers. The most common was looking at the percentage of messages that were rescued from a spam folder. In later years, other
engagement-based global metrics were added to the mix, including the overall engagement of mailbox users with messages that come from a given sender and the percentage of addresses that are being
mailed to which are active mailboxes.
Deliverability 2.0: Subscriber-Level Filtering
Mailbox providers began to realize that a global approach to spam fighting did not take into account the extremely varied preferences of individual mailbox users to messages from a particular source. They started to develop “mailbox level” reputation metrics for mail coming from a particular set of identifiers (domains and IP addresses). These systems decide whether or not to deliver a message to a particular mailbox primarily based on the probability that the recipient will engage with the message.
It’s worth noting that both the global “Deliverability 1.0” and subscriber-level “Deliverability 2.0” filtering systems are being used in combination at the same mailbox providers today. Like all security professionals, anti-spam engineers like defense in depth.
Deliverability 2.0 has profound (and subtle) implications for email marketers. Here are a few to think about:
I hope you find this brief primer on the “new rules” of deliverability instructive. What other implications for email marketers do you see?