What is Deliverability?

It seems natural that the first frequently asked question about Deliverability is “What is Deliverability?”

Simply put, Deliverability (sometimes shortened to D12Y) is the ability of your messages to reach your recipient’s inbox. It’s simple to define, and hard to achieve.

In the early days of email, almost every email you would send would arrive in the recipient’s inbox, because there was no such thing as a spam folder; email was used mostly by academics and engineers at that point, and if your message didn’t arrive, it was most likely because of an outage or an error.

Once email spam became a thing, network operators started having to create and implement systems to identify and block unwanted messages (see History of email spam – Wikipedia). Over time, as these systems became more sophisticated, things became harder for legitimate senders, bringing the importance of deliverability to the forefront.

Deliverability Is Not About Delivery

If you came here looking for the answer to the question “What is Deliverability?” you may find yourself assuming that your Deliverability rate is the Delivered rate you see in your dashboard. It’s a common misconception and a logical one: both are based on the word “deliver”, but the terms represent two different stages in the life of a message.

When your mail server (a Message Transfer Agent or MTA) connects to the MTA at your MailBox Provider (MBP), it attempts to send your message. If the MTA at the MBP accepts your message, that counts as delivered, sometimes also referred to as “accepted”. The ratio of messages attempted to messages delivered is often referred to as your “delivery rate” or your “accept rate”.

That said, not every message that is accepted by an MBP will be placed in the inbox. Most MBPs will perform various tests and checks on your message, and as a result, it could be placed in the inbox, the spam folder, or even discarded completely.

Some MBPs even break the inbox down into more granular segments, with Gmail being a common example of a provider that sorts their user’s mail into categories automatically, assigning messages to categories such as Forums, Updates, Social, and Promotions. The important thing is that these categories are still part of the inbox, and still count as good deliverability.

Deliverability Can’t be Measured Directly

So we’ve defined Deliverability and established its importance, that means we need to measure it, right? Absolutely, but the problem is it cannot be measured.

While there are several data points that we can track on a per-recipient basis, including whether the message was attempted, whether it was accepted and whether someone clicked on the links in the message, there is no reliable way to know whether the message went to the inbox.

The reason why we don’t know is that the MBPs don’t tell us, and we can’t track it without the help of the MBP. We can tell what was attempted and delivered because our local MTA keeps logs of the interactions it has with the remote mail servers and we can report on those. We can tell if someone clicked on the links by wrapping the links in a tracking URL and keeping a record of the click before redirecting the user’s browser to the actual link destination. We can track when a message is opened by embedding a tracking pixel in the body of the message and keeping track of when that unique pixel is requested from our servers, but this is unreliable as not all recipients have images turned on, and new technologies like Apple MPP work to actively obfuscate image opens. When a message goes to the spam folder or the inbox, we don’t see anything, because it occurs within the MBPs systems.

Combining Multiple Indicators to Measure Deliverability

Since we can’t directly measure Deliverability (AKA Inbox Rate), we have to piece together multiple bits of information like a detective if we want to get an understanding of our deliverability.

First, we look at the clicks. If someone clicks on a link in our email, they must have received our message, so we have a reliable indicator that our clicks are a portion of our deliverability rate. This number is trusted, if we have 1000 unique clicks, we know 1000 of our messages went to our recipients’ inboxes. In addition, we can reasonably assume that those 1000 recipients opened our message, even if open tracking says otherwise.

While open tracking is unreliable, we can use our open rate as a trending indicator; while the specific number is not reliable enough to use as a KPI, if it trends up or down significantly we can assume that there are more or fewer people opening our message. If it trends down significantly and stays down, we can assume we need to investigate whether our message has reached the spam folder.

Deliverability is Heavily Dependent on Engagement

While filtering technology started out relatively simple, mostly looking at keywords and blocklists, modern filtering technology is based on how your recipients interact with your messages. It makes sense: who knows the difference between wanted mail and spam better than the recipients themselves?

As a result, the key to good deliverability is to send your recipients mail that they want and expect to receive, mail that they will engage with by opening it, reading it, and clicking on its links.

While there are some other best practices around authentication and traffic shaping, for most senders those are things covered by their Email Service Provider (ESP).

Summary

The success or failure of your email marketing depends on having good Deliverability, the ability to get your messages to your recipients’ inboxes. While you can’t measure your inbox rate directly, pay close attention to your click and open rates as an indicator of how your sending reputation is doing. Keep your recipients engaged to ensure you’re sending the right signals to the filters at the Mailbox Providers.