Let’s face it: The way in which email now makes it to the inbox is different than it was a couple of years ago. The big dogs of the ISP world (Gmail, Hotmail, and Yahoo) have changed the way they look at and value emails, from a content-identification system to an engagement/behavioral model.
If we want our emails to reach the inboxes and eyes of our target audiences these days, we need to care about how we are engaging via email with our contacts database.
Our reputation, which is based on how people interact positively or negatively with the emails we send, is very important. Did you know that 83% of delivery problems are caused by reputation issues? This also means we need to be cognizant of the fact that there is a human being on the other side receiving our messages. If we don’t take the time and energy to personalize and write content that is interesting to them, it will eventually hurt the deliverability rate of our future emails.
Think of it this way: If you host a party at your house and it isn’t so fun, over time, it will be harder to convince people to attend.
Email Deliverability 101
To preface, email engagement comes in two basic forms: positive and negative. The general rule is fairly simple: The more positive the engagement is, the greater the chance of your emails reaching the inbox. On the flip side, the more negative the engagement is, the lower the chance of your emails reaching the inbox.
While the ISPs don’t tell us exactly how they weigh and assign value to each action, we can assume that stronger engagement actions hold more power than weaker engagement actions. Google’s priority inbox feature is a great example of how ISPs are using users' behaviors to learn what emails are and are not important to them.
In the research paper "The Learning Behind Gmail Priority Inbox," the authors from Google discuss some of the different metrics used to evaluate engagement. They bucket hundreds of measurements into a handful of features: social, content, thread, and label features.
"Social features are based on the degree of interaction between sender and recipient, e.g. the percentage of a sender’s mail that is being read by the recipient," the authors state. "Content features attempt to identify headers and recent terms that are highly correlated with the recipient acting (or not) on the email, e.g. the presence of a recent term in the subject."
Again, while the report doesn’t go into the exact ways Google measures and assigns value to a user’s behavior, it is clear that these complicated algorithms take into account a variety of engagement actions when deciding to mark an email as priority. Gmail’s new email tabs are also used to track engagement.
Positive engagement includes, but is not limited to:
- Clicking on the links in an email
- Marking an email as important
- Moving a message from spam to inbox
- Forwarding an email
- Replying to a sender
- Opening an email but not clicking on the links within the email
- Clicking on a link to change your email preferences
- Archiving an email after opening it
Negative engagement includes, but is not limited to:
- Marking an email as spam
- Deleting an email without reading it
- Clicking the unsubscribe link within the email
- Opening an email and then deleting it (vs. archiving it)
We decided to carry out an experiment of our own to test out how email engagement affects email deliverability rates. We had two sets of IP addresses: one to be used by senders who had a proven track record of sending appropriate content/messaging to its email audience (let's call this the engaged IP) and another to be used by the rest of the senders (let's call this the standard IP).
Our research showed the engaged IP had significantly higher email deliverability rates (for those statistics gurus reading this, a P value less than 0.01) compared to the standard IP. This can be explained by the fact that these senders were doing a better job engaging their audience, which, in turn, increased the chances their emails made it to the inbox and were further engaged with.
How Can You Improve Your Email Deliverability?
From this experiment, we learned the trick is to send emails that promote positive engagement and mitigate the effects of negative engagement.
Remember that all of your small actions add up and affect the future probability of successful email deliverability.
Here are three key email engagement tactics that can improve the chances of your emails reaching your contacts’ inbox.
1) Personalize Messages and Segment Audiences
If you are still doing batch and blast emailing, I’ve got a message for you: STOP! The mentality that all of your messages are applicable and interesting to all of your contacts in your database is simply not reality. That philosophy is truly outbound, pushing your content to those who are not wanting to really receive it.
According to MarketingSherpa’s 2012 Email Marketing Benchmark report, 52% of marketers say they have a great need to improve email database segmentation. Therefore, there is no better time than now to better personalize your message and aim deliverability of your content to a smaller, more specific audience so that you are reaching people that want to hear from you.
By changing your email approach to one of inbound marketing, I guarantee you will see a jump in positive engagement behaviors, like clicking links in your email and forwarding it to others.
While it may seem less intuitive, you should send emails to smaller, more engaged audiences since ISPs take into account past email engagement when deciding whether to let the next email you send land in the inbox. The proof is in the pudding -- Lyris’s Annual Email Optimizer Report illustrates that marketers who segmented their email lists experienced a 39% increase in open rates and 24% better deliverability to the inbox.
2) Use Workflows and Progressive Profiling
Workflows allow you to personalize your content so that it reaches the right people at the right time. Progressive profiling can be used in tandem with new workflows so that you can gather more specific information about contacts and treat them differently and more appropriately.
For instance, using progressive profiling, if you ask a question about a prospect’s budget for your product, you can create a workflow trigger so that if they answer "yes," the email is personalized to respond to the increased level of sales readiness and if they answer "no," an email that is more nurturing/content-based can be triggered.
By personalizing the message, there will be higher positive email engagement, like marking the email as important, and less of the negative email engagement, such as not even opening the email at all.
3) Remove the Non-Responders
When there has been no positive engagement (i.e. archiving the email) or only negative engagement (i.e. deleting the email) from a recipient after six to nine months, it is time to gracefully say goodbye.
We suggest you send out one last email, with a call-to-action that tries to re-engage them with the content that you have been sending, and give them a chance here to unsubscribe easily. You can also create an email letting the subscriber know you already opted them out and give them a call-to-action to opt back in. Fab does a great job of this.
It would also be wise to note in this email that if they don’t click the call-to-action that shows they are still interested, they will be automatically removed from the email list. This way, you are pruning your contacts to keep those who are interested and highly engaged while letting go of those who have lost interest.
Again, by getting rid of people who are only bringing down your positive engagement levels, the ISPs will then see your emails as more relevant and place them in the inbox more frequently.
HubSpot conducted this study from May 2 to July 13 of this year to determine if engagement increased if sending only to contacts with a proven track record of it.
Engagement is defined in this study as recipients of an email clicking a hyperlink in the content. Our system tracked that the contact clicked a link and therefore was considered more "engaged" than contacts who did nothing with the message.
We used two separate groups of IP addresses: one actively sending email ("standard pool") and another that had not yet sent anything ("engaged pool").
Upon choosing to email a list, our system used a proprietary algorithm to determine which contacts were most likely to engage with a message based on previous history, given priority in real-time. These contacts received an email sent from the engaged pool. If no prior engagement information was available, the contacts received an email from the standard pool until it was proven.
We noticed a statistically significant increase in engagement during the first weeks of the experiment, as a result of the engaged pool only mailing contacts that had clicked a link in a previous email. By knowing that a contact recently engaged with the sender, we hypothesized they were more likely to take action on the email than other contacts in the list.
The last three weeks of the experiment found us hitting a saturation point, displayed by the sigmoidal curve of our data, where engagement was similar in both pools as we gained more information about previously unknown contacts.
Sending email to engaged contacts increases the likelihood of engagement, though a saturation point may be reached where the difference is no longer substantial.
(Editor's note: A big thanks to HubSpot Delivery Engineer Evan Murphy, who helped me compile this data and conduct necessary research to put together this post.)
Image credit: Ekkehard Streit