Avoid this email strategy to improve your deliverability

How many times a week do you miss an important email to the dreaded spam filter?

I was scrolling through my inbox this morning, and I noticed that several emails I was expecting-and even more that I probably would have opened-had been banished to the purgatory that is my junk folder. As someone who has worked in the marketing industry for years, I know how discouraging those wasted efforts can be to the teams who build those emails. And in enrollment marketing, every invested dollar matters. You can’t afford for your emails to live in the spam filter. Therefore, we test.

At EAB, we test hundreds of marketing ideas every year-across billions of student emails and interactions-because we want to stay ahead of emerging practices and uncover opportunities for differentiation among our college and university partners.

I recently came across an insight that highlights the importance of rigorous, ongoing testing.

Testing a Personal “From”

We finished one of our regular quarterly testing and innovation cycles this past summer. One of the tactics we tested was using a personal email sender, like “Michael Koppenheffer,” instead of an institutional email sender, like “EAB University,” or an office sender, like “Office of Admissions at EAB University.” We conducted A/B split tests where we could compare the performance of each type of email sender while everything else was held constant.

In many cases, the personal sender performed significantly better than the alternatives. “Aha,” we exclaimed. “We have a new best practice!”

Meet students where they are with modern, responsive marketing

When we had the opportunity to launch our student search campaign to a new set of students, we decided to apply some of our test-based observations. We introduced more personal senders in several campaigns and waited for the improved engagement rates to roll in.

An Unwelcome Gmail Surprise

Imagine my surprise when I got a message from our head of email deliverability analytics a few days after our summer campaign launched. “We have a problem,” he warned.

Our email sender best practice conflicted with some of our other preferred practices for email subject lines. When we paired a personal sender (someone’s name) with an urgent-sounding subject line, we were seeing warning signs of those emails going into Gmail spam filters instead of arriving directly in students’ inboxes.

A hypothetical example of this pairing is below:

The rate of spam placement of the personal sender emails was 64% higher than the institutional senders in the same campaigns-certainly a sign that we should investigate more closely.

What was so interesting was that our initial tests seemed to be correct about the value of personal email senders. Personal senders performed very well in some instances. But when the personal sender was paired with more direct subject lines-ones that asked the recipient to take an action or seemed time-sensitive-we were seeing signs of spam risk.

Optimizing in (Near) Real Time

After reviewing the first few days of email campaign activity, our creative and marketing teams leapt into action. Our analytics group did an email-by-email analysis for all the partners we tried this new tactic with, identified the ones that were most problematic, and switched them out to safer senders.

Almost immediately, email deliverability was back on track. We averted a potentially serious issue, and along the way we learned more about the nuances of this specific best practice.

I offer this example not because I think everyone needs to understand the arcana of email deliverability (in fact, many of our partners work with us precisely so they don’t have to delve into the weeds!). But rather to underscore the amount of analysis, testing, monitoring, and in-the-moment flexibility that is necessary to really optimize an enrollment marketing campaign. Email deliverability is far from the most glamorous part of marketing, but if your audience never sees your communications, all your hard work is for naught.