Does Segmentation Really Matter? Here’s One Real-Life Example.

Image created by AI.

I get asked this question a lot: Does segmentation really make a difference?

My answer is usually yes, but with an important caveat. One campaign does not prove anything on its own. Good marketing decisions come from watching performance over time, testing ideas and looking for patterns.

Still, sometimes one campaign gives you a signal worth paying attention to.

This one instance is what I'd call an "N of 1" but still a worthy example of what segmentation can do.

A test in increasing registrants with segmentation

A small liberal arts university partner of mine was about three weeks away from Preview Day, and registrations were trailing the previous year. They had already promoted the event for several weeks through emails and text messages. They needed another push, so they asked me to get creative.

“Creative” to many can mean a beautiful design, an animated GIF, an interactive element, a clever headline or a giant call-to-action button. I went in the opposite direction.

I created an email that looked like a counselor had opened Gmail and written directly to the recipient that morning.

There was no university header.

No photo.

No logo.

No preheader.

No frills.

And the subject line was simple: Are you free on [date], [First Name]?

The counselor introduced Preview Day, shared what the student and family could expect and invited them to attend. The email included practical details, an opportunity to receive an admission decision on the spot and a note that the counselor would genuinely like to say hello in person.

I targeted two audiences: students and parents.

But we did not stop at making the emails feel personal. We also wanted them to be personally relevant.

The audience included inquiries in selected geographic markets who had not registered and fell within the university’s low- and medium-engagement tiers. These were not the students already raising their hands at every opportunity.

Within that audience, we created tailored content for six academic-interest groups that were offering Preview Day-specific events.

The customization was tiny. In most cases, it was only two sentences explaining what students in that academic area could experience during Preview Day. Nursing students, for example, learned that they could visit the facilities and see the labs. Everyone outside those six areas received a general academic message.

Then we watched what happened.

The results

Before the emails were sent, Preview Day had 175 registrants. Afterward, that number climbed to 212. That is a 21% increase in total registrations.

The parent response was especially interesting. The student email went to 35,093 students and generated 70 unique clicks. The parent email went to only 3,458 people but generated 38 unique clicks. Based on the size of each audience, parents were approximately 5.5 times more likely to click.

You know what was cooler? The academic segmentation also produced a promising signal:

• Students who received tailored academic content made up 43.3% of the student audience, but they represented 48.6% of the registrations added after the campaign.

Their observed post-send registration rate was approximately 24% higher than the rate among students who received the general fallback message.

Final thoughts

Is this definitive proof that two customized sentences caused someone to register or that a plain email converted? No.

We cannot determine whether an individual registration came from the student email, the parent email or another influence. This was not a randomized experiment with a control group. It is one campaign, one event and one moment in time.

But it tells us enough to keep testing.

Segmentation does not need to mean building 47 entirely different emails. Sometimes it means changing two sentences so recipients can immediately understand why an event is relevant to them.

And creativity does not always mean adding more.

Sometimes, the creative choice is stripping everything away until the email feels like what it was always supposed to be: one human talking to another.