Have you ever felt like an ad was following you? You go to one site and see an ad for something, then on another site you see the same (or a similar) ad, and then it is on another site, and another?
You may have experienced digital display targeting, or DDT. Usually, the ads are specifically targeted, either based on pages you have previously viewed or searched for, or on other information that marketers are able to get, such as your location or previous purchases.
At Capture, we use DDT to target college ads to students, using information such as a previous inquiry, email address, or physical address to find and target prospective students.
Until now, Capture Higher Ed data analysts have measured the effect of digital display targeting by looking at click-through rates. This gives us a good indication of how our campaigns are performing, especially relative to benchmarks. But is DDT effective in driving students to apply and enroll?
To look for an answer, we recently focused on one specific DDT product—household IP targeting. This product is unique because our third-party IP targeting vendor, El Toro, has a patent-pending technology that matches at the individual household level with precise accuracy. That means they get ads to prospective student households, not their neighbors down the street or next door.
This allows us to analyze the ultimate effect of DDT and see if it influenced students to apply or enroll.
The cool thing for testing purposes is El Toro is only able to match about half the students we give them. Why is this cool? It creates an accidental A/B test. While it’s possible students who aren’t matched are somehow different than those that are, this natural experiment can give us a good idea of the effect of DDT—or at least this specific version of DDT.
To investigate the effect of household IP targeting campaigns, we were able to match 760,461 students in Fall 2015 senior new campaigns across 23 schools. All together, these campaigns produced about 9 million impressions.
This gives us a nice big set of data to look at. If DDT makes a difference, it should be clear. And yes, there’s a clear difference.
The new-to-dra (decision ready applicant) rate for students who were matched and targeted was .72 percent compared to .60 percent for those who weren’t matched and targeted. That’s 20 percent higher.
A regression controlling for distance from campus and number of emails clicked (proxies for likelihood to apply and the effect of our email campaigns) confirms there is a statistically significant positive relationship between digital display targeting and application.
Although these campaigns were geared toward driving applications, it seems they also had some residual effect on enrollment. The new-to-enrolled rate for students who were matched and targeted was .10 percent compared to .07 percent for those who weren’t matched and targeted. That’s 34 percent higher.
That gets us to the big takeaway. We can estimate it takes targeting about 774 students to get an additional applicant beyond our other efforts. Based on this, these digital display campaigns generated about 983 additional applicants we wouldn’t have otherwise.
By John Foster, Data Analyst, Capture Higher Ed