Last spring, the scientists in Capture Labs — the technology development team of Capture Higher Ed — started building applicant models for our partner universities using our comprehensive predictive engine, Envision.
This patent-pending technology, which is the only comprehensive predictive engine in the enrollment management market today, uses machine learning and non-linear analytics to accurately predict application and enrollment rates of prospective students.
It’s one thing for predictions to be accurate in the lab, but we have proof now that Envision applicant rankings are accurate in the real world. Perhaps more importantly, as they are put to use in a real-world environment, they are saving our clients’ money. Here’s a great example.
One of our partner universities, a small, historic Christian college in the Southeast, used an Envision applicant model to rank their students by likelihood to apply, then served online banner ads to the 10,000 students most likely to apply. They mailed postcards to the top 4,500 of those students. Doing this they saved about 65 percent off their Digital Display Targeting costs compared to communicating with everyone in their pool. They saved 80 percent off their direct mail costs.
They started these campaigns on Oct. 3. By Dec. 14, 101 of those students had already applied and been admitted. That’s twice the applicant rate of the rest of their pool. Another 77 of those students have started an application.
Next week, we’ll look at how one of our largest partners successfully used Envision applicant ranking to prioritize direct mail.
By John Foster, Data Analyst, Capture Higher Ed