How To Soup Up the Industry’s Most Advanced Predictive Engine

Capture Higher Ed’s predictive modeling engine, Envision, is the most advanced in the enrollment management industry. And it just got some extra horsepower thanks to Capture Behavioral Engagement (CBE), our cutting-edge marketing automation technology.

Until recently, Envision and CBE existed in relative isolation. That’s not the case anymore. Now behavioral data collected from CBE serve as inputs to the Envision models, allowing us to make applicant predictions not just based on who the student is but on demonstrated interest in the school.

In the first year of a partnership with a school, we were able to accurately predict which students were most likely to apply based on data the school provided as well as about 10,000 data points Capture had about that student’s neighborhood. Using this method, we found that prospective students who we ranked as being in the top 10 percent most likely to apply had indeed applied at seven times the rate of lower-ranked students.

But the second year of a partnership with a school is when things get really interesting.

After a year of collecting data about students’ email response and web behavior, we now know the relationship between that behavior and a student’s likelihood to apply as well as how those things interact with all the 10,000 data points we already know.

By knowing a student’s behavior — and combining that with all the data we already know about who a student is — we are able to take higher education’s most accurate predictions and raise them to a whole new level.

Looking at Fall 2017 predictions, we were already very accurate — above 98 percent. There isn’t much room for improvement above that level of excellence. However, if we included CBE data using our new method, we would have boosted that total close to 99 percent accuracy for those same schools.

What does this accuracy get you?

With an accurate model, a school is able to target students precisely and avoid using expensive or time-consuming forms of communications — direct mail, phone calls or hand-written notes — on students who are less likely apply.

By targeting those students who are most interested in a school, an admissions office is able to reduce waste … so much so that an Envision applicant model pretty much pays for itself.

See for yourself how Capture uses logistic regression, multiple types of data, automatic feature selections, dynamic scoring as well as how we can retrain your model within the enrollment cycle. Take a Free Envision Challenge and let us run your data for free. We’re eager to show you what we can do.

By John Foster, Senior Data Analyst, Capture Higher Ed