I’ve had dozens of conversations with Admissions Directors who feel stuck between a rock and a hard place: trying to increase enrollment numbers in the face of shrinking budgets.
State and local funding for public higher education per student fell by 21 percent from 2000 to 2010. California alone saw $650 million in budget cuts last year.
To add to the headache, not only has the average number of college applications per student grown, but admissions are finding that more and more applicants contact them much later in the process. These trends pose considerable challenges for Admissions Directors to accurately meet their enrollment goals.
How is it possible to meet your increased enrollment goals, when the resources you have to do so are shrinking?
Doing more with less requires innovation. By using data and technology reduce costs, colleges can use “predictive modeling” to transform their recruitment strategies.
An Enrollment Crystal Ball
If you could accurately foretell how likely each admitted student is to enroll, you’d be able to focus your recruiting resources on the students with the highest likelihood to enroll. Say, for example, you could predict that John is 15% likely to enroll. Sarah, on the other hand, is 88% likely to enroll. Focusing on students who are more likely to enroll means you can be much more effective at converting students to enrollment.
What is predictive modeling?
Basically, predictive modeling is a statistical technique. In the corporate world, it is regularly used as a powerful tool to improve operations and forecast specific outcomes.
While predictive modeling has been making inroads into higher ed, there remains tremendous potential for better use in analyzing admitted student data. Noel-Levitz focuses its thought-leadership around how predictive modeling can help admissions teams focus resources on high quality leads and prospective students.
To predict enrollment outcomes, companies like Noel-Levitz use demographic data and combines it with the college’s historical enrollment data. In other words, making a prediction on how a particular student will behave based on others like him or her.
Same budget, bigger enrollment
An admissions counselor equipped with this “enrollment intelligence” can then segment the admitted student pool based on likeliness to enroll. This segmentation allows you to reallocate resources towards students who are more likely to enroll. Taking our previous example, resources otherwise spent on John, who is not likely to enroll, can be reallocated towards Sarah, whom you could convince using targeted outreach.
Below is an example of how to re-allocate resources on students to improve effectiveness.
Figure 1: Enrollment Intelligence allows Admissions to reallocate resources based on yield probability.
Bottom line, predictive data allows admissions teams to increase enrollment while maintaining – or even reducing – costs.
Careful, not all data sets are created equal
Can predictions get more accurate? We think so. But first, two quick definitions:
1) Behavioral data: Information about a student based on habits, preferences and interactions (e.g. questions posed to a counselor.)
2) Demographic data: A student’s socioeconomic statistics (e.g. Asian-American, rural residence, $80k household income).
As you can imagine, predictions based on an admit’s actual actions are typically much more accurate than assumptions based on demographics.
Currently we’re helping colleges track real-time student activity – such as number of friends or logins made – in their Schools App community to predict enrollment. Using this behavioral data to drive your recruiting strategies allows you to focus on those students you can influence most, and consequently enhance the effectiveness of your recruiting resources.
How is your office doing more with less?
We’ve been interviewing institutions both large and small to learn how they are prioritizing recruitment initiatives in the face of budget cuts. I’d love to hear from you by sharing your thoughts in the comments below!