For higher enrollment managers, the heat is always on. Even during a period when they should be enjoying a little relaxation, they have to contend with “summer melt”–that is, students not returning to their colleges and universities once they’ve closed out their first academic year. About 25% forgo the option to persist, a percentage that has been exacerbated by the unrelenting pandemic and new questions about the value of higher education.
Chris Lucier knows a thing or two about melt, admissions, retention and enrollment. He is now the Director of Partner Relationships at Othot but helped lead enrollment management at the University of Delaware and the University of Vermont and was a director of recruitment at the University of Michigan. He sees the challenges year to year for institutions trying to bring students back, especially in 2022.
“Uncertainty is still out there–with college costs, the question about whether college is worth it, debt concerns, and students being able to make $17-$18 working almost about anywhere with a signing bonus,” he says. “Community college enrollment is really low because those are often students that say, ‘I can find pretty good employment.’ The question is, how does that turn around?”
He says the key is predictive analytics, or using data to see into the future to forecast the pivots institutions might need to make to meet students where they are and keep numbers strong. That only comes with a deep commitment to toss aside old initiatives that only focus on past data.
“In higher education, we view it as ‘We can’t do that, we don’t want to change, or I’m not going to change until they change,’” Lucier says. “Through advanced analytics, we have the ability to help inform us on admissions, financial aid and student success.” And that can help give an early edge to those who embrace it.
To learn more about predictive analytics and their benefits, University Business sat down with Lucier to discuss the future of forecasting in higher ed:
Who is being impacted most by summer melt and enrollments, and how can institutions adapt?
Regional publics are feeling the impact of large, land grant flagship institutions. Pennsylvania is feeling that. In Texas, the growth of UT Dallas and UT Arlington is unbelievable. The big are going to get bigger. So it’s important to be able to optimize resources to where they’re going to have the biggest impact. We’ve always looked backward. We’ve used descriptive or diagnostic data to say this is what happened and why it happened. But the reality is, that students are making different decisions, so that doesn’t allow you to understand them. That’s where advanced analytics can help. The ability to know where to apply research and marketing resources strategically is becoming critical and worth an investment, even though there’s a lot of tightness in budgets.
What is in predictive analytics that is really a difference-maker for enrollment leaders?
Computer algorithms can take hundreds of data fields about students and conduct complex analyses quickly, with results the next day. Othot was one of the first companies in that space, and there are increasing numbers of schools and companies starting to move into it. It’s artificial intelligence and machine-learning based, so the model learns. The model learns much faster than we could ever learn by just putting data on an Excel spreadsheet and comparing one month to another.
It sounds simple, but is there more to it than just implementing predictive analytics to prevent melt?
My background is in the military (Lieutenant Colonel in the U.S. Army), and a great plan won’t matter if you can’t execute it well. If you’re a [regional public] university, the ability to make decisions more quickly than everyone else can be critical. You need to know what the student needs before the student even knows they need it, whether it be personal outreach or an additional $1,000. The admissions process is filled with anxiety and pressure. A lot of times students want to make that decision and move on. Sometimes they don’t have the data, whether it be finances, clarity about a certain academic program or clarity about campus life. Our models update every night. So if the probability of enrolling is 40%, we can send a certain marketing piece for additional financial aid, and now it goes from 40% to 80%. The ability to send that piece this week before next week, maybe a parent says, “That’s great, we’re done.”
How far along are most institutions in being truly ready for the future? And how critical to their survival is jumping in now?
I think most schools are getting some help from third-party providers. The level of sophistication of the help, and the level of knowledge, probably differ a lot. Higher education is very risk-averse, very change-averse. It also follows the leader. Institutional leaders need to wake up. The demographic declines of high school graduates will extend into the 2030s and maybe even to the next decade because of low birth rates. You already have concerns about cost and debt concerns. Now, you add full employment to that. Because of the number of institutions we have, institutions will probably start closing faster. There are still going to be way too many institutions all fighting for a smaller population of students. What are those [regional publics], much less regional privates going to do unless they become a lot better at playing the game and optimizing their resources?
Are those who embrace shorter credentials or new curricula in a better position to avoid summer melt?
That is the bigger elephant in the room. Academic program development is the biggest part of enrollment management that has never caught on. I go back to what higher education is not. They don’t like change. Predictive analytics provides the ability to be more tactical in the short term–the week to week, the day to day. At Delaware, when we were starting to use advanced analytics for student success, I said I want to be able to find what the student needs and when, and then deliver it to that student before they even know that they’re in trouble. Then we will make a difference in graduation rates. But the nimbleness to deliver programs is still not there. It still takes two to three years often to go from an idea to offering a program.
How should universities that have catered to the same audience for years branch out?
They have to find alternative methods of delivery. They have to find different programs. And programs are needed four years from now, five years from now. Offer some programs so students say, this is really cool. It even comes down to the names of what you call your academic program. Talk to business leaders. What are the skillsets they are missing right now? Manufacturing may not appeal to 17- to 18-year-olds because kids are hearing manufacturing is dead in America. What can they call that program that is better?
What are some of the positive outcomes of implementing predictive analytics?
One of the biggest areas that predictive analytics can help is student success and retention. We’ve had partners who have been able to increase the amount of need-based aid and impact the enrollment of low-income students by utilizing very complex algorithms. You can be much more precise in how you will award financial aid. How are these students behaving? How are they interacting with your institution? Are they using the library, fitness center or dining facilities? You can be much more proactive in making decisions. We talk about summer melt. The students who are most likely to melt are the underrepresented, underserved populations. Advanced analytics allow you to find out what that student needs. Some will say, well if they go to orientation, and if we send them enough emails, they’ll stay. Maybe they won’t. But if you know precisely what they need, you can deliver it at the right time and keep that student engaged and persistent.