Turn “Summer Melt” into “Summer Growth”
Enrollment management solutions to turn “summer melt” into “summer growth”
Key concepts discussed here are:
- Turning “summer melt” into “summer growth” requires a deep understanding of applicant behavior patterns
- The Institution must learn how to identify the “swing” applicants
- Counselor messaging must focus on the specific differentiators of your institution
For far too many of us, we have just accepted the idea of summer melt – that we will lose more deposits than we gain in June, July, and August.
But does that actually have to be true?
As a career Admissions Director and now the Director of Enrollment Success at enroll ml, I have seen how a more precise data-science-driven assessment of applicant behaviors can identify students who may not be as visible in our traditional views, but who are still displaying significant positive behavioral patterns, even this late in the game. In fact, for one institution we recently identified over 400 “swing students” that their team should be heavily engaging with between now and the start of fall classes to increase their yield.
It has become increasingly clear to me that our summer melt premise is not correct: there are applicants scattered across the pool who are on the verge of a late decision to enroll – but until now we have not had the tools or processes to identify them, understand them, or take the correct actions to secure them.
In the coming weeks, I’ll discuss each of these three areas in their own 10-minute micro-class. Yet there are things you can do today to move towards summer growth for your institution.
Turning “summer melt” to “summer growth” requires a deep understanding of applicant behavior patterns.
Changing to summer growth is not just a mindset shift – we cannot just will our students into different behavioral patterns. Instead, we need a deep and thorough understanding of how a student behaves on their way to enrollment. This is the first step.
Take for example one of the swing students I mentioned above, we’ll call him Logan. Enroll ml’s machine learning engine has identified that Logan has a 75% likelihood of enrollment, even though he has not yet submitted a deposit. Why would that probability be so high without the deposit? How do we help Logan to take that next step?
Enrollment teams have been limited in their statistical analysis of which applicants they should be in contact with. A simple one, two, or three variable query about application status, FAFSA, and visit attendance would separate the admit pool. More advanced strategies may factor in email open rate, unmet financial need, and website tracking data. Perhaps a somewhat generic and unexplainable “lead scoring” algorithm that discounts the daily demonstrations of interest a student is making every day. At best, those approaches would put Logan in a group of students similarly situated, but all of the nuances of his engagement are washed away.
With the accessibility of more advanced data science, machine learning, and increased computational power, we have the ability to process dozens of variables to understand deeper patterns that highlight the positive and negative signals of the applicant. Whenever possible, institutions need to consider not just major milestones towards enrollment (visit, application, FAFSA) but also engagement timings and patterns, participation in virtual or on-site events, and the time series between each action to know which pattern shows strength or weakness of the student’s interest.
The admissions office must learn how to identify the “swing” applicants.
Part of why summer melt seems to have been handed down by fiat is how much students close themselves off to outreach after May 1. We develop an overreliance on anecdata and make incorrect assumptions about the students that have not yet enrolled. The reality, however, is that there are applicants scattered throughout the pool who are still actively engaged, and very close to a positive decision. The key is learning to identify them.
We need to move beyond attempting to measure interest by stages or status – and instead move towards true statistical groupings based on an applicant’s propensity to enroll. By identifying and scoring a much broader set of behavioral, engagement, and profile factors, we have the ability to be more precise than ever before. The chart below (on right) shows the direct statistical relationship between yield rate and factor scores. Relying on a dedicated machine learning engine, this particular example is more complex and precise than most colleges will create on their own. But the general concept is clear – more steps cleared increases the likelihood to enroll. With this understanding and identification, the enrollment team has a far more focused and powerful list to reach out individually to each and every applicant.
However it happens, colleges and universities must figure out how to identify their swing students, and be comfortable putting a disproportionate amount of focus on those students. If your summer growth strategies treat all students the same, regardless of their interest in your institution, then yes, you’ll believe that summer melt is inevitable. But it doesn’t have to be.
Counselor engagement needs to focus on the specific differentiators of your institution
A cornerstone of the enroll ml philosophy is to believe in the work of your admissions counselors, and ensure they have the tools they need to succeed. To do this, we work with our clients to empower admissions counselors to lean into discussions of the unique strengths and perceived weaknesses of your institution.
Consider the example of Logan, above. He has done all the things we would expect an incoming student to do, except pay the enrollment deposit. Why? What is stopping him from taking that last step? It is most certainly not because the college hasn’t been telling Logan that the deposit is the next step – he knows. He just isn’t taking that action. To unlock this, a counselor needs to dive deep into the relationship with Logan and uncover the strengths and weaknesses he perceives about the school.
Far too often, counselors shy away from asking the most important and meaningful questions of the applicant. Perhaps it’s a fear of an uncomfortable conversation, a lack of training on overcoming objections, or an unwillingness to reinforce a strength the student sees that the counselor does not.
Value is one of the critical topics surrounding higher education today – but what is value in the mind of a student? In the simplest form, it’s the combination of two interrelated equations. 1. Do the unique and differentiating strengths of this institution outweigh the perceived weaknesses of the institution and 2. Do I believe that I can achieve a positive return on my investment in this education? These questions should be addressed directly with your applicants, without fear or judgment. The sooner a counselor has this conversation, the sooner they can help offset or even reverse summer melt.
I get it – making the argument that summer melt is not an unchangeable law of enrollment physics is controversial. A few years ago, I would have had a similar reaction. However, I’ve always tried to be open to being persuaded by data that challenges my pre-existing conceptions – and the data I’ve seen tells me that there is potential for positive growth all year long.
