Enrollment Blog

How not to achieve the enrollment results you want

by Michael Koppenheffer

At Royall & Company, we love finding enrollment management strategies that work for our clients—but sometimes it's equally gratifying to find out what doesn't work. That's why we were excited to complete our latest Entering Class Survey, wherein we gathered data from a large cohort of Royall clients to assess the overall state of the market as well as the effectiveness of specific approaches and interventions.

Overall, we were happy to find that institutions who partner with Royall on enrollment management appear to be outperforming the overall market by a substantial margin. The 196 institutions that responded to the survey had increased the number of students placing deposits for the next year's entering class by an average of 6%, and had raised net tuition revenue by 7%.

While we didn't gather data from non-clients for this survey, and other market survey collection instruments don't match our survey precisely in timing, data from the National Association of College and University Business Officers' tuition discounting survey suggests that the change in average net tuition revenue across all institutions was just 1.2%.

Perils of predictive models

Despite this good news, we found at least a few cautionary notes. Perhaps the most significant finding was that schools that incorporated predictive modeling into their outreach approaches and workflow performed slightly worse than those that didn't.

Like so many people in higher education (and beyond), enrollment managers are looking to the promise of data and analytics to improve effectiveness and efficiency throughout the enrollment lifecycle. Software vendors and market consultancies are making ever-bolder claims about the power of predictive modeling.

But institutions in this year's survey that used predictive modeling to target their enrollment campaigns for the 2016 entering class saw a net decline (on average, compared to the average gain experienced by those not using this method) of approximately 5 points in average SAT scores. Likewise, institutions using this method saw little or no change in yield or deposits, compared to the non-utilizing group.

We found this finding intriguing but not surprising, because it's been corroborated by our own internal analyses. We employ predictive models to great effect to target advisor interventions within the EAB student success platform. But when we've applied similar data science approaches to pre-matriculation data and played out the effect across an entire institution's enrollment marketing campaign, our analyses consistently find that a school that pursues a more "targeted" and "predictive" approach fares worse at the end of the day in terms of applications, deposits, and net tuition revenue.

That's not to say that predictive analytic models might not ultimately prove to have transformative applications within enrollment management—just that their success hasn't been proven yet, and institutional leaders should proceed with a healthy amount of skepticism and caution.

Texting and calling campaigns aren't 'off the hook' yet

Our survey also found that certain other enrollment management tactics increased the cost or complexity of programs while failing to produce any significant improvement in results.

For example, 13 respondents introduced institutionally managed text messaging campaigns around yield. The average performance of those institutions in yield-related metrics was not significantly different from that of institutions which did not use this strategy.

More for enrollment leaders: Get the latest enrollment insights in your inbox

Similarly, 15 respondents introduced institutionally managed parent communications focused on yield, and similarly to those above, did not noticeably benefit.

Thirty-four institutions increased the number of phone calls to admitted students; this tactic seemed to have little effect except for a mild drop (-0.7 points) in SAT scores while their comparison group experienced a slight lift (1.3 points) in scores.

Again, as with predictive models, these findings don't necessarily mean that telephone outreach or text messages can't be useful, high-impact channels for reaching students. But they underscore that the channel itself certainly won't create success on its own. At the very least, enrollment leaders need to be vigilant about how well their teams are using each channel, not just whether they are using them.

Interested in a snapshot of this year's freshman class?

See how your institution compares to other incoming freshman classes.

Download the survey

  • Manage Your Events
  • Saved webpages and searches
  • Manage your subscriptions
  • Update personal information
  • Invite a colleague