Enrollment and student success leaders know there is no single blueprint for increasing retention and persistence. Meaningful progress rarely arrives as one sweeping reform. More often it comes from identifying a specific student challenge, testing a thoughtful intervention, and measuring whether it actually worked—then doing it again.

That was the throughline of a recent webinar, Two Campuses Prove Data
Can Drive Retention, which features two higher education institutions that use Liaison’s Othot data-analytics platform to support student success by leveraging AI and predictive analytics:

At Augusta University, which is currently operating under a 2022–2027 strategic enrollment management plan, retention has moved from roughly 72% at the plan's outset to 78%, against an 80% target. Semester-to-semester persistence has climbed from about 91% four years ago to 94.5% today, just shy of a deliberately ambitious 95% goal.

Data-Driven Results

The story of Indiana University of Pennsylvania (IUP) starts from a different place. The university was never a stranger to student success work—it had programs for students entering below a certain GPA, programs for high achievers, programs for particular affinity groups. What it didn't have was scale.

"Within these pockets of excellence, we didn't have an approach that was truly scalable," said IUP’s Strategic Advisor for Student Success Paula Stossel. 

The consequence showed up in the data: Retention and persistence rates hovered at roughly the same level, year after year.

A new strategic plan built entirely around student success, with explicit backing from the president and cabinet, created the opening to rebuild. IUP took a core model grounded in national best practices, tested it against what was and wasn't working locally, and carried it across campus for feedback. Hundreds of comments from faculty, staff, and students later, the infrastructure launched in fall 2023 with several deliberate design choices, including:

  • Institutional responsibility for student success, meaning it would be owned holistically rather than by one unit.
  • A strong academic partnership.
  • A dedicated data coordinator.
  • Seventeen student success “navigators,” each carrying a specific caseload and working with students through graduation.
  • A year-round advising center to fill the gaps left by nine-month faculty contracts.

Critically, the model was designed for every student — not only first-time undergraduates, but dual-enrollment high school students and master's and doctoral populations as well.

The gains came faster than expected. For example, fall-to-spring retention in the first-year increased by five percentage points, from 85% to 90%, almost immediately.

Data-Informed Decisions Made in the Moment

At Augusta, a $1.8 million presidential student success investment funded, among other things, a cluster of data analysts embedded directly in departments so decisions could be made in the moment rather than queued behind requests to institutional research.

What that looks like in practice is the marriage of quantitative and qualitative analytic insights. Augusta examined students from last year's first-time, full-time cohort who didn't register for fall, and the model surfaced patterns, including a higher share of asynchronous courses and greater commuting distance.

"That allowed us to paint the picture in a lot more vivid color, "Singleton said.

At IUP, Othot's likelihood scores and other contributing factors helped leaders decide where to deploy human attention across the student lifecycle instead of spreading it evenly and thin. For example, the Othot-powered model made possible to focus on students who should be sophomores but haven't earned sophomore status yet and then consider targeted interventions.

Getting Insight to the People Who Need It

Predictive scores only matter if they reach the person sitting across from the student. For example, IUP puts likelihood scores in the hands of navigators and student success coaches in real time, after significant investment in helping those practitioners interpret the full picture behind the number. Broader campus access is curated deliberately—filtered datasets matched to specific colleagues and specific strategic questions.

These two different institutions had two different starting points—a portfolio of targeted programs at Augusta, a ground-up infrastructure rebuild at IUP—but shared several key strategies, including:

  • Naming the specific problem, such as focusing on students in specific majors with GPAs below certain threshold, or those who haven’t progressed to sophomore status on schedule, rather than focusing on the broader concept of retention alone.
  • Introducing an intervention designed to address the specific problem.
  • Making sure meaningful insights reach the correct decision makers in real time.
  • Adding human layer to outreach and interventions.

To hear the full stories of how IUP and Augusta achieved important retention milestones by using Othot’s analytics insights, watch Two Campuses Prove Data Can Drive Retention.