Webinar Rewind: Turning Data Insights Into Actionable Enrollment Decisions

Artificial intelligence (AI) and machine learning (ML) are no longer just buzzwords in higher education—they’re powerful tools that help enrollment professionals forecast student behavior and design interventions that improve outcomes. In the recent webinar, Data Insights & AI Literacy: Practical Strategies to Jumpstart Predictive and Prescriptive Analytics in Enrollment Management, Liaison’s Data Science Curator, Alexis Higgins, and Data Science Manager, Ashton Black, described several actionable strategies for integrating AI into enrollment workflows that don’t require campus professionals to have a coding background.

The session began with an overview of AI fundamentals that broke down the differences between AI capabilities in general and machine learning in particular. While AI encompasses tools like natural language processing, image recognition, and large language models such as ChatGPT, machine learning uses historical data to make predictions or inform decisions. As Black explained, supervised learning—covering classification and regression—is commonly used in enrollment analytics to forecast outcomes like enrollment likelihood, retention, or FAFSA submission. Unsupervised learning, like clustering, helps group students based on shared characteristics without predefined outcomes, supporting exploratory data analysis.

Attendees learned about key concepts central to effective modeling, including features (data elements used for predictions), target variables (the outcome being predicted), training data (historical records), and model performance metrics. The session also highlighted challenges such as leak variables—data that’s only available after an outcome occurs—and overfitting, where models perform well on historical data but poorly on new cases. Understanding these elements ensures predictive models remain robust and actionable.

The conversation then shifted to prescriptive analytics, which goes beyond prediction to recommend actions that can influence outcomes. Using decision tree examples, Black demonstrated how modifying financial aid packages could increase the likelihood of a student enrolling from 5% to over 45%. These examples illustrate the power of prescriptive analytics to optimize strategies at both individual and cohort levels.

As Higgins emphasized, high-quality, well-curated data is the foundation of any successful model. Using a “why, how, and what” framework, she encouraged enrollment teams to align data collection with institutional goals, ensure data is maintained and granular, and select variables that are representative of the populations being modeled. This includes personal, demographic, behavioral, and interventional data—the latter two being especially critical for understanding how students interact with the institution and how interventions affect outcomes. Proper timing, auditable processes, and reproducible pipelines ensure models remain reliable even amid staff turnover or system changes.

Finally, the webinar addressed deployment strategies—transforming insights into action. AI and ML outputs are most effective when combined with human judgment. While models can process vast amounts of data and generate recommendations, enrollment staff provide the qualitative insights needed to make the best decisions for individual students. For instance, while a model may suggest financial aid adjustments, a counselor may know additional context—such as a FAFSA delay attributable to family circumstances—that influences the best course of action.

As the experts noted, integrating AI into workflows requires thoughtful planning. Teams should consider where models fit in daily operations, who the end users are, the format of predictive outputs, and how often models and their effectiveness will be re-evaluated. By continually collecting data from enacted interventions and feeding it back into models, institutions can refine predictions and improve outcomes over time.

For institutions just beginning their AI journey, the panel emphasized starting small: focusing on data quality, building simple predictive models, and learning iteratively. Even without a dedicated data science team, teams can experiment with tools, leverage online resources for guidance, and gradually expand the scope of predictive and prescriptive analytics.

Regardless of an institution’s size or academic focus, predictive and prescriptive analytics offer enrollment teams a pathway from insight to impact. With curated data, thoughtful modeling, and human judgment, it’s possible to make more informed decisions, optimize student interventions, and ultimately improve enrollment and retention outcomes.

Whether a team is just starting to explore AI or looking to refine existing strategies, the frameworks and examples from this webinar provide practical steps to harness the power of AI in higher education.

Ready to learn more? The webinar is now available on demand, so watch Data Insights & AI Literacy: Practical Strategies to Jumpstart Predictive and Prescriptive Analytics in Enrollment Management today.