It’s no secret that data analytics is critical for the future sustainability of higher education. But what will it take to unleash the full potential of data analytics?
For one, it requires the adoption of artificial intelligence methodologies — such as predictive analytics and machine learning — to automate and optimize data. And out of all the higher education departments, enrollment services may be likely to adopt AI tools at a faster rate.
Predictions of a deep decline in university enrollment by 2026 were already emerging before COVID-19, with the pandemic throwing in an additional, more urgent element of uncertainty. In May, a McKinsey student enrollment survey found that “15 percent of students are very likely to defer by at least a semester, and up to 45 percent are very likely to look for a different school.”
What Can Universities Do to Re-Enroll Students?
Even before the pandemic, a significant number of students were dropping out of college every year. A 2019 report from the National Student Clearinghouse Research Center found that 36 million Americans have college credits but no degree, translating to countless lost revenue for higher education.
Re-enrolling students who have dropped out is key to securing the future of higher education. To help, one educational technology firm, ReUp Education, is using AI in a particularly resourceful way.
ReUp Education applies a combination of predictive analytics and machine learning to find — and communicate with — stopped-out students (those who take a break from college with the intention of resuming their studies later) who have a higher likelihood of returning to campus. During the spring 2019 semester, the firm’s AI platform helped 30 universities re-enroll more than 8,000 students and regain approximately $25 million in tuition.
Connecting Incoming Students with Advanced Chatbots
Even before the COVID-19 pandemic required the use of emergency communication solutions, universities were adopting AI-powered chatbots for a variety of tasks. One of them is to streamline the complex process of enrolling incoming students.
Chatbots are simple to deploy and update, and they are familiar to college-age users. Approximately 40 percent of millennials already interact with various chatbots on a daily basis. More advanced, conversational AI chatbots have the potential to bring dramatic results in enrollment numbers and related cost savings.
At the University of Oklahoma, the SoonerBot, launched in May 2018, has been given major credit for securing the largest freshman class in the school’s history in 2019. Similarly, Georgia State University saw solid results in reducing “summer melt” by 22 percent within the first semester of implementing an AI-powered chatbot. That number has since grown to more than 30 percent, which has led to hundreds more students enrolling each year at GSU.
Predictive Analytics Refine Student Support That Boosts Retention
The use of AI tools in the recruitment process has resulted in enrollment and retention going up and costs going down. Harnessing the power of predictive analytics to impact the recruitment process and the retention of current students has led to a dramatic increase in class size at Taylor University.
The college implemented Salesforce Einstein for data visualization and then turned to an in-house predictive modeling tool. In the past six years, this approach has resulted in two of the college’s largest freshman cohorts ever, says Vice President for Enrollment Management Nathan Baker.
“While every recruitment year is unique, predictive analytics have helped us improve our business processes, making us more efficient with our recruiting expenses and personnel time,” he says.
Nathan Baker Vice President for Enrollment Management, Taylor University
As potential applicant numbers dwindle, colleges also have more reason to focus on retention. After the University of South Florida implemented a predictive analytics platform, the institution saw “continuing improvements in student persistence and graduation rates for first-time-in-college students,” says Assistant Dean of Advising and Analytics Melissa Irvin.
“Retention increased five points between the 2010 and 2018 cohorts, while graduation rates have been even more dramatic: 24 points in 6-year rates and over 19 points for 4-year rates,” she says.
She adds that USF also has several projects in development to support the use of machine learning in admissions, including robotic keystrokes to optimize processing times and automation of transcript optical character recognition.
AI tools can also lower the costs of marketing research and communication by accurately pinpointing students who are most likely to apply. Taylor University was able to effectively allocate recruiting expenses, primarily with direct mail costs. “Using rich data systems, we can more strategically target prospective students with mail that is applicable to them, making sure our recruiting dollars have a further reach,” says Baker.
Although AI tools can shift and modify the roles of enrollment management staff, they exist to enhance and support employees, not to replace them.
At Taylor, Baker says, engaging with AI platforms “expanded our need for personnel to focus on data collection, quality and predictive analytics, while streamlining enrollment management business processes.”
Data-crunching tools still require a human at the helm, and Irvin says that “one important element in South Florida’s approach is that the increased use of data and technology has not superseded the value placed on human intelligence. It is the symbiosis of the two that supports our continued success.”