For decades, California’s state universities operated under a simple premise: Build programs, and students will come. The state’s population growth and economic expansion meant recruitment was mostly a matter of outreach. Then the COVID-19 pandemic hit, and for the first time in the California State University System’s existence, enrollment began declining across multiple campuses.
At Sonoma State University, the situation became critical. The campus had dropped 40% below its enrollment target over eight years. The university needed to completely re-engineer its recruitment approach to attempt to reverse declining enrollment.
The solution came in the form of a predictive analytics platform that transforms how California State University (CSU) campuses not only identify but also effectively recruit prospective students. Rather than casting the traditional big net, the university can now precisely target students most likely to enroll, dramatically improving recruitment efficiency in an increasingly competitive landscape, according to Ed Mills, the former vice president for strategic enrollment at Sonoma State during the project rollout and the current interim chief enrollment officer at Cal State San Marcos.1
The transformation began with confronting the reality that the recruitment strategies that worked in previous eras were struggling in the new reality. “The Cal State System forever just used the big net approach,” says Mills. “The approach was ‘If you build it, they will come,’ and you really can’t do that now.”
The challenge was particularly acute for mid-tier state universities caught between the expanding University of California system above and the increasingly sophisticated community colleges below. “If you take more from the top and more from the bottom of our funnel, then we’re going to shrink.”
Additionally, today’s students expect a different type of outreach than previous generations. They’ve become accustomed to receiving immediate, personalized responses in other areas of their lives, a far cry from the traditional model where universities could rely on basic outreach efforts.
“This generation of students as consumers is quite different. They’re used to real-time response,” Mills says. “This is the Amazon generation. Their expectations of what a school can do compared to what most schools can do are very different.”
The mismatch creates competitive vulnerability. “They’re kind of a more discerning consumer in many cases. If you can’t respond or reply, or they don’t see you, they may turn their attention to another school.”
Implementing the analytics platform required wrestling with data. In the past, without data, there was a void. Now, oftentimes, with too much data, there is still a void from the perspective of actionable insights. Mills knew that the three years of post-pandemic recruiting data, from 2022 to 2024, wouldn’t lend itself to a meaningful trend analysis on its own.
“Those three years of post-pandemic data were only going to show our three largest drops in the school’s history, almost 10% a year,” Mills says. “So, I also wanted to look at some pre-pandemic data where the school had a very different student body.” The team had to not only reach back to pre-pandemic data to understand the university’s traditional student markets, but also couple that with deep insights on individual student households and geographic patterns to get a more rounded picture of the landscape and how to get tactical in a way that students and their families have come to expect.
The project faced significant operational hurdles beyond just analyzing historical trends. Budget constraints had led to a voluntary retirement program that reduced the information technology team, leaving a small staff that was already stretched across basic operational maintenance tasks.
“The team that was left was really a skeleton crew that did everything they could do just to keep the wheels on,” Mills says. “Asking them to do additional work, though they wanted to do it, was difficult because they were so thinly staffed.” Having actionable insights became imperative in allowing Sonoma State’s team to move faster within their current environment.
Once operational, the analytics system enabled dramatically more effective recruitment campaigns. The platform analyzes variables such as grade point advantage, geographic location, and historical yield patterns to identify prospective students most likely to progress through the enrollment funnel from applicant to admission to actual enrollment.
One early success demonstrated the potential of integrating tactical insights at the individual level: When planning an event at College of Marin for transfer students, the team used the analytics tool to identify 200 prospects with high-yield probability scores rather than conducting broad outreach. “We did a calling campaign and we more than doubled the number of students that attended the event,” Mills says. “If we had just done an email blast to applicants from that area, we would never have had that success.”
The precision targeting proved essential given resource limitations. With recruitment staff also stretched thin, the university couldn’t afford inefficient broad-based campaigns. “You’re throwing a net in the ocean, and you don’t know if there are fish there or not,” Mills says. “You just can’t do the big-net approach anymore. You have to be strategic, and you need sophisticated tools to guide you.”
The success at Sonoma State has attracted system-wide attention as other CSU campuses face similar enrollment pressures.
“There’s a lot the CSU can do,” Mills says. “We are basically the main generator of the professional workforce in California.” The system is preparing to deliver recommendations to the Board of Trustees with “fairly bold things that I think we as a system can do that can transform strategic enrollment efforts.”
As enrollment pressures intensify across American higher education, predictive analytics platforms are becoming essential tools for institutional survival. The CSU System’s experience shows that even resource-constrained public universities can adopt sophisticated data-driven recruitment strategies if they’re willing to drop outdated assumptions about student recruitment in the modern era. While it will take a few years for Sonoma State to rebuild its enrollment, it is now much better positioned to use data to inform and enhance its strategic recruitment efforts.
The transformation from “If you build it, they will come” to precision-targeted recruitment represents more than just a tactical shift. It reflects higher education’s broader evolution from a seller’s market to one where institutions must actively compete for students using the same data-driven approaches that define success in other industries.