At EDUCAUSE, president creates an independent Office of Analytics to drive performance management.
[Editor’s Note: Our editorial picks include stories the editors believe either highlighted an important issue in 2015 and/or signaled the beginning of an escalating trend or issue for 2016.]
It increased student enrollment while reducing student recruitment expenses, increased course completion and persistence rates, and informed policy changes across campus with effective results. It’s also completely autonomous and spearheaded by the college President.
It’s called the Office of Analytics at University of Maryland University College (UMUC), and it began with the vision of President Javier Miyares, who wanted to not only unlock the potential of institutional data across multiple areas, but turn the data into a profit for the college.
UMUC, an open access college with 85,000 student enrollments each year, found themselves dealing with challenges many institutions currently face: a focus and delivery on student success, decreasing funds, an evolving business model, and a completion agenda. And just like other institutions, UMUC knew that data had the potential to help ease the challenges.
Invest in a solution
“The problem is most presidents have heard the word but don’t know how to execute,” explained Miyares during a session at last week’s EDUCAUSE 2015. “We have less than 10 percent of revenue coming from the state and had a 50 percent decline in enrollments in 2012. We had to cut 60 million from the budget and fire 300 people, and that’s when we knew we had to take what we had left and invest in the priority: analytics.”
The Office of Analytics, which has a team of 14 people (half functional data analysts that build analytics dashboards and work with users, and half backend data architects), was designed to be a unit independent of IT and accountable to no specific department.
Using a mix of skills in data engineering, analysis, and business intelligence, the Office focuses on analytics subject areas such as executive and academic program dashboards, marketing, enrollment management, retention, financials, and service centers. A service catalog includes predictive analytics, metrics dashboards, forecast modeling, Ad Hoc reporting, operational reporting, and variance analysis.
Ad Hoc reporting, explained Darren Catalano, the Vice President of Analytics for UMUC, is perhaps the most dynamic service as it interprets data gathered from source systems, replicates the data into higher-performance computing, and then normalizes and models the data in a central repository to be presented in a series of interactive dashboards. An analyst is then assigned to each functional area to interpret the data to users and identify key trends.
“Our approach is to demonstrate the “art of possible” to the institution,” said Catalano; “in other words, to make complex data simple.”
According to Miyares, there are 5 lessons in leveraging analytics to deliver what’s possible:
(Next page: 5 lessons)
1.Organize for Performance
Thanks to the creation of Catalano’s position at the college, the CFO had already seen the value of funding an Office of Analytics. According to Miyares, the work of the original analytics team had helped to predict the drop in enrollment that occurred in 2012,, saving the college millions thanks to advanced planning. However, that didn’t mean small successes by the Office weren’t needed to continue financial support.
“It’s important to get the approval and ongoing support from the CFO,” said Miyares. “So go into the mindset that small successes can make a case for a larger buy-in.”
Catalano explained that the Office itself had to focus on internal organization for success as well, specifically by ensuring a data infrastructure was created, not just a report repository.
“We combine data from multiple areas and systems, like by integrating the college’s LMS data, and then deriving variables for interesting analyses outside of what a typical LMS could do,” he explained. “For example, combining event data with enrollment and seeing how these relate to retention by examining whether or not what initially attracts a student to the college is still what keeps them there post-enrollment.”
Engaging in Skunkworks projects, presenting visually engaging tool sets, and focusing on proactive analytics also helps to not only foster data exploration, but help the university by ensuring performance efficiency.
Miyares emphasized that outside of creating organizational goals, it was important to him to make clear to IT and departmental leadership that the one source of data from the college came through the Office. However, this then led to the Office’s critical organizational task of gaining the trust of senior leaders by presenting clear, actionable data reports everyone could understand and helping to instill a culture of working toward progress.
2.Empower Leaders to Use Data
“’In God we trust. All others bring data,’ is funny, but it’s a mantra that’s useful,” said Miyares. “You have to make clear, especially to faculty, that decisions should be based on data and that this data will be shared and accessible. Think of it as data-driven decision making but with no data-driven agenda—because the Office is independent!”
He also said it’s important that institutions change the culture on holding data.
“Every morning I have a routine: Espresso, and then send the data to all of the departments, making sure that the data is the same for every department—no holding back the data,” he noted. “The four metrics I always share are: number of new students, success of course completion rates, revenues and persistence.”
The Office made good on its organizational goal to produce quick results for CFO support by increasing new student enrollment in fall 2014 by 20 percent, while reducing recruitment expenses by 20 percent due to better targeting and smart spending. “This alone justifies the funding for the Office of Analytics,” emphasized Miyares.
The Office also increased the undergrad successful course completion rate by 7 percentage points and the persistence rate by 4 percentage points over the last 4 years. Data was also used to inform policy changes ranging from length of courses to course registration procedures.
“It’s our goal to retire legacy reporting over time, as well as to build the brand!” said Catalano; for example, through case studies, journal articles, and, of course, conference sessions.
The Office of Analytics is currently working on a market solution based off of their work called HelioCampus, which will use new computing functionality to deliver data insights to each specific institution via a flexible data infrastructure platform.
“I’m not going to say it’s easy to build a successful analytics culture,” said Catalano. “It takes a long time and we never could have done it without Javier. But it can be done.”