The Michelangelos of Data
Universities turn mountains of data into strategic Business Intelligence.
Photo Credit: HUGH SYME
55,000 Students who apply to Florida State University each year
30 Minutes it takes for FSU to update its list of qualified applications using BI programs
Rick Burnette spends most of his waking hours considering the relationships between things such as batches of data, ideas and university processes.
At Florida State University, you might call him the master sculptor of data. In a higher education environment typically slow to adopt technology that has proved critical to the corporate sector, he and his team have created a masterpiece out of what might otherwise have remained a quagmire.
About six years ago, Burnette, director of student information management, began building a system that would elegantly integrate all kinds of data about its 40,000 students with information from a wide range of sources throughout the organization. Today, he’s honed it into a sophisticated business intelligence (BI) system that has transformed many of the school’s strategic operations.
That’s no small task, given the hesitance on the part of many higher ed organizations when it comes to embracing new BI systems, experts say. This hesitancy stems in part from the mess many colleges and universities found themselves in about a decade ago, says John Voloudakis, managing director at Huron Consulting Group of Boston.
“Having invested millions in putting information in data warehousing and ERP [Enterprise Resource Planning] systems, they had no idea how to get it out,” he says. Universities had amassed mountains of data, but were clueless about how to use it.
Even with the variety of BI tools available, implementing them in complex higher ed environments can still prove tricky. Things that work well in the product-oriented corporate world typically need to be fine-tuned for use by education administrators, and it is not unusual for BI vendors to collaborate with universities to customize new product offerings in this area. After all, creating a successful environment for students is a far cry from, say, tracking the progress of a pint of ice cream.
Yet at Florida State, the same type of BI software used to track the product cycle of a container of Ben & Jerry’s Cherry Garcia ice cream is used strategically to track the progress of students from application to enrollment to graduation. Florida State also uses its system to make business predictions and answer key questions by tapping into unique, integrated views of data from admissions files, financial aid, curriculum, and academic and other records that specify a student’s interests, career path, background and future needs.
Although retailers may use BI software to determine the most profitable use of display space in stores, Florida State uses it to recognize the most effective way to reach high-achieving students and retain them.
BI Defined
What exactly is BI? Nicole Engelbert, education technology analyst at Datamonitor, an international consultancy, defines it as software that mines massive amounts of data for patterns, trends and opportunities. In the higher ed world, this also requires integrating financial, enrollment, data administration, admissions, registration, housing, marketing, recruitment and a vast store of information from many areas. “It’s on the rise in higher ed, but being adopted slowly,” Engelbert acknowledges. Indeed, in 2007, higher ed institutions will spend only $14 million on BI. Yet those who are ahead of the pack are reaping big benefits, she says.
In the case of Florida State’s Burnette, the inspiration came during enrollment-management meetings, in which perceptions of an enrollment manager who was also a biologist were far different from the nuanced data coming from other areas of the organization. “It became apparent people were just saying, ‘We feel … ’ or ‘It looks like … ’ or ‘Well, we sense … ’ ” Burnette says. “Suddenly, these hunches were challenged. The provost and enrollment manager recognized we need to use data as efficiently as possible to get a handle on ways to improve our admissions profile as well as retention.”
BI lets Park University keep a close eye on enrollment figures, Director of IT Sara Freeman says.
Photo Credit: EARL RICHARDSON
Higher ed organizations in more competitive markets are typically more likely to be ahead of the BI curve. “They’re more motivated,” Datamonitor’s Engelbert says. “It’s a leadership decision, to initiate change. The best BI organizations are places where the dean or the provost are saying, ‘We need to be more effective.’”
Slice and Dice
These systems can slice and dice vast amounts of data to measure subtle things, such as what impact a 1 percent tuition increase per year will have on alumni donations over the next five years.
Florida State uses a BI system from Business Objects, which includes a simple drag-and-drop interface. The Business Objects system connects to the school’s customer relationship management (CRM) system and other databases. About 55,000 students apply to Florida State each year. With its BI system, the school can update lists of qualified applications in 30 minutes.
“We used to rely on programmers to pull out data and massage it into fixed reports showing how many students applied in state versus out of state, by ethnicity and so on. Now I have a user interface that is intuitive,” Burnette says. “In admissions and records alone, we have probably 20 people using Business Objects on a weekly basis. They refresh it each week to see where we are.”
With the system, Burnette helped administrators create a decision matrix that employs data from several databases to determine which prospective students have the best probability of success at the school.
Among other things, the system allows recruiters to track prospective students from high schools, including an academic profile, contact information and outcome of recruitment efforts for each.
BI has revealed important trends in higher ed institutions. Universities often measure their success in two ways: enrollment and management of students through their lifecycle. How many of the students you brought in actually graduated? And how long did it take them?
BI applications, such as the one Florida State employs, enable university officials to look at trends associated with students who don’t graduate. What factors caused these students to fail?
Huron’s Voloudakis says BI is indeed a powerful tool for proactively solving retention problems. “You might be able to find trends in the data,” he says. “If a student requested to move their housing twice in one semester, it’s likely they’re having personal problems; if you have a student change their payment status to needing financial aid, that might indicate there’s a problem at home. Maybe a parent lost his job and can’t pay any more. Maybe the student had to drop out and go back to work.”
Alerts Warn of Problems
BI systems let academic advisers generate weekly reports identifying students at risk and flagging them. Even better, BI dashboards might show an alert when an adviser turns on his or her computer. Students exhibiting at-risk behavior would be flagged in yellow.
“Advisers can now intervene proactively before the student has failed,” Voloudakis says.
Park University in Parkville, Mo., has 43 campus centers around the country and more than 25,000 students worldwide. Executives and administrators wanted a mobile
system that could be updated instantly so they could see the most current enrollment information while on the road. By integrating systems, information managers at the school created a BI tool that pushes reports to key executives on their handheld smart phones.
“Our administrators travel a lot,” says Sara Freeman, Park’s director of IT. “We also hold enrollment five times a year, more frequently than most schools. Our execs needed to have information at their fingertips, anytime, anywhere.”
Freeman trained the university’s president and six vice presidents on the systems. She says BI enables them to keep an eye on enrollment, a critical factor in the school’s success. Enrollment data is updated twice a day.
“We’re a very tuition-driven institution. Being able to monitor enrollment, ask questions if something doesn’t look right and motivate certain campus centers has been key. We’re also able to monitor our financial well-being,” Freeman says.
Using Key Performance Indicators (KPIs) dashboards, university leaders can view many slices of data at once. For example, they can compare current enrollment with the previous year’s enrollment at the same point in time. The data is highlighted in red, yellow or green for an immediate visual read on which campus centers are not doing well.
“That red really sticks out,” Freeman says. “It blows everyone’s minds when they see it. The president and VPs were not technology-savvy prior to using these tools, and were amazed at their simplicity and power.”
The initial system gave Freeman and her team ideas about an admissions system they’d like to launch next. “We’d like to start with inquiries and work through to ‘applied,’ and to ‘admitted’ and ‘confirmed.’ We’d do that by admissions counselors so we could challenge them to get their numbers in line” using the same color alerts, she says.
Florida State’s Burnette says without BI, the massive stores of data in any organization are like huge hunks of marble before they are chiseled into a meaningful shape. CIOs and information managers can use BI tools to help sculpt the best possible insight into a strategic future.
“We realized the advantage in the future is not necessarily in marketing in the traditional sense, but on having a grasp on data — being able to use it to make decisions,” Burnette says. “The more we understand our data, the better off we are in terms of predicting how well we’re going to do and therefore can make mid-flight changes in our strategy.”
Attitudes about Business Intelligence
BI is still an emerging tech area in higher education.
North America has the most positive orientation toward adopting BI, but Spain and Italy appear to have a strong interest as well.
Mid-sized schools have the most positive view toward applying BI.
Institutions are more likely to adopt BI if their IT budgets exceed $1 million.
Institutions with a growing IT budget are more likely to use BI apps or plan on buying them in the near future.
VOI versus ROI
Unlike the corporate world, in academia the benefits of a BI system cannot easily be measured by return on investment, say BI pundits. A better measure is VOI, “value on investment,” says John Voloudakis, managing director at Huron Consulting Group of Boston. “There aren’t a lot of efficiencies to be gained given the distributed nature of how a college or university works, especially the bigger ones. In higher education, you’re looking at a broader spectrum of value.”
Universities use BI to make people more effective in their work, versus more efficient. That means they can have the same number of people providing even better service.
“Now instead of having people doing data entry, you can have them analyze that data,” Voloudakis says. “With BI, instead of having to struggle to find the information you want out of a bunch of greenbar reports that come out of a legacy environment, you’re able to answer exactly the question you want to answer when you need it.”
Higher ed organizations using BI now spend more time applying information to decision-making.
For example, if a student comes in to an administrator’s office asking how many more semesters he’d have until graduation if he changed his major, instead of saying, “I’ll get back to you,” and running a report, the administrator could answer an ad-hoc query immediately.
At FSU: How Data Figures in the Student Mix
By data mining, “we’ve done extremely well in improving admissions profiles and retentions,” says Rick Burnette, director of student information management at Florida State University.
But the more success an institution has meeting targets through use of student-enrollment data, the harder it is to make gains. “When you talk about moving from an 85 percent retention rate to an 88 percent retention rate, that’s a lot different than moving from 88 percent to 90 percent because you start running into a ceiling,” he says.
The same is true for admissions. The school moved from an SAT average of 1146 to “probably a 1210 for this fall,” Burnette says, “while at the same time growing from 4,500 students in a class to 6,200 entering freshmen. We’re increasing quantity and quality.”
Instead of looking at students as a monolithic, homogeneous group, BI enabled Florida State to look at them as “being what they are — a very diverse population,” Burnette says. It was able to determine that, since out-of-state students are harder to retain, the show rate for the new year had to take into account what percentage of accepted students are out of state, and other categorizations.
The show-rate data can also be parsed in many different ways, based on sex, ethnicity and so on.
This year, FSU’s goal was to enroll 6,200 freshmen and to distribute those between the summer and fall semesters at preset numbers. “As it turned out we enrolled 6,222,” Burnette says. “Our margin of error was pretty low.”
