A BI solution was in order. Even so, NUHS didn't rush into deploying one. "During a two-year time frame, we discussed what we needed and how we were going to achieve it," adds Werosh, who was the executive on campus tapped to "get the ball rolling" because much usage would be centered around the registrar's office.
After the investigative period, NUHS invested approximately $200,000 in a BI system supplied by Business Objects.
The benefits have been multifold. Today the registration process is clearly viewed instantly. "We can process information on the spot," Werosh reports. NUHS can even determine the peak registration time during the day. For example, most online student traffic falls between 9 a.m. and 10 a.m. "We added additional staff at that time to handle the traffic. Student workers were added to key areas and computer labs since more online use would lead to more questions. Others worked the telephones and were able to call back immediately to those students who had contacted online, but who wanted follow-up," he says.
Florida State University adopted BI several years earlier, according to Rick Burnett, director of student information management. "We saw data as the low-hanging fruit and wanted to start using it to figure out the best students to target." To that end, the university, which enrolls 40,000, uses a BI system that links to its CRM system from Talisma, its databases, and other resources, to determine which classes are filling up the fastest. "An alert goes out when a course is 80 percent full." This, in turn, affects room scheduling, and gives fodder for studying other course offerings.
BI has also added a layer of sophistication to the admissions strategy. Officials at FSU, which receives 55,000 applications each year, once had to cull through each file and hand-code information. If at the end of the decision cycle there was a change in the admissions criteria (such as the need for slightly higher GPA requirements), decision-makers would have to retouch all the files again.
Setting a BI query avoids all that. A new list of qualified applicants can be produced within 30 minutes.
HIGHER ED BI TRENDS
"Business intelligence is in its infancy," notes Jim Strickulis, product marketing manager for Jenzabar, a technology company that offers BI functions via executive dashboards, datamart applications, and other analytics. "Higher ed usually follows about 10 years after business adoption."
An IHE like Ohio State might be well ahead of that curve because it is a large institution, but midrange schools-a common client base for the technology company-will move slower because of cost.
BI systems can range in price from five figures, to even into the millions, depending on the size of the institution and the number of users, says Engelbert. "IHEs are probably the most conservative market. If there is any ambiguity, it slows adoption to a halt."
Her observations are partly based on a Datamonitor survey conducted in mid-2006. At that time, only 12 percent of the 50 IHEs surveyed had already purchased a BI solution. Another 18 percent said they would implement BI in 2007. Sixteen percent said they'll implement at some point, but not this year. And 54 percent said they don't plan to purchase a BI solution.
"The vendor community has to do a better job of communicating and provide better examples of how BI is used," says Engelbert. She adds some words of caution for vendors: In the Datamonitor study, higher ed executives said that service and price were the most important criteria for selecting a BI system. The emphasis on service signals that they know BI is a sophisticated technology that may require much instruction and finessing to work well. Engelbert adds, "Institutions really want and need a strong vendor partner." cliché
Technology Tips
The consulting group BearingPoint advises defining the problems you want to solve before building a business intelligence solution. Anticipate not just current needs, but future ones as well. To that end, the company has a list of BI challenges and best practices: Challenges
Multiple versions of the truth
Limited time for data analysis, too much time wasted on data gathering
Disparate definitions
Data redundancy
Unclear data ownership rules
Inconsistent/incomplete information
No standard reporting tools
Development focus on providing detail data
High IT development costs
Best Practices
Single version of truth
More analysis
Common data definitions
Well-defined data ownership
End-user toolbox and common portal,
self-service
More emphasis on analytical applications, such as dashboards