Consider the IQ of your IT: Are you making use of business intelligence to provide higher-level analytics for decision making?
By David Tobenkin
"It's helped us remain one of the most efficient institutions in the state in terms of lower cost and high quality," says Gavin Leach. "It's greatly extended the reach of my one-person office," says Joe Dehart. "It was a significant reason for the accreditation team to comment on our level of institutional understanding on our use of data," adds Janice Forsstrom.
So, what is the "it" that these business officers are referring to? It's business intelligence (BI)—computer-based techniques for analyzing a department's or organization's raw data and providing reports, dashboards, and other measurements on which to base smart decisions.
While it's not a new technique—and many institutions have for some time made use of data warehouses to consolidate, retrieve, and review information—a number of factors are driving more sophisticated use of data analytics tools. These include the increasing demands of accrediting agencies for more analytical information; the need for more measurements across institutions; and, increasingly, the push for alignment of funding with performance. The inability to access operational data, or lack of efficiency in doing so, places institutions at a competitive disadvantage.
While systems operate from a common quantitative and informational platform, the sophistication of the tools and the degree to which they have been implemented by higher education business officers varies widely by institution, and even within the same institution, says Theo Bosnak, director, higher education and academic medical centers, at Attain LLC (formerly senior director, education and research, at Oracle Corp.).
"In our business," says Donald Dement, general manager, strategy and planning, at SunGard Higher Education, "we ask institution leaders just how motivated their people are to use business intelligence to solve problems." Right now, says Dement, while many schools have a culture of assessment, most are really looking to achieve a culture of performance. "That's not just semantics," he notes. "A culture of assessment could mean collecting data only at the end of the term. That's lagging data rather than leading data—collecting information as an end in itself, then stapling and filing it all away. It's another thing to collect data as a catalyst to drive change and make informed decisions, which are characteristics of a performance culture."
Establishing such an environment requires setting performance metrics and establishing a few key measures, says Dement. Data based on these criteria tell institution leaders what happened (perhaps a rise in enrollment) and why (new courses, increased marketing efforts), and allows them to project future outcomes and make appropriate institutional decisions to support them.
"What institution leaders want depends on where they are on the maturity curve," agrees Attain's Bosnak. "They generally want analytics through a dashboard, with data points dependent on the individual's role. A senior executive or president of a midsized school wants to monitor the pulse of the school through retention rates, endowment performance, and measures of overall business health. The financial officer wants to track student receivables, administrative expenditures, and grant dollars—and so on for leaders in other departments. Analytical tools can be tailored to these customized needs."
The most common applications reported by business officers and industry experts who commented for this article relate to performance monitoring and metrics, program management, energy optimization, accreditation agency requirements, reporting mandates, and financial systems upgrades.
Here's a look at a number of colleges and universities that are at various points of the BI spectrum.
Northern Michigan University (NMU), Marquette, has been steadily transitioning from an information-based reporting
system to a BI-based one over the past eight years. "We're quite far along the road," says R. Gavin Leach, NMU's vice president for finance and administration, who explains that the institution started implementing data analysis tools as far back as 1995 and began expanding business intelligence functionality in 2003. "When we initially started," says Leach, "we were creating reports to support daily operations. From there, we began offering data to support various stages of planning. Now, we are focusing heavily on performance monitoring and analytics of both internal and external criteria."
Some institutions choose to forge their own paths with proprietary systems that they develop.
Performance metrics and dashboards. "Internally, we've identified key performance indicators [KPIs] in four important areas." says Leach. "Those are enrollment, student success, financial resources, and human resources, and we're creating a dashboard and metrics that our university executive team can use to monitor performance in each of these areas.
"For external data analysis," he says, "we're developing a dashboard that quickly shows us how we compare with other peer institutions' and state universities' key operational areas, such as research cost, instruction, student services, administration, plant operations, and financial aid." NMU imports data to its system from a statewide public database that includes data going back to 1977.
"NMU is also expanding the use of dashboards and metrics management tools that summarize conditions at a high level," says Leach. "Using red, yellow, and green status indicators, we can show performance and how we're trending over time. We've been taking weekly snapshots of our data and we're using them to present a picture of performance."
Admissions. Leach says that at present the university's admissions office is the biggest user of BI. "Staff can easily monitor applications, admitted-student counts, and orientation reservations," he explains. Specifically, the university can, for example, look at the number of admitted students for this fall and compare that to the number at the same point in time last year. The administration office then can drill into that data and look at it by region, by college, or by major. "We can use the status indicators: green, if our admits are up; yellow, if they are within a certain range; and red, if the number is below where we were last year or below the target we've set," explains Leach. "That is valuable information."
Academics. Another area where more people are using the system is in academic departments. Last year, the IT business intelligence team, led by Felecia Flack, director of information services in the department of information technology, worked closely with department heads to determine what their data needs were related to course and program planning. "Suites of new reports were created," she says, "and the department heads now have access to information that they can use to make better business decisions for their departments."
Energy optimization. Leach says that a new area of BI use in the past two years has been in the area of energy optimization: "We're using BI to examine where our courses are being conducted and tying that data to our heating, ventilation, and air conditioning systems. This information allows us to schedule classes in the most efficient buildings and the most efficient zones within the buildings, and to optimize our building schedules." The result: energy savings over the past few years of approximately $200,000 annually.
The university uses SunGard's Banner administrative systems and the IBM Cognos Business Intelligence software tools for financial reporting, analyzing data, and creating dashboards. Most data is stored in Oracle tables and the majority of administrative reporting is done using the university's standby database, which also serves as the institution's disaster recovery server.
Staffing. Because of increased efficiency, NMU provides students and families great value and high quality, which is an important consideration for students selecting a university, Leach says. "Our tuition is the second lowest in the state and our staff productivity level is among the highest. BI tools have allowed us to control costs and find areas to invest in."
With respect to staff, says Leach, "that means we can look at area and cost and compare those figures with those of other institutions, which lets us see how and where we should staff. Where we've had high employee costs, for example, we've automated some of our processes, so fewer people need to monitor these areas. Managers have the information at their fingertips. Given that appropriations for all universities in Michigan have been on a steady decline since 2002, we must look at all operations to see where we can reduce costs—and at tracking enrollment, which is our revenue side."
Reporting mandates and accreditation requirements. Janice Forsstrom, vice president of administration and finance at North Shore Community College, Danvers, Massachusetts, says that her college's implementation of business intelligence in 2007 was driven in large part by detailed reporting mandates. "We have a statewide accountability system, which measures the colleges against 32 key performance indicators in many different areas, such as enrollment, retention, program completion, and some financial measures," says Forsstrom.
"For us right now, everything is about student success. As CFO, I feel that this is the way that institutions have to go. You need a plan to get to this point, given all the accountability and measurements that are imposed on colleges. Everyone is asking for a wide variety of data and you need it to make good decisions. You need a strategic reporting team to address issues such as security (who has access to what?); data definitions (what do you mean by 'full-time student'?); and a governance process as you move forward."
Implementing BI on the campus is a continuing process, Forsstrom says. "We're not all the way there yet, but we have made a lot of progress and we have gotten to the point where we have changed the culture." She says that the institution's business intelligence solution includes SunGard Higher Education's operational data store (for ad hoc reporting) and warehousing systems (for trend analysis), both of which integrate with its SunGard Banner system. Web-based reporting tools allow for a wide number of departments and users to create custom reports and more readily drill down into the data.
Different Speeds, Different Needs
Implementation of business intelligence is generally an evolutionary process, with varying implications dependent on institution size and number of locations.
Getting a single campus in sync. At North Shore Community College, Forsstrom says, areas of the institution have implemented business intelligence at different speeds. "The institutional research office is very involved in it—they'd gotten requests for information all the time," says Forsstrom. "The enrollment area was highly invested to get information, as was finance. BI has met our expectations, but we're not at 100 percent. We need to further focus on academic areas and program review, providing assistance to departmental areas to help them adapt to the newer reporting tools."
Forsstrom says a gratifying moment in the migration to analytical tools came during the institution's 10-year accreditation visit a year ago. "One of the comments the accreditation team made was to commend the level of institutional understanding on our use of data. That was huge for us and validated that we were going in the right direction. The next steps are to continue to grow our ad hoc reporting user base and finalize our warehouse trending analysis system. Our longer-term objectives include dashboards and scorecards for management."
Expanding analytics to multiple branches. Just as BI penetration varies by institution function, so, too, does expansion of the process to multiple campuses in college and university systems.
In 2008, the State University of New York (SUNY) began a universitywide data analytics initiative. By 2010, the university hired Marco Cestaro as its business intelligence solutions specialist and charged him with driving BI implementation throughout the 44 of SUNY's 64 campuses that use SunGard's Banner product.
"Although the chancellor has made her universitywide goals clear, each campus has unique challenges," says Cestaro. "These include establishing more efficient processes, communicating better about the details of the current budget, reviewing the budget on a more regular basis, and ensuring data consistency." As for the latter, Cestaro explains, human resources and the academic dean can have incompatible numbers in their proprietary systems based on certain assumptions. "When asked, for example, 'How many faculty do you have in this budget line?' the HR manager may say there are five full-time employees, knowing that he or she had hired that number of staff for the particular positions. The provost may say that two people are on sabbatical and two are not teaching this semester, so from that perspective, there is only one faculty member available to teach."
There is a wide range of BI maturity levels at SUNY's many campuses, says Cestaro. Some have hired people to do data manipulation, yet no one on campus uses the data for analysis. Others are much further along.
"We are working with university presidents down to administrative assistants to build a culture that can make the best of these tools," says Cestaro. "Many have good information technology, approaches, or processes but are not using them effectively. Staff may spend 80 percent of its time getting the information together and only 20 percent using it. In such cases, I advise those individuals on the kinds of technologies they can implement to help flip that ratio."
Cestaro says university leaders also hope to use BI analyses to align universitywide priorities with the work being done in various areas and levels of the SUNY system. "If you set up good strategies, metrics, and reporting standards—and everyone uses the same information set—managers know what's expected of them. Maybe they are aware that performance will be measured on 10 things. Ideally staff will work every day to feed information from the managers up to the executives. This allows senior management to see what activities and programs are competing for time and resources. And over time, we all hope that the number of competing priorities is reduced as executives reevaluate plans and continually reallocate resources based on solid data analysis."
An Overarching IT Overhaul
In some cases, institutional upgrades to new financial systems have prompted business intelligence advances.
Conversion to a common financial system. "The California State University System has a statewide data warehouse project for which all campuses began in July 2010 to convert to a common financial system," says Amir Dabirian, vice president of IT and chief information officer for California State University, Fullerton (CSUF). "The data warehouse that was installed came with data analytics and Oracle tools. We upgraded the Oracle Business Intelligence Suite Enterprise Edition (OBIEE) to provide more functionality, change the reporting queries, and improve the look of the dashboards. Features like purchasing, expenditures, and budgeting are now all available on the dashboard. Our goal is to enable the college to make data-driven decisions, particularly in tough budget times."
What has emerged is a far more fluid tool, Dabirian says. "Before, we had access to a lot of static data, without the ability to compare and contrast. Now, at our fingertips, we get all the data we need from summary level to detail level, which we can export to Excel and then perform more data manipulation, such as comparing several departments' performance. It changes the paradigm. Business analytics is not just reporting; it's the way you use the system to make decisions effectively."
At CSUF, says Dabirian, "business intelligence is used for ad hoc reporting to ensure compliance with CSUF, CSU, and State of California fund source requirements. Besides providing operational direction and oversight, the combination of financial enrollment and market data recently guided CSUF in the relocation and expansion of our campus in Irvine."
Expansion of access to data. Des Moines Area Community College, Iowa, has revolutionized its access to data since the institution installed a business intelligence solution in 2005. "This has greatly extended the reach of my one-person office," says Joe Dehart, executive director of institutional effectiveness, and assistant to the president. Dehart is also the programmer who implemented the business intelligence solution, an adaptation of an SAS business intelligence tool.
"Previously, people at the institution had to call me and request data," Dehart says. "This product lets them look at information and produce reports on a self-service basis. It puts common data out there and users can generate aggregations, means and modes, and their own customized reports, and save everything to a server. Since users can get information directly, rather than having to call around for it, the time it takes to access information is greatly reduced. In a couple hours, you can generate a new report and everyone on campus can see it."
Dehart says that the college has achieved and exceeded all the analytic goals cited for the system when implementation started in 2005. The broad goals included (1) providing users with data sets that they could aggregate, analyze, and run reports from, without coming up with incorrect results; (2) providing a platform for various measures specifically created for the college; and (3) pushing data to users as well as letting them pull data from the system.
More immediate results. The system has greatly increased the sophistication of the community college's approaches in tackling its rapidly evolving planning issues at the 24,000-student institution, such as enrollment, Dehart says. "Two years ago, for example, we realized that we needed to hire an enrollment director. Information we could act on regarding enrollment prior to the start of the semester gave us the ability to initiate plans if we fell short on our projected enrollment numbers.
"We did hire a director of enrollment. And last October, for example, we started seeing that enrollment looked soft for spring, with numbers weakening as the economy recovered. So management went to work determining why it was soft and initiating plans to remedy the shortfall. We ended up with a 7 percent increase over the previous spring term."
The Customization Question
Attain's Bosnak says that one trend in business intelligence solutions available to institutions is a reduction in the degree to which solutions must be customized. "Institutions may have written lots of customizations in the past, since they were looking at their unique situations," says Bosnak. "Now, they are moving toward standard business processes at the same time that applications have matured. That means more adaptability and analytics in BI solutions. It also reduces maintenance costs and allows institutions to retire old proprietary shadow systems. We see that during upgrades, which are now more user-friendly."
On the other hand, some institutions choose to forge their own paths with proprietary systems that they develop. Among them is Rensselaer Polytechnic Institute, Troy, New York, which designed its own tools 12 years ago. The institute has since continued to refine a fully proprietary data warehouse with business intelligence tools that source SunGard Banner-housed data. The system was built for $2 million and costs $500,000 annually to maintain, says Eileen McLoughlin, assistant vice president for planning and budgeting.
"We believe we can better understand our institute's needs [than an outside vendor]," says Jeff Stark, manager of data warehousing for the 7,100-student university. "That assumes, however, that you have talented staff, warehousing experience—in our case, 12 years' worth—and a deep knowledge of the enterprise resource planning system. Twelve years ago, the vendors did not have a solution that met our needs. Since then, vendor-supplied solutions are showing promise. They offer a jump-start to your BI architecture, but we are concerned about being bound to their environment. We prefer the flexibility of being able to respond to the ever-changing needs of our institution." Stark says the institution relies upon four full-time employees, as well as leveraging of other computing and data experts on campus, to maintain and develop the system.
Also, important, Stark says, was an interactive business intelligence deployment process that focused on delivering rapid, incremental improvements to the system. "We had a very robust strategy for how to build and deliver each data mart," Stark says. "It was very iterative, with the concept of delivering small but high-quality business values as soon as possible. We don't just walk away from users for months at a time—we quickly determine what they need, set priorities, and deliver something in three- to four-week increments. They begin to understand business intelligence and realize how they can use it."
As is the case with many institutions, campus admissions has been a heavy user of business intelligence at Rensselaer, Stark says. "The vice president for enrollment management has spearheaded real initiatives to improve service quality and consider student demographics and enrollment selectivity," says Stark. "The only way to do that is to understand where we were and monitor individual classes as applications come in. We have nearly tripled applications since 2005, and to monitor our progress we needed access to information on a daily basis.
"We have been able to accommodate an increase since then from 26,000 to 110,000 inquiries and from 5,500 applications to 14,500 applications, despite having one less employee to process those applications. We can easily monitor the number of confirmed enrollees and how much they are depositing on a daily basis, which helps us manage our wait list.
"In the past, tracking enrollment tended to be a more static process—we'd look at the data after May," says Stark. "Now the vice president is constantly looking at the numbers. 'Let's go through applications again—did we miss something?' They can see where they are and if there is something we should look at again. Do they need another recruitment trip? It's now a much more iterative process."
"We have also pushed out a faculty dashboard," says Jack Mahoney, Rensselaer director of institutional research. "It's a quick assessment for every department and dean allowing them to look at their department and school by number of courses and sections taught, research expenditures made, and students counseled. The departments can look at the load, see a division is overloaded, and move resources from another division. It's not always popular, but it allows us to be more responsive."
Timing Is Still Everything
Implementation of BI solutions requires a major expenditure of time and resources, and sometimes business officers decide that the timing is not right.
Senior leaders at Ithaca College, Ithaca, New York, for example, recently used a consultant to thoroughly investigate implementation of a new strategic business intelligence solution, but have decided to defer implementation for now for several reasons, says Carl Sgrecci, vice president for finance and administration. First, the institution's provost left at the end of 2010, and Sgrecci wanted to wait for the appointment of a new permanent provost so the college's allocation of time and resources on the project can be driven by a unified leadership team.
In addition, given that the BI solution under consideration came with a hefty price tag, possibly $1 million or more—and the college is also considering implementing a new enterprise content management system—it may be necessary to choose between the two, Sgrecci says. Finally, Sgrecci notes that the success of any business intelligence solution will depend in part on management's ability to ask the right questions.
In the end, says SUNY's Cestaro, there is no one solution for getting business intelligence right. "Effectiveness of data analytics is driven by the business culture, and every culture has a somewhat different take on it."
DAVID TOBENKIN, Washington, D.C., covers higher education business issues for Business Officer.