Solving the It's My Data Mess

Solving the It's My Data Mess

How a business intelligence system helped a small private college improve its data-driven decision-making

Several years ago The College of St. Scholastica, a Catholic Benedictine school in Duluth, Minn., purchased a business intelligence (BI) system to improve its ability to make data-driven decisions. Along the way, we learned some important lessons that have strengthened us, and that may be of use to other institutions.

A bit of background might be helpful. St. Scholastica was founded in 1912 by the Benedictine Sisters in Duluth. It was a women's college until 1970, when it became coeducational and independent, although still sponsored by the Sisters. The school has a modest endowment and is tuition-driven. We offer programs in the liberal arts and several pre-professional areas; more than half of our students major in health care fields.

Over the past decade, St. Scholastica has changed considerably. We entered the education market for working adults, opened three new campuses around Minnesota, added Division III football, expanded our masters programs, added two clinical doctorates, and entered the online market. Enrollment has grown from 2,000 to 3,900; faculty and staff have increased by 65 percent. We have more than doubled the number of traditional undergraduates living on the mother campus. The budget has grown from $27 million to $69 million. We anticipate more growth as we expand online programming.

The Plan's Price Tag

These changes are a result of ongoing strategic planning, and they have caused us to improve our planning process. We have been clear all along that mission and vision should shape strategic priorities, and these priorities should inform annual college goals and departmental action plans. We operate on a belief that the budget is the price tag for the plan and have gotten better at connecting everything in the plan with a budget number, and vice versa.

We had the right insight, but we struggled to get clean and consistent data.

Consequently, we began to realize the importance of data-driven assessment of results. If we committed a hundred thousand dollars to a retention initiative, did we achieve the anticipated results? If we increased our marketing budget by a half-million dollars, can we tell what difference it made? Looking at our set of aspirant institutions, what benchmarks can we establish from them and how should we strive to realize them?

Several years ago, we began annually reporting important dashboard indicators to our trustees. The effort to refine and expand these measurements led us to establish key performance indicators (KPIs) for every goal in our strategic plan. KPIs give a snapshot of where we are, a trend line showing the direction in which we are moving, and a benchmark we are trying to achieve. For example, we want to realize a four-year graduation rate of 60 percent; we are currently at 53 percent and have been moving slowly but steadily in the right direction in recent years.

Clarifying Needs to Meet Challenges

The process of improving our KPIs caused us to look at purchasing a BI system. We had the right insight—use data to measure the effectiveness of planning and budgeting—but we struggled to get clean and consistent data. Until recently, we faced the following challenges:

  • Departments tweaked and manipulated data to best serve their own interests. Finance might define "full-time student" in one way to meet its needs; the registrar's office might define the term slightly differently to meet its needs.
  • Dozens of homegrown "datamarts" emerged, each good for parochial tactical purposes, but increasingly unable to converse with one another at the all-college and strategic level.
  • A mess had resulted. We had differing definitions of basic terms, an inability to report historical trends, increasing use of IT time to reconcile reports and data, and a lengthening lag time for receiving important data for strategic decisions.

It was in this context that we purchased a BI system. The first lesson we learned was that selecting the right tool required us to be clear about our needs. We were after data that was reliable and accurate, dynamic, serving constituents across the institution, conveniently accessible online, and scalable.

We chose a tool that allowed us to create a data store to collect discrete information (Jane Doe's year in school, major, financial aid package), a data warehouse to collect aggregate information (how many sophomore English majors on the Duluth campus are receiving Benedictine scholarships?), and reporting tools that allow any budget manager or administrator to access the store and warehouse to create reports.

Transcending the 'Turf' Factor

Sound wonderful? It is. But putting the BI system into operation was not easy. The difficulty was not technological or quantitative; it was political: Who controls the fundamental data definitions?

In the meetings of a large committee established to implement BI, it became obvious that people whose primary interests were tactical and departmental couldn't agree on common definitions. Senior administration had to take hands-on control—our second lesson learned.

We formed an institutional reporting committee (IRC) consisting of the vice presidents of academic affairs, finance, and enrollment management, as well as the chief information officer and the institutional researcher. It moved "top down" to establish data definitions to meet the broadest set of organizational needs, to establish reporting standards, and to guide process improvement. It was not the democratic process that campuses often prefer, but it worked. Overall, people are pleased with the results. When registrar, admissions, financial aid, and student accounts staff access the data warehouse, they draw from exactly the same data. Reports are timely, accurate and consistent—and provide the data we need to move forward confident in our mutual understanding.

In the end, arriving at common data definitions requires attention to college strategy. Data should serve strategy, not just discrete tactical purposes. This was our third lesson.

One of our key priorities is to increase enrollment in nontraditional offerings. The IRC retooled our data definitions to accommodate this strategy. We can now view and analyze unlimited combinations for headcount or student FTEs or credits taken by program, campus, or term. A tactical need for operational efficiency yielded to a need for strategic information.

Pricey, but a Quick Payback

A BI system can be costly ($250,000 for us up front) and require significant staff training (intensive for key users for three to eight weeks; ongoing, more moderate, for a year). But the payback is generous and quick. Data reporting has increased ten-fold. Time spent retrieving information has been reduced 50 to 75 percent. Collections work that took a monthly half day now takes less than a minute. The controller saves over eight hours each semester on reconciling numbers. And so forth. Best of all, data is reliable and consistent, allowing for accurate and integrated planning, budgeting, and assessment.

Larry Goodwin has been president of The College of St. Scholastica (Minn.) since 1998. He is a member of the Board of Directors of the Council of Independent Colleges.


Advertisement