Free the Data

by Alison Bolen

Swap out 'dean' for 'CEO' and the issues are the same. Even top-ranked universities need business intelligence and the ability to distribute information to the people who need it. That's why Eric Donohue of the University of Washington is using SAS to "free" the university's data.

What do managers of the information architecture for a public company have in common with administrators running a data warehousing program for a top-ranked university? The unceasing pressures to provide faster, better information to ever-growing numbers of business users, analysts and decision makers.

Eric Donohue, manager of the data warehousing program at the University of Washington, sheds some light on how his department deals with these pressures. Recently, he discussed his department's long-term goals and shared his vision for business intelligence (BI) with the editors of sascom.

What are your goals for data warehousing and business intelligence at the University of Washington?

Donohue: Our vision statement is "Free the Data," which really means distributing the data to those people who legitimately have a reason to use it, and improving the data for those intended uses. We have many levels of business users at this university; all of them are screaming for more information to manage their local business requirements. We want to provide information to deans, academic planners and university administrators that will allow them to be creative in the ways they think about the future of the university's development, as opposed to reflecting on the past.

How do you see your ETL solution as the first step toward achieving your mission to "Free the Data"?

Donohue: It is the first step in making the data available for distribution. The more access an organization has to common information, the more likely it is that the organization will run efficiently and well with happy employees who do better work. I also believe that the second step – the data use – is one of the harder pieces of this whole process. If you don't have good BI solutions sitting on top of your data warehouse, then what you end up with are multiple reporting tools designed for purposes other than data analysis. It is very important to have a well thought-out business intelligence platform on top of a data warehouse.

Because otherwise it's just data, not information.

Donohue: Right. You can look around a room, and there are a lot of things that will never register as information. Information is anything you see that actually matters to you – like the chair you want to sit in, the computer you want to use. But you may not notice the grain patterns in your door or the 48 different colors in the spectrum in your carpet. That may not matter to you at all. We have that kind of detail in our data, but you probably don't want to see it. So we have to have something in the middle that allows you to get to what you need.

What are the benefits of business intelligence?

Donohue: The goal is to put good tools into the hands of the people who need them to make sound decisions for their organization's future. Decision makers need to have common tools that allow them to get to the grit of what the organization is doing. BI speeds up the process of knowing more about the nature of your business. The best business intelligence tools on the market easily display information and insight to people who don't have deep analytic skills. A common BI solution set allows you, based on a common data set, to have one point of reference for your organization's data. So if a dean asked his department chairs to project future course offerings for the next five years, the faculty could go out and make independent judgments based on a common database and using a simple interface, without being forced into becoming real institutional information analysts.

At SAS, we like to talk about moving Beyond BI^™ , to actually providing deeper levels of analysis. Explain the importance of incorporating predictive modeling and forecasting capabilities in a BI solution.

Donohue: What ends up happening – by virtue of having a good data set and a good distribution mechanism for displaying analyses – is that you are able to start thinking about what-if scenarios much more quickly than ever before. What if we wanted to do this differently? What if I need to change that? Those kinds of things are terribly valuable, because now I can just sit down at my desk to slice and dice the enrollment history for the last five years to determine the effects of an enrollment cap. Or I can explore financials to define baseline budgets for a curriculum that focuses more heavily on technology-based education. I can understand how that would affect the overall academic portfolio. Without a powerful BI solution, that question would not get asked and answered very easily, because it would take months to assemble that kind of information.

And having access to that information in the first place helps you to know what types of what-if questions to ask.

Donohue: Exactly! So data mining is also another piece of all of this, which involves looking for the patterns in the data that we wouldn't otherwise recognize. With data mining tools, we can set the data free and have the machine start looking for patterns that we wouldn't already recognize. Then we can have those things reported back to us so we can start investigating in whole new ways that we probably never thought about before.

Can you describe some of your long-term goals?

Donohue: I want the people who are responsible for making decisions about how the university operates to have the opportunity to make those decisions fully informed. I want them to be able to have valuable information available to them in a time frame and in a workspace that makes that information useful to them. I want the thinkers that are responsible for the university's future – the academic planners – to really be able to pose the questions that matter. Like how do we fund education? How do students participate in that funding? What kinds of opportunities could we make available? If we are going to handle financial aid, which we all do, what are the things that students really need – as opposed to what we just perceive they need? I want innovation and thought. I want to give them the ability to ask the questions they've always wanted to ask and to find answers to the things they never thought they would be able to ask.

I can tell that you are excited about these ideas. What about others at the university? What are their reactions to these plans?

Donohue: People are certainly looking for this. Ultimately, better information can improve outcomes for our students. That is really where this university and others who are interested in these things are focusing. How do we best serve our student populations? Through better financial management, better student management, better course management, better space management and more.

Bio: Alison Bolen is a corporate communications specialist for SAS.