Higher Ed Growing into BI, Data Warehousing
- By Linda L. Briggs
As a long-time expert in data warehousing and business intelligence (BI), David Wells has seen the technologies gradually come into their own in higher education. Now a consultant--Wells was formerly director of education for The Data Warehousing Institute (TDWI), a sister institution of Campus Technology--he continues to advise colleges and universities on data warehousing and BI issues.
Two years ago, we spoke with Wells
about the slow progress colleges and universities were making in BI and data warehousing, the reasons behind the measured pace, and about a few superstar schools that were setting great examples.
When we re-visited those topics with him recently, Wells said he sees higher ed making progress in BI, especially in the area of predictive analytics, in which data is analyzed to predict future trends. BI and data warehousing leaders that Wells cited two years ago, including the University of Illinois and Rensselaer Polytechnic Institute, are still setting the standard, but have been joined, he said, by schools such as the University of Texas at Austin and New Mexico State University.Campus Technology: Two years ago, we talked about the relatively slow advances being made in BI and data warehousing in higher ed. What has changed since then?David Wells
: In general, in the industry, there is a much stronger focus on the analytics aspect of BI and data warehousing. In corporate America, the attention is on business analytics, which increases the challenge [for higher ed].
Higher ed is a business, yes, but it is more complex than a business, and it is not always about the bottom line. So, as BI grows more toward a heavy business focus, it becomes more challenging for higher ed to find where they fit into it.
If we don't want to talk about it as Business Intelligence in higher ed, we can call it Institutional Intelligence. The same technology applies, and the same need for analytics and measures of metrics. We just think about the driving forces in an institutional culture.... CT: So higher ed is making some real strides in its use of data warehousing and BI technologies?Wells
: I think they are starting to play catch-up, though I don't have anything but anecdotal evidence to support that. I see increased interest. I don't know whether that increased interest has yet translated into activity.... Economic circumstances are probably creating some drag.
But I see lots of institutions right now asking questions like, how do I gather the right requirements to get value from BI technology? CT: When a school asks you that sort of question, does that mean they already have BI tools in place, and are now trying to understand how to use them more efficiently and effectively?Wells
: Sometimes, yes. Sometimes it's a function of BI tools becoming increasingly integrated in the mainstream technology. The major database vendors, for example, now offer that functionality as part of their database management systems.
You just have to do is look at Microsoft SQL Server and the recent growth in analysis services, reporting services, data integration services.... All the major [database management system] venders have recognized that business intelligence capabilities are, to a large extent, really data management capabilities.
The major technology vendors--IBM and Oracle and SAP--are buying the big BI players. So BI has really become part of the technology mainstream.CT: Speaking of all the consolidations we've seen recently in the data warehousing and BI space, how are those moves affecting higher ed? It sounds like part of the answer is that BI tools now ship with DBMS products.Wells
: That's part of it. The other effect is that it's probably going to make the open source vendors more aggressive and hungry. In particular, that may have an influence on higher ed, which has always been more amenable to open source than corporate America.CT: The last time we talked, there were just a handful of schools leading the charge in this area. Has that changed? Can you name some schools that are doing noteworthy things in BI and data warehousing?Wells
: I still look to the University of Illinois as one of the really powerful leaders. The
University of Texas is doing some good things at their Austin campus; the University of North Carolina has done some good things.
New Mexico State University has taken a score cards and dashboards approach that seems to be unique. One of the challenges for higher ed has been that there don't seem to be many really good, widely accepted sorts of industry standard performance matrixes for higher ed institutions.
For example, a standard measure of performance that you could compare across institutions would be: Is the institution performing well in the area of preparing incoming students? Is the institution performing well in enrolling underserved minorities ... or in its effectiveness and efficiency in funding utilization?
[There's very little] definition of the categories you would measure, and then [metrics to] quantify how well an institution is doing.... I think UNM has made some significant strides in setting standards. Just by defining them for their own institution, they have perhaps set a bar for others to aspire to.CT: You clearly talk to a number of higher ed institutions about these issues. What advice could you offer to a college or university trying to make progress in the areas of DW and BI?Wells
: I think the No. 1 issue is simply getting started. Higher ed has such broad scope that it's a challenge to figure out where to spend your energy. The second challenge is how to gather requirements, which is partly a question of systems and processes, but also human processes and dynamics.
When I teach requirements gathering, I tell people that gathering requirements is simple. You get the right people together and ask the right questions. However, simple does not mean easy. The analogy is losing weight: It's simple, right? Eat less, move more. But simple is not easy....
So in gathering questions, it's get the right gathering requirements, it's bringing the right people to ask the right questions. In higher ed, both the right people and the right questions are different than in the corporate world. You cannot rely on BI best practices from the corporate world.
What I try to do is break the higher ed's space down well enough that we can talk about who and what in a fairly granular way. So, if we want to apply business intelligence to institutional research, the questions we should ask are different than if we want to apply BI to recruiting and admissions, or to financial aid or to campus safety.
The key to higher ed getting around this is to itemize those topics as if they were investments in a BI portfolio. "Now we're going to invest in campus safety; now we're going to invest in recruiting and admissions; now we're going to invest in student engagement. Iif you can work it at that level of granularity, it becomes less imposing to... get the right people and ask the right questions.CT: Given all that, where should a school get started?Wells
: They should start where they have pressing need; that will vary from one school to another. If I look at Virginia Tech, they have reason to focus on issues such as safety. If I look at a little school like Seattle University, a tiny Jesuit School that is like [a beautiful] oasis in the middle of a city ... their pressure is to grow enrollment to where they are a residential campus with finite boundaries.
So you have some interesting challenges there in terms of enrollment management, resident management, and so forth. The right place to start is with whatever is pressing to your governing body or board of regents, assuming that the pressures they are feeling are reflective of the interest of the communities that higher ed institutions serve.
That, I think, fundamentally defines the difference between corporate America and the world of higher ed. With corporate America, all the driving forces end up being directed toward the bottom line: "How does it affect profit?"
In Higher Ed, fundamentally, it should come down to "How does it affect service?" It's about "How well do we serve the community, how well do we serve the students, how well do we serve the social purpose that makes education so essential?"CT: That's a very different perspective from looking solely at the return on investment from a particular product.Wells
: Yes, absolutely. "Are we serving the students, are we serving a community, and how well?" That's ultimately what should drive us toward metrics that are meaningful for higher education.... We can create analytics, but they don't fit the model of conventional business analytics. The return on investment is student engagement. It's time-to-degree; it's a student's post-graduation success.... We are measuring people instead of dollars.