Q&A: Making Good Use of Analytics in Higher Ed

Tech leaders at Arizona State University and the University of Arizona explain the rewards of using analytics well.

In this second half of a two-part interview, BI This Week discusses BI and analytics in higher education with two leaders in the field: Henry "Hank" Childers, senior director of enterprise information and analytics at the University of Arizona, and John Rome, deputy CIO and BI strategist at Arizona State University.

Rome has spent more than 20 years at ASU and is a pioneer in data warehousing in higher education. He built ASU's first data warehouse in the mid-1990s and has since coached both educational institutions and industry on building data warehouses, best practices, and the strategic importance of BI. Rome now focuses on dashboards, design, academic analytics, and most recently, big data.

Hank Childers has served as project director of the enterprise systems replacement project at U of A since 2008, overseeing a campus-wide replacement of most of the university's administrative computer systems. He also oversees the BI solution implementation, expanding the availability of business information to managers and executives. Childers also has a wealth of managerial and technical experience in financial services, hospital information systems, and university administrative data processing.

[Editor's note: Read part 1 of the Q&A here.]

BI This Week: One issue with analytics can be finding skilled staff for this very popular area of IT. You both have had success in using student staff. What kinds of skills do you look for?

John Rome: I think we both have good stories about students; I'm really impressed with Hank's [program]. At some point at ASU, we realized that students were an untapped resource -- and an inexpensive one. We have a great relationship with student workers; we've had up to 10 students working on BI in ASU, so yes, I think there's a huge opportunity with students.

Hank Childers: There are two dimensions in the way we work with students. One is [that] we've tended to employ more graduate students in our BI team than any other units that I've worked with or know about here. We usually have two to four graduate students at any given time. In almost all cases, they're students in our business college who are majoring in or studying MIS, so for us, it tends to be MIS, not computer science, where we find people.

We've also hired a number of graduates of our graduate program in MIS as permanent employees. ... In my experience, it's relatively unusual in higher ed to have a good bridge between the administrators and the academic side. ... In bigger institutions especially, I think you'll find that there are lots of people who know something about BI but they're working on the academic side. If you can find a way to make that connection, that can be helpful.

Rome: We initially focused on graduate students as well. We also tried to diversify to having more undergraduates, partly because we could keep them a bit longer. It feels good for us to not only help mentor students in terms of getting skills, but in addition to the coursework, they're getting real live experience. When they leave our institution, there's never been a case in which they haven't been hired right out of the gate. That feels good, especially for the team, knowing that in addition to us hiring these students, when they go out into the marketplace, they're ready to go.

Childers: I think we could take a lesson from John and look at bringing more undergraduates into the mix. We've had great experience with grad students, and they do go on and get fabulous jobs. I can't think of a single one who hasn't gone on to get a good job. It also feels good, as John said -- we're connecting the administrative side of our work with our academic mission and that just feels very satisfying.

Rome: A lot of this centers around finding the talent. Typically, because industry commands so much in terms of salary, it's hard for us to hire ETL developers, seasoned data modelers, or seasoned data warehouse people, for example, so we have to look students as a potential pipeline.

For those outside higher education who are looking for potential employees with analytics talent, is there enough of an outreach between private industry and higher ed? It sounds like you both work fairly closely in placing students.

Rome: Actually, in our case, I think most of the students have found jobs on their own. Their portfolio and the interview typically centers around this-is-what-I've-built-at-ASU, so we haven't really done much in terms of a job pipeline.

Childers: I may not be in the right position to really see this, but it seems that private industry is fairly well tuned into the value of the graduates in these programs. We haven't had to help very much in placement other than acting as a reference. It seems that industry is finding these people; it's just that there aren't enough of them.

Rome: The other thing that's happening is that universities are recognizing the great need in the marketplace for these skills. At ASU, we have set up our W.P. Carey School of Business. We have an analytics graduate program that's going to be teaching things like big data. It's starting this fall, and I imagine that there's going to be a huge demand. I see a lot of other institutions doing the same thing in providing these types of programs.

You're both well-known in this area, and you've accomplished a great deal on your own campuses. When you talk with colleagues from other schools, are there errors you see committed when it comes to analytics and BI in general?

Childers: Yes, I think so. One error is to treat BI as too much of a technical issue without being sufficiently grounded on the business side. When people talk about data quality issues, it always raises a question in my mind. What's really going on here? I don't always accept that [problems with data quality] is the right explanation.

Another thing that is really important is to consciously work on relationships with the people that can make use of the data. Maybe that's a variation on a theme throughout this interview. Analytics is not fundamentally a technical issue as much as it is a business one. The best word is probably a "relationship" issue. I feel that sometimes people do not pay the right attention to that.

Rome: I've noticed, too, that many institutions that want BI don't necessarily want to make the investment towards BI. They can bring in a system, set up a project, plan for it, hire staff, hire a consultant. Then, at a lot of institutions that I know of, they think the rest happens organically. Sometimes it does, but I noticed that plenty of places just don't understand the investment needed to have a strong BI program. I've always said, four to eight percent of your IT staff should be focused on the BI initiative; at many institutions, they might have just one person doing this.

Also, there's certainly some of what Hank talked about. Lots of places focus on the technology -- even we were guilty of that. We focused so much on building the data warehouse, having conformed dimensions, and making all this nice data out there and we never really worried about the customer or delivery. When we changed that mindset and started building dashboards for easy consumption that was a milestone in focusing on the customer. That's when I think the value of what we brought to ASU really grew.

What about getting senior management to agree to the value of analytics and to spend money on it? You're that key high-level person at your institution who understands the value and is pushing analytics forward. Do you have tips for other schools on getting presidents, provosts, and deans to understand the value of data and analytics?

Childers: In higher ed, my sense over the last few years is that there's been a change -- that senior administration is often demanding [analytics] these days. ... There is definitely more demand now. For example, we demonstrated some dashboards that John's team had built... and that certainly got their attention. It's kind of a friendly competition. They said, "Wow! If ASU can do that, why aren't we doing that?" It puts us in the position of saying, "Well, we can do that, and here's what we need."

Seeing something that somebody else has done is enormously persuasive. Dashboards were one of those topics that, when talked about abstractly, just didn't carry the weight of actually seeing it and saying, "Wait a second. We can do that."

Rome: I think Hank is right. In fact, with several institutions, we've actually shown our dashboards -- either we've gone there or they have come to ASU. Once you get your leadership to actually see it, I think what Hank said is true -- senior leadership wants this. It's just that they often don't know what they want until they see it.

One of the nice things about talking to people in higher ed is the openness -- you're both so willing to share what you've done and how you got where you are. Can you direct me to a few resources that show some of what you've done at your respective universities?

Childers: There's the website for our BI team at U of A, which is called Enterprise Information & Analytics, and I can be contacted at

Rome: ASU's dashboards are online at , and I can be reached at

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, & Team memberships available.