What does the individual development plan of the future data professional look like? What skills are needed to stay relevant, and are there differences in skill sets for operational reporting vs. analytics? I sat down with three members of the data community at the University of Minnesota to get their thoughts on these questions, as well as the challenges and opportunities they see in their work with data, reporting, and analytics. Our conversation boiled down to three key tips, which are shared below, along with excerpts from our conversation.
First, let’s meet who you’re hearing from:
John Vlk is a Data Analyst and Team Lead for the Academic Support Resources Student Data and Analytics Team. The Student Data and Analytics team supports the University community by making high quality data available and shareable. The team also supports the Student Success Analytics initiative by understanding and utilizing data in creative and innovative ways to get better insight into students’ experiences, improve retention and degree progress, and make a positive difference in their lives.
Colin DeLong is a Senior Data Scientist & Director of Application Development for the College of Liberal Arts at the University of Minnesota, and directs the operations of the Office of Application Development (OAD). Over nearly 20 years, he has led numerous Enterprise-level projects in the areas of student services technology, enrollment management, data mining, and collegiate/departmental analytics.
Brian Krupski is the Service Owner for Enterprise Data and Analytics within the Office of Information Technology at the University of Minnesota. Brian received his degree in Computer Science from Clarkson University, and brings more than 20 years of IT experience in development, process improvement and program management in healthcare, retail, and higher education. Brian’s service provides support for legacy warehouse and reporting platforms, as well as overseeing initiatives to develop a dimensionally modeled enterprise data warehouse and establish BI platforms to better inform strategic decision making at all levels of the University community.
Tip 1: Take time to see what’s out there
Colin, Brian and John all agreed a critical part of an individual development plan should carve out time to research changes in technology tools and trends related to data and analytics. Reading white papers, technology sites, or connecting with interest groups and peers at other institutions are all meaningful ways to stay connected.
Colin: You need to zoom out periodically from what you’re doing every day. There’s so much going on in data and reporting, and computing has become so cheap. You have to ask yourself, “Is the way I am doing this stuff the way I should continue doing this stuff?” After you do your research, the answer is often, “no.”
Brian: You can make it work (approaching your work with the same toolset) but it doesn’t become economic or scalable. Sometimes you try something and you’re not ready, sometimes the technology isn’t ready, but you need to carve out that space for learning about what’s out there and what’s changed. Doing small experiments and proof of concepts should be continually incorporated into work plans.
Colin: We only have so much ability to continue putting out Excel pivot tables, so that has brought about other changes. Look at Tableau transition at the University of Minnesota. Now everyone needs to know dashboarding on some level. And that’s a thing. By either Tableau or Power BI, OBIEE...you need to think about this and have it in your toolset. Your leaders will go to other institutions and see what they’re doing, and they’ll come back and ask you to do the same thing.
Tip 2: Challenge yourself to tell a story
Regardless of the type of reporting you’re working with, you’re telling a story. The presentation of data ultimately leads back to a question that is being asked; “How many students are going to graduate this fall?” can have both an operational and analytical answer, but both benefit from an approach that communicates the story of the data to the consumer and spark additional questions about what the data means. Get versed in telling data stories in different ways. Colin noted that dashboards are now “a thing,” and operational reporting, predictive, and prescriptive analytics benefit from storytelling which can be achieved through dashboards. John spoke about this in more detail, using an example of looking at student success from both an operational and analytical viewpoint.
John: If you notice a student is off track, do you take action to get them on track? You might use an operational report, and how the data is presented can impact how efficiently you’re able to do that. Or, do you figure out what got them off track in the first place. For some people, each question is more important and the other, but ultimately, both questions are important.
Tip 3: Move beyond “looking backwards” with your reporting approaches
The group talked at length about the need to advocate for time and resources in the predictive and prescriptive analytics space. Reporting approaches don’t exist in a vacuum--they feed each other, and data professionals can influence how resources are allocated in the data and reporting space. For example, if you see yourself as working in operational reporting and you want to stay in that space, you can have an important role in operationalizing the findings from predictive analytics. This means you need to understand how they two fit together, and advocate for the appropriate allocation of resources where needed.
Colin: Operational reporting for sure has its place, and not enough space has been carved out for predictive modeling. With operational reporting, there’s an expectation that this has to happen. For predictive modeling, yeah that has to happen, but we don’t tend to allocate the necessary resources to do it. You have to make it a priority. Someone needs to make leadership aware of what the potential is. People, individually, have to also advocate for themselves.
John: Do we think operational needs inform analytical needs or is it more analytical needs inform operational needs?
Colin: I think they feed each other.
Brian: Yeah, it’s bidirectional.
Colin: I could see operational needs coming out of the analytic output.
John: I think analytics can better shape the understanding of, “This is what is happening,” get to the understanding of why something is happening, and then the operational reporting is defined to support keeping whatever “it” is from continuing to happen.
I asked the group for their final thoughts on how data and reporting professionals should approach their individual development.
John: If you want to remain an operational analyst, you have to be the go-to person on whatever data you’re working with. You can do that. You can be indispensable. But, that’s a lot harder to obtain, rather than moving to a different space.
Colin: PeopleSoft data can’t be all things to everyone. Think about PeopleSoft as the biggest piece of your administrative data ecosystem universe. Historically, there’s a tendency to think about how PeopleSoft can do everything, but there are other systems. So, learn to think about PeopleSoft as the most important of many different systems that you have to talk to. In order to do what we need to do with the data, we need to take it out of PeopleSoft and integrate it. Focus your skillset on what you need to enrich PeopleSoft data with other data, and get versed on ethical use of data as you do this.
Brian: Get exposure to new technologies and new ways of thinking on a more continuous frequency to inform your perspective and approach