Data Governance: It's a Process

Starting this month, Vaughn Hopkins, Executive Director of Institutional Research, Planning, and Analytics, Delaware State University and Henry Zheng, Senior Associate Vice President for Strategic Analytics, The Ohio State University, will share a series of monthly articles with AIR colleagues on the topic of data governance. Data governance is not a new topic to those in the AIR community. As recently as 2019, at the AIR Forum in Denver, data governance was an important topic. In fact, there were five workshops and concurrent sessions with data governance in the session titles. There was also a workshop titled “Data Governance: A Primer” by Braden Hosch and a speaker session titled “Sorting It All Out: Data Governance, Policy, and Compliance” by Bethany Miller, Becky Brodigan, and Colleen Wynn. Data governance is prevalent because it is a fundamentally important topic to IR practitioners and is a process that is difficult to implement.

Joe Correia blogged about how industry experts estimate that 80% of data governance projects fail. Anecdotal evidence from conversations with AIR colleagues at national and regional meetings largely support the claim that data governance is not as “turn-key” as some vendors would like you to believe. It is also not something that is doomed from the beginning. Our monthly short articles will try to demystify the concept of data governance by explaining the key elements of data governance, the stakeholders involved, the implementation process, and the key success factors. We will also share some of the successful examples.

So, let’s start this series by defining what data governance is in the context of higher education. NYU’s data governance program quotes the definition from Dataversity: “Data Governance is a collection of practices and processes which helps to ensure the formal management of data assets within an organization. Data Governance often includes concepts such as Data Stewardship, Data Quality, and others to help an enterprise gain better control over its data assets, including methods, technologies, and behaviors around the proper management of data.” This definition first and foremost clearly identifies data as organizational assets.

As such, data need to be systematically managed, governed, and utilized for organizational objectives. However, legacy data issues, functional silos, self-claimed data ownership, and multiple technology platforms often make coordinated data management and shared access to data resources difficult. A key condition for an effective data governance program must start with an unambiguous and strong consensus that data are shared assets not owned by any individual offices. Lehigh University’s data governance policy has perhaps one of the strongest statements. It says that the “university is the owner of all institutional data. No single person, department, or other unit within the institution ‘owns’ any institutional data. Departments and other units within the University, however, have operational-level responsibility for subsets of institutional data.”

Data governance, by recognizing the enterprise ownership of all data assets, is about the collaborative efforts by data stewards, data managers, data analytics professionals, and other stakeholders to develop policies, procedures, and standards to manage the quality, access, and utilization of data as organizational assets. Data governance is not a short-term project, but a continuous process and part of the organizational infrastructure.