Data Governance: An Implementation Checklist
One of my favorite books is “The Checklist Manifesto: How to Get Things Right” by Atul Gawande. In this book, Dr. Gawande explains that in our ever-changing world, many processes are complicated with inter-connected parts and dependencies. The level of complexity often exceeds our ability to process and react correctly, consistently, and carefully. Considering the complexity facing us, Dr. Gawande makes a compelling argument that a simple checklist approach can avoid a lot of headache, complication, and more importantly, potential disasters. Can you imagine an airline pilot starting a flight without going through a pre-flight checklist or a surgeon starting an operation without looking carefully at the patient’s charts?
Data governance implementation is a complicated process, and it may make sense to build a checklist to remind ourselves what needs to be done and what to avoid when starting the implementation process. Based on what I learned through three different implementation processes and the shared knowledge of IR practitioners and industry experts, here is my attempt to develop a data governance implementation checklist:
Secure stakeholder buy-in and ensure active involvement: Data governance is an organizational process, and it touches all key data assets across many functional areas. Therefore, it is essential for the entire organization, including the Information Technology (IT) office, to provide the necessary knowledge and participation. Good working relationships are key to effective implementation of a data governance program (Northeastern University). Before and during your data governance implementation, check to make sure that your stakeholders (especially senior leadership and IT) are onboard and engaged. You cannot do this alone.
Assemble the right team for data governance: Data governance is a journey and not a destination. Your team should not be there because they are assigned to the committee and must attend. I have seen a data governance committee that had so many members that routinely only 30–40% of the membership attended meetings. Consequently, the messaging and priorities are not delivered effectively, and the data governance work is followed inconsistently. Make sure your data governance committee is right sized to be effective and focused on getting things done. The core members of the team should include leaders who can influence decisions on data quality, accessibility, security, and transparency.
Find the right technology solutions: Different universities have different technology infrastructure and data asset management capabilities and resources. If you are going to acquire a data governance software or system to help streamline the data governance process, make sure that you see through the sale pitches to look at the fit. Check references from universities with similar sizes, missions, and resource bases as yours. What works for a large research university may not work well for a liberal arts college. Ask both the vendors and the references tough questions before signing the contract. When possible, ask for a testing period to try it out before committing to buy.
Decide if you need a dedicated data governance manager: At a time when resources are tight, hiring a full-time or even a part-time data governance manager may seem like an expense that is not warranted. Many IR directors or staff members are asked to step in to play the role of a data governance manager. Without a dedicated person, the data governance function is by and large carried out by a committee. A dedicated data governance manager is fully responsible for the details of the implementation and accountable for its effectiveness post-implementation. Whether to have a dedicated person is something that should be explored with the decision makers early on.
Develop a phased approach to implementation: Rome was not built in a day. Institutional data governance process touches almost all aspects of the data management operations, and it is worth the time to discuss with key stakeholders to decide the right phasing of the implementation process. To be effective, data assets should be prioritized by data domains and even data segments within each domain. The data governance committee should prioritize data domains based on institutional priorities and ease of implementation. At the beginning of the process, taking on low-hanging fruit makes great sense. Small wins build experience, confidence, and collaborative energy. At one university, their implementation process started with donor relationship and admission contact data sub-domains. Both are priority areas for the university, and the domain leaders are motivated and passionate about organizing their data assets effectively. Such an alignment of organizational needs and data domain focus helped plant the seeds for future success.
Communicate the needs and benefits to stakeholders: For most of the people on campus, data governance is often viewed as something that only involves the technical people and not something that will affect or benefit them. To ensure support and sustained development, effective communications strategies must be developed to engage and inform the campus community. There is nothing more effective when data glossary and dictionaries, data governance process, data security, and data accessibility processes are clearly documented and published at a single web location. Well-documented policies and processes that are accessible can ensure that data governance remains consistent as changes occur in the organizations and programs.
Data governance implementation is a complicated process, and it may make sense to build a checklist to remind ourselves what needs to be done and what to avoid when starting the implementation process. Based on what I learned through three different implementation processes and the shared knowledge of IR practitioners and industry experts, here is my attempt to develop a data governance implementation checklist:
Secure stakeholder buy-in and ensure active involvement: Data governance is an organizational process, and it touches all key data assets across many functional areas. Therefore, it is essential for the entire organization, including the Information Technology (IT) office, to provide the necessary knowledge and participation. Good working relationships are key to effective implementation of a data governance program (Northeastern University). Before and during your data governance implementation, check to make sure that your stakeholders (especially senior leadership and IT) are onboard and engaged. You cannot do this alone.
Assemble the right team for data governance: Data governance is a journey and not a destination. Your team should not be there because they are assigned to the committee and must attend. I have seen a data governance committee that had so many members that routinely only 30–40% of the membership attended meetings. Consequently, the messaging and priorities are not delivered effectively, and the data governance work is followed inconsistently. Make sure your data governance committee is right sized to be effective and focused on getting things done. The core members of the team should include leaders who can influence decisions on data quality, accessibility, security, and transparency.
Find the right technology solutions: Different universities have different technology infrastructure and data asset management capabilities and resources. If you are going to acquire a data governance software or system to help streamline the data governance process, make sure that you see through the sale pitches to look at the fit. Check references from universities with similar sizes, missions, and resource bases as yours. What works for a large research university may not work well for a liberal arts college. Ask both the vendors and the references tough questions before signing the contract. When possible, ask for a testing period to try it out before committing to buy.
Decide if you need a dedicated data governance manager: At a time when resources are tight, hiring a full-time or even a part-time data governance manager may seem like an expense that is not warranted. Many IR directors or staff members are asked to step in to play the role of a data governance manager. Without a dedicated person, the data governance function is by and large carried out by a committee. A dedicated data governance manager is fully responsible for the details of the implementation and accountable for its effectiveness post-implementation. Whether to have a dedicated person is something that should be explored with the decision makers early on.
Develop a phased approach to implementation: Rome was not built in a day. Institutional data governance process touches almost all aspects of the data management operations, and it is worth the time to discuss with key stakeholders to decide the right phasing of the implementation process. To be effective, data assets should be prioritized by data domains and even data segments within each domain. The data governance committee should prioritize data domains based on institutional priorities and ease of implementation. At the beginning of the process, taking on low-hanging fruit makes great sense. Small wins build experience, confidence, and collaborative energy. At one university, their implementation process started with donor relationship and admission contact data sub-domains. Both are priority areas for the university, and the domain leaders are motivated and passionate about organizing their data assets effectively. Such an alignment of organizational needs and data domain focus helped plant the seeds for future success.
Communicate the needs and benefits to stakeholders: For most of the people on campus, data governance is often viewed as something that only involves the technical people and not something that will affect or benefit them. To ensure support and sustained development, effective communications strategies must be developed to engage and inform the campus community. There is nothing more effective when data glossary and dictionaries, data governance process, data security, and data accessibility processes are clearly documented and published at a single web location. Well-documented policies and processes that are accessible can ensure that data governance remains consistent as changes occur in the organizations and programs.