Carmen McKenzie of the Washington State Board for Community and Technical Colleges (SBCTC) offers practical advice for establishing data governance and standardizing data definitions, security, policy, management and data quality.
Data is everywhere and in higher education there is certainly A LOT of it. There is data about students, faculty, employees, research, extracurricular activities, and residences, to name a few. As institutions are looking to improve their services and increase their cost savings, they need to apply effective strategies to manage that data. This is where Data Governance comes into play. It serves as an institutional function that sets the parameters and creates the processes for managing data-related issues and facilitating data-driven decisions. Developing a data governance framework is no easy task. There are so many questions about ownership, stewardship, and custodianship, as well as inconsistencies across how different departments define data and business processes. To tackle a Data Governance Program, institutions need careful planning, the right resources, and a solid information infrastructure with the appropriate tools.
There are plenty of best practice examples out there, whether you look at Educause, vendor case studies or blogs. At Alliance 2017 in Salt Lake City, I had the privilege to monitor the session “Data, Data Everywhere” presented by Paula McDaniel from the Washington State Board for Community and Technical Colleges (SBCTC). This is where I met Carmen McKenzie, Data Services Director over at SBCTC. She graciously agreed to answer a few questions regarding her experience in working in data governance. In this interview with Carmen, we explore the various business and technical issues that SBCTC juggled with on their way to setting up a very successful data governance program. Carmen also provides plenty of insights for how to address these issues along with lessons learned. The Technical and Reporting Advisory Group thank Carmen for her time and valuable input.
Anna Kourouniotis: How long have you been working in data governance?
Carmen McKenzie: Officially, I became involved in data governance in 2011 as our system set out with a goal of implementing a data governance program. Unofficially, I’ve been involved with data governance since I began working at the Washington State Board for Community and Technical Colleges in 1998. I say this because my positions have always been focused on data management, data accuracy and data security.
Anna Kourouniotis: In your own words, what is data governance?
Carmen McKenzie: I believe data governance is the organizational management of all things related to data that includes decision rights and the authority to enforce. Data governance creates a culture of data quality by combining data management, data quality, and data policies through a system of decision rights. Data governance is not a one-time effort; rather, it requires ongoing monitoring to support continuous improvement. It deals primarily with orchestrating and standardizing the efforts of people and processes to optimize data integrity and quality.
Anna Kourouniotis: How many people are involved in the data governance program at WSB?
Carmen McKenzie: There are 16 official members of our Data Governance Committee. The committee is comprised of two representatives from each of our seven commissions in addition to the co-chair from the State Board and the co-chair from our Research and Planning Commission. The Research and Planning Commission is the sponsor of our program. The seven commissions represent all functional areas of our college system including: student services, information technology, instruction, human resources, finance, communications and institutional research. The membership of each commission includes at least one representative from each of our 34 community and technical colleges. These representatives are typically Vice-Presidents or Deans of colleges.
Most commissions also sponsor one or more councils, for example, the Admissions and Registration Council falls under the Student Services Commission. These councils represent the operational functions of the colleges and are designated as our system’s data domain stewards. The membership of the councils have a similar makeup as the commissions in that each college is represented. The members of the councils are regarded as the subject matter experts. The Data Governance Committee leverages the councils and their subject matter experts to develop proposals that will then be provided to the data governance committee for review and approval. The number of members on the sub-committees fluctuate depending upon the topic at hand.
Anna Kourouniotis: Was everyone clear about what data governance meant?
Carmen McKenzie: At the beginning, I do not believe everyone understood what data governance meant and most had differing ideas. When the system began to discuss implementing a data governance program, we established a working group called the “Data Task Force”. We quickly realized that data governance can hold a different meaning depending upon your place in the organization.
We also realized that there are different data governance program focus areas such as the standardization of data definitions, security, policy, management and data quality. It took several months of discussions to come to an agreed upon the definition and to determine which program structure and focus that would work best in our system.
Anna Kourouniotis: Was there an event that triggered WSB’s data governance program or was it an evolutionary process?
Carmen McKenzie: A number of events created the opportunity to formalize data governance in the system. As a result of a study done by our state’s legislature, the Business Affairs Commission (BAC) was charged with standardizing the chart of accounts definitions for the financial data of all of the colleges in our system. The Research and Planning Commission’s Data Stewards Committee was charged with standardizing data definitions and improving data quality.
At this time the system also began work to implement a new ERP system replacing the existing student management, HR and Finance systems. A leading research and advisory company recommended a data governance program be established at the onset of this work.
These three events occurred at approximately the same time providing the environment and desire to establish data governance.
Anna Kourouniotis: What was the biggest challenge in your data governance implementation?
Carmen McKenzie: The biggest challenge in implementing the data governance program was determining the structure of the committee. Do we create new committees and working groups? Or do we leverage the existing organizational structure of our commissions and councils? Or do we leverage the governance structure of the ERP project? We finally agreed upon using the Non-Invasive Data Governance model developed and copyrighted by Robert Seiner. This model allowed us to leverage our existing commissions and councils and only form one new system committee, the Data Governance Committee.
Once we settled on a model and crafted a draft charter, we were able to obtain system buy-off fairly quickly including an approval from our college presidents to move forward.
Anna Kourouniotis: What was the biggest surprise?
Carmen McKenzie: I think the biggest surprise has been how easily the system accepted and endorsed the concept of data governance. Initially I expected only certain areas would welcome a new governance model such as staff who work closely with data in IT and Institutional Research, but was surprised to see it easily accepted in all other functional areas.
Anna Kourouniotis: What activities and deliverables did you need to prepare prior to implementing data governance?
Carmen McKenzie: Much time was spent retrofitting Bob Seiner’s Non-Invasive Data Governance model to align with our system structures, but it was well worth the effort. Once we had a model, we prepared a draft charter that included the purpose and vision of the committee as well as scope, guiding principles and committee composition. We also created a one-pager (that was really 3 pages!) that explained the overall structure and scope of the committee.
These three documents provided enough context for our college presidents to approve the structure. Before they gave us their approval however, they did ask us to determine a commission to sponsor the committee. This was the first time the system had proposed to form a cross-functional standing committee that did not directly report to an individual commission. The Research and Planning Commission agreed to sponsor the committee and we began soliciting representatives from each of the commissions and preparing for a kick-off meeting.
Anna Kourouniotis: What did your roadmap look like?
Carmen McKenzie: The very first step on our roadmap was to design the program by determining what data the committee would govern and which data governance area would be given the primary focus. Having a federated system of 34 distinct colleges, you can just imagine the vast amount of data that is created. It was decided to break the plan into two phases.
During the first phase, the committee would govern the data housed within the system’s data warehouse which is used for state and federal reporting as well as allocating funding from our state’s legislature to the individual college districts.
In preparation for the implementation of the new ERP system, the committee would focus their efforts during phase one on the standardization of data definitions. This focus area was very important because we were moving from individual application instances into a single instance for all colleges. Also part of phase one was to develop a business process that included a standard method for documenting decisions as well as a communication strategy.
Phase two would expand the data spectrum of the committee to include the ERP transactional data in addition to the data warehouse. This phase would also include an expanded focus of measuring the quality of the data. We are just now entering into phase two. We currently have three colleges on the new ERP system and will be migrating the rest of the colleges onto the new system over the course of the next four years.
Up next on the roadmap is to determine how to measure and monitor the quality of the data as well as the data governance program itself. This year we are undergoing an evaluation of the program from our sponsoring commission.
Anna Kourouniotis: How long did it take you to get your data governance program to a good place?
Carmen McKenzie: It took approximately two years from the formation of the task force until the committee was producing proposals to standardize data elements.
Anna Kourouniotis: How is the structure of your data governance team? How did you staff the team?
Carmen McKenzie: We have two co-chairs one from the State Board and one from our sponsoring commission. We then have two representatives from each of our seven commissions. These representatives are tasked with approving or disapproving proposals that are created by sub-committees we call Data Domain Stewards. The majority of work researching and determining a consensus based solution as well as developing the proposal is actually done by the Domain Stewards. I staff both the committee and most of the domain steward sub-committees.
Anna Kourouniotis: What are some tips or lessons learnt that you can provide to other institutions who are thinking about establishing data governance?
Carmen McKenzie: Gaining executive sponsorship and being given the authority from the executive level to make decisions concerning data was probably the most important factor to the success of the committee.
One other tip is that it’s very important to assign a tenacious leader to the committee. I would also say that a lesson learned was that it takes time to develop proposals with cross-functional buy-in in such a large system. Data governance is not something that occurs quickly, not if the desired output is a well formed solution with cross-functional input.
Anna Kourouniotis: What technologies did you implement?
Carmen McKenzie: We have not purchased any new technologies related to data governance. We have developed a metadata repository application to support data governance related to our new ERP (PeopleSoft Campus Solutions, Human Capital Management and Finance applications). We have also looked at a few data quality measurement products, but at this point are in the process of developing our own metrics.
Anna Kourouniotis: How do you measure the impact that data governance had?
Carmen McKenzie: That is a very good question! We are currently undergoing an evaluation from our sponsoring commission which will hopefully provide a measurement of impact.
Anna Kourouniotis: What are your governance principles?
Carmen McKenzie: Our guiding principles listed in our charter are:
- Keep it simple. Simplicity makes data easier to track, validate and to find coding issues.
- Consider the effects of the decision to modify criteria from multiple perspectives including the student, college, district, SBCTC, and national viewpoints.
- Understand the purpose – Is the data used to report externally, for state level analysis, college level analysis or funding decisions? Is it used nationally?
- Implementation timeline – Consider how the timing of changes impact how the data is used and viewed. Implement changes beginning in summer quarter whenever possible.
- Consider the effects of the change on trend analysis and historical reporting.
- All recommendations and decisions should be fully documented to include the intent, purpose, definition and impact. They should also be accompanied by a diagnostic evaluation of why there are issues with the current definition, how the data were represented under that definition, and how the data would look under the new definition.
- Whenever possible, pretest any coding change with targeted populations.
- Consider funding and policy impacts.
- Consider future needs, not just current or past needs.
- Align with national approaches and best practices whenever possible.
- Consider any unintended consequences.
Anna Kourouniotis: How did you sustain your governance efforts over multiple years?
Carmen McKenzie: One factor that helped us sustain is that our college system has become very data driven over the past several years. Consistent leadership of the committee is also a factor. Tenacity is very important as it is easy to get distracted by daily tasks, so keeping data governance at the forefront is very important. Probably the most important factor is focusing on the data that impacts state and federal reporting and the funds allocated to the colleges. If there is money involved, there will be interest in the topic.