Enterprises are taking Data Governance—planning and programming to control and secure data—more seriously. In a recent DATAVERSITY® Trends in Data Management Report, 76 percent of organizations have a Data Governance program in place or such a plan for the future. This makes sense as companies focus more on security and business growth. Data breaches have increased. Ransomware families have grown 700 percent since 2016 and attacks are up 365 percent from last year. To compete, companies rely on good Data Governance working with Data Management.
However, firms face a big challenge carrying out a cohesive Data Governance framework where its reputation and results differ widely across the organization. Andy Hayler, CEO of the analyst firm The Information Difference, stated in the Data Governance Trends article, “the implementation of Data Governance is very patchy.”
Participants in the Trends in Data Management Report also back up that Data Governance programs vary, depending on the business area, with older technologies tending to have more Governance maturity. Lack of Data Governance traction, according to Stan Christiaens, a Forbes Technology Council Member and Founder and CTO of Collibra can be attributed to “images of police and bureaucracy.” But as Data Quality deteriorates and frustration mounts over the course of some projects, Data Governance becomes very attractive and ends up implemented as project-based solutions.
This haphazard Data Governance execution carries risks across organizations, especially as they plan to release new technologies. For example, over 80 percent of firms plan on using digital twins and deep learning by the next year or two, requiring intensive computing resources. Also, business have and will continue to shift application software into cloud-based alternatives. As data accumulation, storage, usage, and transfer increase in the cloud or hybrid cloud solutions, risks of data breaches and misinformation also rise, through additional insecure access points and bad data integration.
A cohesive Data Governance, based on a Data Strategy adopted across the board, reduces the danger of compromising data safety and sets up organizations to successfully leverage newer applications. Having a successful, united Data Governance depends on reconnecting Data Governance through Data Strategy, branding good Data Governance, and getting executive support.
Connecting Data Governance through Data Strategy
Just because a company has a Data Governance framework it used with a mature technology project, like a data warehouse, does not mean it is sufficient for newer technology initiatives, like machine learning. New business requirements need to be considered, especially where system integration is necessary. For example, Data Quality must be good for all data sets, across the entire enterprise, before machine learning can be applied to a new venture.
Danette McGilvray, President and Principal at Granite Falls Consulting, said, “The cold brutal reality is that the data is not good enough to support machine learning in practically every company.” This is only one of many business needs that crop up before succeeding at such an undertaking. Revisiting Data Governance prior to starting a new data project reduces exposure to mistakenly overlooking prerequisites, and moves toward a unified Data Management approach.
Rethinking older Data Governance plans alone does not necessarily lead to a more coherent Data Governance. According to Donna Burbank of Global Data Strategy, Ltd., good Data Governance results from aligning business strategy and Data Strategy. A Data Strategy assesses how well an existing Data Governance plan supports the business as a whole. The Data Strategy guides responses and activities toward building the Data Governance framework, such as using Burbank’s matrix, shown below:
Companies are already taking the lead in connecting Data Strategy to Data Governance. CEMEX, a global leader in building materials, linked Data Strategy with Data Governance to become more data-driven and to better take advantage of newer technologies. Through understanding Data Strategy provided by a fully integrated platform, Gerardo Reyna, CEMEX Data Architecture Manager, saw where Data Quality hampered integration and began to see how the data architecture needed to evolve. Upon gaining better Data Strategy insight, CEMEX, as Burbank said, “built out collaborative Data Governance, [with] small and agile-based digital teams to prioritize projects and align different internal customers toward a common goal.” As a result of joining Data Strategy through Data Governance, Data Governance has become more unified across the enterprise. The foundation for new data project success is in place.
Branding Good Data Governance Across an Organization
As mentioned earlier, many people across the enterprise have negative perceptions about Data Governance, based on bad experience and wanting to explore or move forward without hassle or make-work on a project. This makes implementing data projects with newer technologies challenging.
Overcoming Data Governance’s reputation can prove difficult but is necessary to receive company-wide support. If an organization gets behind Data Governance, it would find it easier and more natural to consider at the beginning of a new data project, rather than when Data Quality problems emerge months down the road. Several organizations have experienced success in positively reinforcing Data Governance benefits.
According to Jason Simon and Dan Hubbard—who shared in the 2019 CIO Award for innovation in technology given to the University of North Texas (UNT)—diagrams and visuals have power in explaining Data Governance to key stakeholders. Simon and Hubbard also said, “Be sure to have a consistent look and feel for all aspects of your program, including communications in all forms and training announcements and documents.”
The UNT approach of spreading Data Governance’s message also worked for Freddie Mac, a home mortgage financer that struggled four or five times to unite Data Governance. Jenny Schultz, a Director of Data Governance at Freddie Mac, led a Data Governance branding initiative after ensuring the framework had credibility of “peers, stakeholders, and thought leaders in the industry.”
A Data Governance positive image was fostered through branding. She created a simple message communicating Freddie Mac’s Data Governance: empowerment, simplification, reuse, and data control. Early on, Jenny Schultz and her team validated Data Governance activities to executives and business partners by focusing on business needs and publicizing successes along the way. In part due to branding a Data Governance plan that worked for its people, Freddie Mac had a high level of buy-in to that model, with more engagement across the company, resulting in a cohesive Data Governance model.
Getting Executive Buy-in to a Data Governance Framework
In addition to branding a good Data Governance, getting executive buy-in and authority to make data-driven decisions makes for Data Governance unity across the board, especially at the beginning any big data project. Carmen McKenzie, Data Services Director of the Washington State Board for Community and Technical Colleges (SBCTC), found this out through setting up a Data Governance structure for 34 community and technical institutions. Legislators mandated a formalized Data Governance, “standardizing data definitions and improving Data Quality.” Concurrently, colleges required establishing Data Governance as recommended by some knowledgeable consultants before rolling out a new ERP system.
The combination of establishing a Data Governance and planning for the new ERP system happened in a short time frame, making it challenging for McKenzie and her people to decide which Data Governance framework would fit best across all the community college entities. McKenzie and her committee chose a Non-Invasive Data Governance approach. Seiner’s blueprint introduces Data Governance by leveraging best practices and structures within the business, instead of disrupting employees or forcing them to comply with a new program.
McKenzie’s task force encompassed Data Governance planning within existing college councils. This approach gained support quickly from the colleges’ presidents. The SBTC has developed a metadata repository algorithm to support Data Governance related to the ERP and is holding back purchasing any technology. This will allow Data Governance to spend more time developing Data Quality metrics.
Cohesive Data Governance needs to come first, before pulling the trigger to start a data project or using new technology. With a new venture, there are just too many opportunities from data breaches and bad data from poor Data Quality with a patchwork Data Governance framework.
The point at which a company decides to implement any project with new technology, it should consider connecting Data Strategy with Data Governance, branding Good Data governance that has support of people and buy-in from top management. Changing business requirements, usually including a greater need for integration, demands no less.
Image used under license from Shutterstock.com