Three Approaches to Data Governance
There are three distinct approaches to implementing data governance in an organization:
Command and Control
The first approach follows the notion that people must be assigned to be data stewards, immediately making the role feel like it is over-and-above the activities already called for in a person’s job description. But even still, people are told that they must govern the data, no matter what it takes.
The second approach can be compared to the movie “Field of Dreams” where the most memorable line is, “If you build it, he will come”. Organizations that follow the traditional approach build up their roles, processes, and data management style hoping that people in the organization will gravitate toward the roles and processes, but without a level of assurance that the people will adopt a governed environment. People are basically told that they should govern data better than they do already.
The third approach is built on the premise that people are already governing data, but they are governing data in an informal manner, leading to inefficiencies and ineffectiveness in the way data is managed. In the Non-Invasive Data Governance approach, people are shown that they already govern the data informally and are directed to the value they will receive from a more formal level of governance.
People naturally rebel against the idea of being governed. Therefore, the selection of the appropriate approach to governing your data is so important. Data governance is known in some circles as “People Governance” because it is people’s behavior – how they define, produce and use data – that is being governed. In other words, the data will do what we tell it to do, so we must govern people’s behavior if we want to improve the quality, value, and understanding of the data. Therefore, the approach the organization takes to govern the data (and the people) can make or break whether the data governance program is accepted or rejected by the organization.
I have been known to say that “the data will not govern itself.” Let me add to that with, “the documentation about the data, or the metadata, will not govern itself either.” Most of us have experienced data and metadata that has been left ungoverned. Why? Because people are not held responsible for the quality and/or value of the data or the documentation. As a result, there is no way to improve the efficiency and effectiveness of the way data assets are being leveraged.
Formalizing Data Governance
The discipline of data governance must focus on knowing who these people are, helping them to make more actionable decisions, and empowering them to become better stewards. People who define data must know what it means to define data better, and that includes providing meaningful business definitions for data and managing how often data is replicated across the organization. People who produce the data must know what quality data looks like, and they must be evaluated on the quality of the data they produce. And the no-brainer. People in the organization who use the data, must understand how to use it, and follow the rules associated with using it appropriately. That means data consumers must follow the protection and privacy rules, the business rules, and use the data in the ethical manner spelled out by the organization.
While people already define, produce, and use data, data governance requires that these people consistently follow the rules and standards for the action they take with that data. The rules and the standards are important metadata, data about the data, that must be recorded and made available to the people across the organization to assist in the discipline of data governance. The most effective way to manage this metadata is through a data catalog that engages users, enhances the value of the metadata, and formalizes people’s interaction with the data.
Governance With Non-Invasive Data Governance:
• Data Steward responsibilities are identified, recognized, formalized, and engaged with according to their existing responsibility rather than being assigned or handed to people as more work.
• The governance of information is applied to existing policies, standard operating procedures, practices, and methodologies, rather than being introduced or emphasized as new processes or methods.
• The governance of information augments and supports all data integration, risk management, business intelligence, and master data management activities rather than imposing inconsistent rigor to these initiatives.
• Specific attention is paid to assuring senior management’s understanding of a practical and non-threatening, yet effective, approach to governing information. This approach will be taken to mediate ownership and promote stewarding of data as a crossorganization asset, rather than the traditional method of “you will do this.”
• Best practices and key concepts of this non-threatening approach are communicated effectively, compared to existing practices, to identify, and leverage strengths and enable the ability to address opportunities to improve.
• Metadata is made actionable by formalizing responsibility of data documentation while making information about the data available to people who will benefit from its use.