Metadata Governance is easiest to understand when you separate the term into its two parts – Metadata and Data Governance. Ask any organization that excels in metadata management (or provides thorough documentation of their data, information, and records) whether or not they govern their metadata and they will surely respond affirmatively. These organizations make certain that people are formally accountable for the metadata – because it is a known fact that the metadata will not govern itself.
“The metadata will not govern itself.”
Make Your Point with Analogies
Start with analogies about what it takes to manage anything. A person cannot manage (let’s say “govern”) their finances without having quality information about their finances. A librarian cannot manage the library’s resources without quality information about the resources. A RIM Manager (“Records and Information Management”) cannot govern the organization’s records without quality information about the records. It seems to make sense that, to successfully manage and govern anything, you require quality information about that “thing”.
“To govern any ‘Thing’ requires that you
have information about that ‘Thing’.”
Further Your Point with Resources
Continue with resources that are required to manage anything. A person cannot govern their finances unless they (or somebody) are accountable for managing their finances. A librarian cannot govern their resources unless they (or somebody) are accountable for managing the resources. Your RIM Manager cannot govern the records unless they (or somebody) are formally accountable for the governance of the records. To successfully manage and govern anything requires that somebody is formally accountable for the governance of that “thing”.
“To govern any ‘Thing’ requires that you have somebody
that is formally accountable for the management of that ‘Thing’.”
Let’s start with the separate definitions:
- Metadata are the facts about the data stored in IT tools that improve the business and technical understanding, and therefore value, of that data.
- Data Governance is, first and foremost, the discipline of enforcing authority associated with the effective management of data.
Merge these definitions together and Metadata Governance is the discipline of providing effective metadata resources. Metadata Governance is, basically, the effective governance of the metadata. Are you still with me?
I heard it argued recently that metadata governance is just as important, if not more important, than plain old Data Governance. Data Governance practitioners will be the first to tell you the importance of thorough data documentation to improving the value of our most critical data assets.
With all this being said, the truth is that I rarely see organizations that have a “Metadata Governance Program”. Data Governance Programs are difficult to sell to management and are already hard enough to ascertain the resources and funding that are required to be successful. Good luck selling your management on the need for a Metadata Governance Program. Instead I suggest that you convince people that metadata is a valuable resource and that the governance of data will depend on having quality documentation about the data, in the form of metadata. So … NO to Metadata Governance Programs but YES to the need to govern the metadata. Remember …
“It’s worth repeating that the metadata will not govern itself.”
Hopefully at this point you have questions concerning the basics for governing your metadata. Simply put, the basics for governing your metadata are very similar to the basics for governing your data. To demonstrate this thought, I will use the five core components that I describe in my Non-Invasive Data Governance Framework that I describe beginning in this article.
The five core components of successful Data Governance (and Metadata Governance) are:
- Roles (Authority)
- Processes
- Communications
- Metrics
- Tools
Roles (Authority)
The first foundational component of a successful Data Governance program (and thus the governance of metadata) is the definition of roles and responsibilities. The manner in which roles are defined is a predictor of the effort required to govern the data or metadata. Assignment into roles often presents push-back when the effort is over-and-above existing responsibilities. Identification into roles encounters less push-back as people see themselves in the roles that they have been slotted. The recognition of people into roles is a direct manner of acknowledgement of the part each person plays in the program.
In NIDG (and now NIMG) – Roles are typically represented through an Operating Model of Roles & Responsibilities. The familiar pyramid diagram is represented in the first column of the NIDG Framework. The Operating Model and accompanying artifacts provided as part of the approach include a detailed description of formalized responsibilities, escalation and decision paths, how roles are formally engaged in processes, and communications that are shared with each level.
As stated earlier, the metadata will not define, produce, and use itself. Therefore, it is important that we define the roles and associated responsibilities of the people that will be held formally accountable for the governance of the metadata. Think of it as formal responsibility for defining the metadata that will be managed, as well as the formal responsibility for producing the metadata and hopefully the people that are formally responsible for using the metadata to assist them in their job functions.
Processes
The second foundational component of successful data governance is the way the Roles are applied to processes. The notion of the “singular governance process” misrepresents the fact that processes are a primary component of governance success. There is not a single metadata process that is governed; rather there are a series of processes to which governance will be applied.
In NIDG – Data is governed through repeatable processes that reflect the appropriate level of formal accountability throughout the process. The same must hold true for the data documentation and metadata if the organization is going to improve their knowledge and confidence in their data through the metadata. Data Governance focuses on getting the “right” person involved in the “right” step of the process to deliver the “right” result regardless of the process focus – issue resolution, protection, quality, project-focused. Metadata Governance becomes the application of formal governance to metadata processes.
Communications
Communication is an important component of a successful governance efforts. Raising the data and metadata awareness of every person that defines, produces, and uses data (and metadata) is critical to achieving successful governance. Education must focus on policies, handling rules, best practices, standards, processes, and role-based governance activities.
In NIDG – Communications play a role in every aspect of governance definition and delivery. Communications must be thorough and measurable. Communications must focus on formalizing accountability for the data and metadata processes mentioned above: issue resolution, protection, quality, project-focused, or any other application of authority to how data or metadata is managed.
The Communications Plan must mirror the Roles component described above. Communications must include orientation, on-boarding, and ongoing subjects focused toward the specific audience utilizing available communication instruments.
Metrics
Governance efforts must be able to measure their impact on the organization. This is the responsibility of the Support role often called the Data Governance Administrator or DG Lead and Team. The impact and value of governing your data and metadata may be financially quantifiable, but this may not always be the case. Measuring efficiency and effectiveness improvements require benchmarks of present state and the governed activity of measuring and reporting results.
In NIDG, organizations measure improvements in governance through collecting and reporting the number of issues recorded and addressed, while also quantifying the value of issue resolution, quantifying education, awareness, and certification of handling rules and incidents that are formally attended.
Data and Metadata Governance metrics and measurements must be auditable and demonstrable to management and authorities when requested. Organizations typically count the reusability and understandability of data definition through the metadata, the ability and speed to access the “right” data at the “right” time, the production of high quality data, and the proper usage and handling of data (which can all be found in the metadata).
Tools
Tools of governance enable the efforts to deliver value to the organization. Organizations use tools they develop internally as well as tools that they’ve purchased to fill specific needs of their governance efforts. The tools that are developed or purchased are based on practicality, ease-of-use, and the specific goals of the Data Governance program or the governance of your metadata.
In NIDG – Tools are used to formalize accountability for the management of data and metadata and improve the knowledge of the data rules and processes required to govern data. Tools are used to record and make available metadata in order to improve the understanding and quality of data across the Enterprise.
The Data Governance Tool market is growing as the definition of Data Governance expands to address authority enforcement over big data, smart data, metadata, and all data used for analytics. Prior to investing in new technologies, organizations should clearly state their requirements, consider leveraging existing tools and develop tools internally to address specific metadata needs of their Data Governance program. Metadata Governance tools and Data Governance tools tend to share a lot of capabilities.
“One more time remember that the metadata will not govern itself.”
Conclusion
Simply put, metadata governance is the formal application of the discipline of data governance for the data about data – in other words … the metadata. If you take one thing away from this article, consider these two things 1) Metadata is very important to Data Governance success and 2) the metadata will not govern itself.
* Every thesaurus I checked recognized the term “govern” as an acronym for the verb “manage”.
