As a second line of defense data governance leader, a critical responsibility is to author, publish, and implement effective data governance standards. Perhaps the most difficult part of this task is finding a reference point or foundational framework on which to base the standards. In my experience, this is due mostly to the dearth of publicly available source material or examples. Also, if you’re anything like me, I worry constantly that what I produce is inadequate and lacking in some critical aspect. In this article, I will lean into that insecurity by leveraging Cunningham’s Law to solicit your harsh and withering feedback on a proposed metadata management standard.
If you’re not familiar with Cunningham’s Law, pause now to Google it because it’s amusing and at the very least you can say reading this wasn’t a complete waste of your time. The Law simply states that the best way to get the right answer on the internet is not to ask a question; it's to post the wrong answer. Accordingly, the following is what I will assert is the best enterprise metadata management standard ever written. Prove me wrong!
Standard Title Metadata
Management
1. Standard Statements
The enterprise shall design, implement, and maintain metadata management processes, procedures, and control activities to ensure high-quality metadata. The processes must determine how metadata will be created, maintained, integrated, and accessed and include the following elements:
a. Data Stewardship: Establish roles, responsibilities, and accountability for managing metadata across the organization.
b. Metadata Governance: Develop and enforce policies, standards, and guidelines for metadata creation, maintenance, and usage.
c. Metadata Strategy: Define a comprehensive strategy to ensure metadata supports organizational goals and aligns with business needs.
d. Metadata Quality: Implement processes to validate, cleanse, and enhance metadata to ensure accuracy, completeness, and consistency.
e. Metadata Lifecycle: Establish procedures to capture, store, update, and retire metadata throughout its lifecycle.
f. Metadata Integration: Integrate metadata across different systems, applications, and platforms to ensure interoperability and data consistency.
g. Metadata Security: Safeguard metadata from unauthorized access, modification, or deletion through appropriate access controls and encryption measures.
h. Metadata Discoverability: Enable effective search, discovery, and retrieval of metadata assets to support data governance and decision-making processes.
i. Metadata Documentation: Document metadata definitions, lineage, relationships, and other relevant information to promote understanding and data lineage.
j. Metadata Change Management: Implement change management processes to track and manage metadata changes, ensuring proper communication and impact analysis.
But wait, there’s more! Everyone knows that nothing goes better with standards than key performance indicators (KPIs), as there is inevitably someone around to remind you that what gets measured gets managed. So as a bonus, here is a selection of KPIs that one might use to monitor and enforce performance against the above standard.
• Metadata Completeness: Measure the percentage of metadata elements or attributes that are populated or documented accurately in relation to the total expected metadata elements.
• Metadata Accuracy: Assess the accuracy of metadata by comparing it with the actual data it represents, measuring the percentage of metadata elements that are consistent and reflect the true state of the data.
• Metadata Consistency: Track the consistency of metadata across different systems, databases, or data sources, ensuring that metadata elements are aligned and coherent.
• Metadata Relevance: Evaluate the relevance of metadata by assessing its usefulness in supporting business processes, data analytics, reporting, or decision-making activities.
• Metadata Timeliness: Monitor the timeliness of metadata updates, measuring the time taken to capture and propagate changes to metadata elements after data updates or system modifications.
Deeper Dive! How could we measure metadata timeliness?
• Metadata Update Lag Time: Measure the time taken to update metadata after a corresponding change occurs in the underlying data source. This can be calculated by tracking the timestamp of the data change and comparing it to the timestamp when the metadata is updated. The goal is to minimize the lag time between data updates and metadata updates.
• Metadata Refresh Frequency: Determine how often metadata is refreshed or synchronized with the underlying data sources. This can be measured by tracking the intervals between consecutive metadata updates. For example, if metadata is refreshed daily, the goal would be to ensure that metadata accurately reflects the current state of the data within a 24-hour window.
• Metadata Latency for New Data: Assess the time it takes for metadata to be created or captured for newly added data. This can be measured by comparing the timestamp when new data is introduced to the system with the timestamp when corresponding metadata is created. The objective is to minimize the delay in capturing metadata for new data, ensuring it is available and accessible in a timely manner.
• Metadata Usage: Analyze the utilization of metadata by tracking the number of times metadata is accessed, queried, or referenced within the organization, indicating its value and adoption.
• Metadata Compliance: Measure the organization's adherence to metadata governance policies, standards, and guidelines, ensuring that metadata is managed according to established rules and best practices.
• Metadata Integration: Assess the level of metadata integration across systems, applications, or platforms, measuring the degree of interoperability and consistency achieved.
• Metadata Impact: Evaluate the impact of metadata on data quality, data governance, data lineage, or other key aspects of data management, assessing its effectiveness in improving overall data practices.
• Metadata Maturity: Gauge the maturity of metadata management processes and capabilities within the organization, using maturity models or frameworks to identify areas of improvement and track progress over time.
So here is your chance, light your torches and sharpen your pitchforks. Tell me why this isn’t the greatest enterprise metadata management standard you have ever laid eyes on.
