Metadata Management
Define and update metadata that indicates the information life cycle stage and access control criteria for both business and technical data.
Improvement Planning
Practices-Outcomes-Metrics (POM)
Representative POMs are described for Metadata Management at each level of maturity.
- 1Initial
- Practices
- Manually acquire meta data for specific purposes.
- Engage and allow subject matter experts to use meta data where necessary.
- Outcomes
- Metadata is mostly not present.
- Data structures may not meet the immediate need, and are inefficient, inconsistent and lack the ability to cater for future requirements.
- Basic data structures may be inefficient or facilitate duplication of data without the means to synchronize updates.
- Metrics
- % of data sources for which unstructured meta data acquisition is manual.
- % of data sources for which meta data exists.
- % of KPIs unanimously defined.
- 2Basic
- Practices
- Gather meta data based on defined schema, structures or specific data dictionaries.
- Develop application specific data dictionaries and widely apply them.
- Outcomes
- Data structures meet the immediate need and standard approaches to acquisition are taken.
- Quality improvements and better use of data is evident in well documented areas.
- Metrics
- % of data sources for which un/structured meta data acquisition is manual.
- % of data sources for which meta data based on a dictionary schema exists.
- 3Intermediate
- Practices
- Augment meta data acquisition by automated tools where possible.
- Rationalize metadata into a single master meta data repository for core business functions.
- Develop expertise on meta data and data modelling across the business and ensure all work together.
- Outcomes
- Data structure containers can be populated automatically in some cases.
- Business and technical meta data models enable better use of higher quality data for both IT and business stakeholders.
- Expertise is developing by both business and IT.
- Metrics
- % of data sources for which un/structured meta data acquisition is partly-automated.
- % of data sources for which meta data has been standardised and moved into a central master repository.
- 4Advanced
- Practices
- Extensively automate meta data acquisition.
- Agree and ensure the extensibility of meta data architectures.
- Map meta data to business processes for context.
- Identify dependencies between meta data.
- Outcomes
- Data structure containers can be populated automatically in most cases.
- Rich descriptive meta data enables data exchanges both internal and external to the organization that can be quickly and efficiently put together.
- Metrics
- % of data sources for which un/structured meta data acquisition is fully-automated.
- % of data sources for which meta data, and their inter-dependencies are mapped to the business processes
- 5Optimized
- Practices
- Fully automate meta data acquisition.
- Develop data acquisition tools to automatically map meta data back to business processes and to identify meta data dependencies.
- Use metadata to model the impact of proposed changes.
- Outcomes
- The risks from change are managed in terms of understanding the full impact of changes and the implementation and deployment requirements.
- Subject mater experts are relied on less and documented meta data is the source of knowledge about data.
- Metrics
- % of data sources for which automated, structured meta data acquisition is mapped back to business processes.
- % of data sources for which meta data can be wholly relied upon to inform change control management