“Information Governance is the setting of objectives to achieve valuable outcomes by people using information assets in a process that considers both risk and time constraints.”
Information Governance (IG) must have Process. The Process must consider IG Objectives, Outcomes, People and Assets. Theses are the critical first five dimensions of the Seven Dimensional IG ramework (7DIG).
Again, we can use the magic number seven to subdivide the Process dimension. Seven transitive verbs, Acquire, Validate, Store, Protect, Update, Publish and Dispose cover most governance operations. It is a continuous life cycle, capable of being monitored and controlled at every stage:
- Acquire: Data acquisition, in the 7DIG Framework, includes action around analysis of the information assets and data modelling. The context of data collection is vital to onward processing and re-use in further processes. Quarkside subscribes to the concept of Master Data, Operational Data and Derived data. Master data is relatively static reference data. All data has to be acquired and subjected to further governance processes.
- Validate: Incorrect data causes inefficiency, often accounting for 80% of administrative effort on systems; but far worse is the impact of poor information on decision making and information sharing. Good governance requires metadata and the use of standards to be embedded in the culture. Validation implies comparison of input data with a standard that is enshrined in metadata. Even paper-based publications are subject to validation against standards of grammar and probity before they are published.
- Store: Imagine data stores as silos. Individual grains of data are added until required for further processing. Vast volumes of operational data are stored for subsequent processing. Documents are added to filing cabinets or archive shelves. In the best regulated environments there are custodians who know where the data is and how to retrieve it.
- Protect: Data protection and security of access is an industry in itself. It makes sense to protect any valuable asset and information is no exception. Identity management, likewise, is an all pervading topic. It has to cover the identity of the data subject and the identity of the data investigator. For information sharing between agencies, accurate data matching depends on the quality of subject identities.
- Update: Over the course of time, there are changes to facts and figures. Records need to be retrieved and modified. There are elements of feedback to make corrections as a result of performance monitoring and analytical processing. Derived data can be added to the data stores.
- Publish: Data should be published only to those who are entitled to use it or see it. This could even be open data provided to the general public, such as the £500 contracts with local government.
- Dispose: Oft forgot is the need to delete data. The DPA requires that data should be held only as long as necessary. Whereas this is probably true of major corporate systems, this governance step may not always extend to private document files, emails and spreadsheets. Many operational documents can be safely shredded in less than ten years; others, such as children in care records, have a statutory limit of 125 years. The key to a good disposal policy is the Information Asset Register, wherein the metadata should include the disposal policy. Transfer to the National Archive should also feature as a category when documents may have historical importance when not required by a local authority or other public body.
Governance matters, but it cannot be a universal set of rules. Neither do frameworks guarantee good governance. Frameworks can only provide simple diagrams and checklists, they cannot provide the thinking or knowledge needed in any specific context.
So why bother promoting a framework? The justification is that that requirements are so diverse that a team is needed to cover all aspects at sufficient depth. People need at least an overview of some specialist issue. Hence the importance of a MECE approach that does not drill into the detail, but tries to cover all important topics (aka dimensions).