Understanding Data Governance and Stewardship, Part 3
David wishes to thank Anne Marie Smith, director of Education at EWSolutions, for her invaluable contribution to this month's column.
Because each organization is unique, one could argue that there are as many data governance methods as there are organizations implementing this process. Nevertheless, certain points are universal in governance and can be used to examine and scrutinize any methodology/process.
Sponsorship and commitment to the goals of data governance and the processes to implement it should be foremost. Lack of senior management commitment has killed many governance programs and doomed many data stewards. Analyzing the organization's culture is also very important. Can data governance become part of this organization's culture? Will standards be applied equally across the organization? Does this organization have an enterprise view of data that would foster the development and maintenance of a data governance program?
Technology provides the way to manage and deliver data for the users and consumers. Therefore, choosing the right set of products is essential for a successful data governance program. This includes databases, enterprise application integration (EAI) and extract, transform and load (ETL) tools, metadata management and master data management products, among others. The abundance of technical options means that an organization must also develop a plan to choose the right ones for their needs and must plan to integrate these tools with the processes that they serve to support data governance objectives. Establishing a managed metadata environment (MME) can help an organization track, monitor and record changes to its data and documents. One of the most vital functions of any MME is to provide the technical architecture and processes to manage corporate knowledge (including governance).
Process development is the core focus for data governance, in conjunction with implementing the technology of an MME. Processes for governance include coordinating, managing and monitoring the development and use of organizational audit and control procedures and data standards and policies. One critical area involves managing metadata standards, because metadata allows the organization to understand its data sources, definitions, uses and relevance. All data standards and policies should be controlled within the data governance program, so policies are consistent and changes can be sent throughout the organization when necessary, using the technology of the MME to facilitate this coordination and communication.
Two groups with responsibility for an organization's data governance program are the data governance council and the data stewardship team. A data governance council consists of management representatives from business and IT along with the data stewards. They:
- Coordinate and direct data governance strategies and processes across the enterprise;
- Ensure that data governance strategies and processes support the organization's mission and objectives;
- Develop and direct data standards across the organization and within projects;
- Assign roles, responsibilities and authority and implement governance through a number of organizational layers;
- Provide mechanisms for coordination, communications, information sharing, prioritization and conflict resolution within the organization and across projects; and
- Provide accountability for the successful implementation of all governance efforts, whether at the enterprise level or within lower organizational levels (division, group, project).
Data stewards are the glue that holds a governance program together. Stewards define subject area boundaries in conjunction with IT and business leaders; collect feedback and enhancements for specific subject areas; resolve data integration issues; act as the conduit between business and IT; and serve as quality control point for the subject area.
In the context of their defined subject area domains, data stewards define/describe business data elements; define data domain values; establish and validate data quality rules; identify and help resolve data quality issues; help develop data domain business rules (algorithms, calculations, processing requirements); and define security requirements.
Data stewards for each functional area should be identified and given training in the basics of data and metadata management. The governance council sets the standards, policies and procedures for the data and metadata that is required for their organization, and stewards are the enablers of the data governance council and the executors of the standards and policies that the governance council develops. The data governance council should be collaborative in all information systems application activities (package evaluation and implementation, application development and maintenance) to manage the impact the activity will have on the integrity of the organization's data under their stewardship.
In the final analysis, data governance is a complex topic and should be examined carefully before developing your organization's program. Having the services of an experienced group of information management professionals can make the difference between a successful data governance program and a failed one. Success can be measured by how well you can answer the questions posed here, and if your organization understands its data and values the role that data can play in achieving and maintaining a competitive advantage.