Data Governance Blog

How to Govern Data – A Simple Method to Implement Governance

Some time ago, while working with a large data warehouse, I formulated a simple method to govern data. The method was introduced as part of our Basel II work and was required by our Basel II Risk Management Policy. Subsequently as we were validating our plans for a data governance program, it was reviewed and subsequently endorsed by Gwen Thomas of the Data Governance Institute. She loved the idea so much that she presented it at the 2008 DAMA conference. This simple method is a Data Governance Agreement.

The agreement picks up where Service Level Agreements (SLA) leave off. Most SLAs establish the expectations for hardware, operating systems, networks, and software applications. Normally, they focus on availability, issue resolution, and responsibility. What they do not address is how do you govern the content and link that governance process to corporate policy.

That is where a DGA comes into play. In a very simple document, the responsibilities for data content, data quality, and issue resolution are spelled out. The document is applied to each major data movement within an organization that involves a change in custody. The scope of the document is to focus on the quality and availability of data content.

In the document; data stewards are identified, processes for resolving data quality issues are delineated, and responsiveness is specified. Beyond that, each organization can extend the document as they see fit to meet their needs. Most importantly from an IT perspective, the document specifies the role of business in maintaining data quality and manages the expectations placed on IT to assist them.