Data governance is the cornerstone domain of the enterprise data management discipline. The functions of data governance are essential to any successful enterprise data management initiative
Many organizations are pursuing numerous data-oriented initiatives, such as customer integration, master data management, data warehousing and business intelligence, etc. These efforts all require a thorough understanding of the data needed to answer business questions, satisfy business needs and make it possible to pursue the goals of broader and deeper understanding of the data necessary for operations and decision-making. These needs all point to the reason many organizations are embarking on a data governance program.
Enterprise data (or information) management is the collection of processes and technologies that enable an organization to govern and administer its data assets. Enterprise data management is a function of both business and information technology and the governance of data and its usage is an important aspect of any enterprise data management effort. In fact, a good EDM approach will begin with the realization of the need for data governance, making the implementation of a data governance program one of the first activities of EDM.
Has data governance always been an essential aspect of EDM? Actually, data governance has been a cornerstone of enterprise data management since the enterprise approach to data management was first promoted, but it was buried in the various activities that comprised what is now called “enterprise data management.” It was not until organizations began to realize that data is an asset to be cared for across the organization that “data governance” became visible in its own right.
Connection to Other Data Management Functions
Data-oriented efforts only succeed when the organization has a deep and thorough understanding of its data and its context. Therefore, data management activities are necessary to give data stewards and data consumers the context and usage capabilities of the data captured, acquired and stored within the organization’s systems. “Context” is provided by metadata, which is the all the relevant information concerning the data instance. For example, the data element / attribute “Customer Name” could have metadata about the type of data (e.g., character), the length of the field, the source of the data (entered, sourced from another system, etc.), the date the data was created and the date it was last updated, etc.
If the attribute were a calculation, the metadata would include the algorithm used to calculate the value and other associated information. Data stewards need all of this metadata to determine the usefulness, validity, and other quality aspects of the data they are responsible for. Metadata management is one of the essential activities of an enterprise data management effort, and managing metadata properly and consistently is important to the success of a data governance program and to the performance of a data steward’s tasks. Data stewards and EDM specialists work together to identify, store, and make metadata available and useable.
Other pillars of EDM include master data management, data quality, enterprise data modeling and data sharing, and each of these domains uses data governance to complete its mission. Master data management (MDM) also known as Reference Data Management, focuses on the management of reference data that is shared by several disparate IT systems and business groups. MDM is required to enable consistent computing between diverse system architectures and business functions. Data governance supports the identification of the master / reference data, provides relevant metadata for it, identifies the owners and stewards of the master data and enables the organization to use this reference data across the enterprise.
Enterprise data modeling was one of the first activities promoted by the adoption of an enterprise data management approach. It is responsible for modeling all of the organization’s data from a business perspective and, historically, usually consumed resources without showing much value for the massive effort. Why? Most who study the field of EDM speculate that the enterprise data modeling efforts of the 1980’s and early 1990’s were led by IT and not the business, ignoring the fact that the business owns the data, not IT. When enterprise data modeling is initiated by data governance programs and involves data stewards (business people with responsibility for the management of data in their subject area) the data modeling efforts move smoothly toward a usable result: the identification of the enterprise-level data along with the context and relationships of that data. Many organizations start or re-start an enterprise data modeling project as one of the main activities of their data governance program to realize numerous business and IT goals.
Data quality is not a separate, single activity and cannot be seen as only part of a data governance program. Data quality must be incorporated into every aspect of a data governance program, since a prime function of data governance and data management is to improve and maintain the quality of the data. To be successful, quality must be measured continuously and the results incorporated into the enterprise data management process and into all data-oriented business activities. Data governance without a focus on data quality and the continual improvement of data quality is not true data governance, nor is it effective enterprise data management.
Perhaps “data governance” is really the recognition of the processes of enterprise data management, perhaps it is a new function. Regardless, data governance is a crucial aspect of any organization’s efforts to use their data assets wisely, and should be a major initiative for any organization that wants an effective approach to data management for its enterprise.