Data Governance

Data governance decides who is responsible for an organization's data and how it should be managed. It sets rules for how data is defined, collected, stored, shared, protected and retired, and assigns people to make decisions and resolve problems. Without governance, the same term can mean different things in different departments, quality problems go unfixed because no one owns them, and sensitive data spreads into systems where it should not be.

Interest in data governance has grown with privacy regulations, the volume of data organizations hold and the use of analytics and artificial intelligence, which depend on data that is trustworthy and properly controlled. For enterprise architecture and IT finance, data governance protects the value of data investments, reduces regulatory risk and makes it possible to use data confidently across the business rather than within individual silos.

Definition of Data Governance

Data governance is the exercise of authority and control over the management of data. It defines decision rights, accountability, policies and standards, and the processes that ensure they are followed.

It is distinct from data management, which covers the day-to-day work of handling data. Governance sets the rules and responsibilities, and data management carries them out. Practices such as master data management and data catalogs are common ways of putting governance into practice.

Key Roles

A data governance council or committee, made up of senior business and technology leaders, sets priorities, approves policies and resolves disputes.

Data owners are senior business leaders accountable for specific data domains, such as customer or financial data. They decide who can use the data and what quality is required.

Data stewards handle day-to-day governance within a domain, maintaining definitions, monitoring quality and resolving issues.

Many organizations also appoint a chief data officer to lead governance and data strategy across the organization.

Main Areas of Policy

Data definitions and standards ensure that common terms, such as customer or revenue, mean the same thing everywhere they are used.

Data quality policies set expectations for accuracy, completeness and timeliness, along with processes for measuring and fixing problems.

Security, privacy and retention policies control who can access data, how sensitive data is protected and how long data is kept. Lineage requirements ensure that the origin of important data can be traced, often using data lineage tools.

Importance of Data Governance in Enterprise Architecture

Data governance makes data reliable enough to support decisions. Reports, analytics and AI models are only as trustworthy as the data they use.

It reduces regulatory and security risk. Privacy laws and industry regulations require organizations to know what personal and sensitive data they hold, where it is stored and who can access it.

It also reduces waste. Clear definitions and ownership prevent teams from building duplicate datasets and reconciling conflicting figures. Analysts in many organizations spend a large share of their time finding, checking and correcting data before they can use it, and governance reduces that effort.

Connecting Governance and Architecture

Data architecture defines where data is created, stored and moved, while governance defines who is responsible for it and what rules apply. The two need to be designed together.

Enterprise architects use governance decisions to choose systems of record for each data domain, design integrations and set standards for new applications.

Governance requirements should be part of architecture reviews, so that new systems are designed with ownership, quality and security in mind rather than having them added later.

Linking data domains to the applications that hold them shows where governance effort is most needed. The application inventory can record which data domains each application holds, along with their sensitivity, so that owners and stewards know which systems fall within their scope.

Common Challenges

Governance can become a bureaucracy that slows work without improving data. Programs that focus on policies rather than outcomes often lose support.

Business leaders may see data ownership as an IT responsibility, leaving stewards without the authority they need. Data quality depends on how data is entered and used in business processes, so ownership has to sit with the business to be effective.

Programs that try to govern all data at once spread their effort too thinly. Starting with the domains that matter most to the business usually works better. Early results in those domains also build the case for extending governance further.