Introduction to data governance

Everyone in your enterprise has a responsibility in data governance processes. Data governance includes securing your data, organizing it, defining the access permissions, and determining the way your organization uses data. Data governance is vital for business intelligence and data analytics, and it should be prevalent throughout the modern analytics workflow. In this introduction, we will review the pillars of data governance and give the lowdown on why it’s essential. Then we’ll give you the first steps to implementing data governance in your business.

Data governance includes defining the roles and responsibilities of managing data and deciding how and what data you will use for your business purposes. Governing data access and organization provides for better data analysis.

Benefits of data governance

Data governance and data stewards make it easy for analysts to connect to the right data

Having a robust data governance program can empower your business and IT teams to interact with data—with both the agility the business demands and the data security IT needs. Imagine that an analyst or a team leader can find, access, and explore accurate and reliable data that they need, when they need it—confidently creating visualizations and reports to share with their teams. Every benefit of having actionable insights comes from having sound data governance.

Challenges of data governance

Data governance requires a lot of planning, decision-making, and monitoring. Clearly documenting processes for how you use data will set you up for success in using data for business decisions. When companies do not govern data correctly, security and compliance is at risk, analysis can be incorrect, and crucial IT resources are misused. Data governance is demanding because it involves teamwork, investment, and resources.

Pillars of data governance

We know data governance practices won’t look the same in every organization. Your data governance will look different according to your needs and capabilities. You may decide a centralized authority owns data or determine that data access is delegated or self-governed. Whichever type of program you develop, always ensure it’s agile and iterative. We’ve outlined different models of data governance here. These models are based on the pillars or crucial sections of data governance.

Data quality

Screenshot of a data quality warning in Tableau Server

Screenshot of a data quality warning in Tableau Server

High-quality data is accurate, complete, reliable, and relevant, and helps you serve your business. You can manage data quality by developing a checklist to review how data is ingested, edited, and stored in your enterprise. Ensure data quality close to when your systems ingest it, so the data quality is better for analysis. Data source management and metadata management are two critical practices within data quality.

Data source management

Screenshot of editing data source permissions in Tableau

Screenshot of editing data source permissions in Tableau

Managing data sources is a large part of ensuring data is accurate. Within data source management, you’ll do data mapping and consolidation. You’ll remove duplicate data or control necessary duplicate data. You’ll also decide how business teams will connect with their data. Tinuiti centralized over 100 data sources using Tableau Prep, increasing their compliance and reducing analysis time.

With some data analytics platforms, you can interact with data live as it’s on the server. A live connection requires high-speed databases or data warehouses. Or, you can extract data from the servers to your machine and interact with it there. Extraction can take longer, but it doesn’t put as much strain on your databases. Data source management also involves developing a program to refresh data so it is kept up to date.

Metadata management

Data source management also includes metadata management. An important function of metadata is to give each data source a description so business users of all technical skill levels will understand the context of the data. Business teams can see what is inside the data source and add field descriptions and comments. This metadata needs to be standardized and managed, much like a data source itself. Data stewards should decide who can create metadata descriptions and who can publish or share them across the enterprise.

Data security and compliance

An integrated data lineage makes it easy to understand exactly how and where data is being used

Data-driven organizations must balance two challenging needs: empower business users to interact with and use data while ensuring data is secure. User access can be via desktop, mobile, and web. Your organization will have to customize its data security permissions to comply with all federal, state, local, and global privacy regulations. Data security processes include defining and labeling data sources by their levels of sensitivity and creating secure access points. JPMorgan Chase has more than 30,000 licensed Tableau users in one of the highest-regulated industries and is still able to create secure connections to their data.

Data stewardship

Data governance is an enterprise-wide effort, but your company needs one or more data stewards to captain the ship. Data stewards will ensure data access, security, quality, and that the pillars of data governance are upheld. Stewards will work with IT departments, analysts, and end-users to keep the program aligned with business goals. Stewards also monitor how teams use data sources and whether teams are using the data efficiently.

Data stewardship includes monitoring and measuring the efficacy of data governance. It’s best to be proactive in scheduling when you refresh your data extracts. You should monitor how your systems ingest raw data.

Transparency

All of the policies and procedures you developed in your pillars should ladder up to a system of transparency. Data stewards and governance leaders should inform team members of how and why they govern data the way they do. The data governance team shouldn’t make all decisions behind closed doors—to get everyone to support the process, they need to understand it.

First steps to implement data governance

Implementing data governance begins with making crucial decisions or requirements for the people and policies governing your data. Data governance strategy requirements may vary according to your needs, but this is an excellent place to start.

  • Choose your data stewards or the administrators in charge of your program.
  • Define the processes and policies that work for your enterprise.
  • Choose your data governance solutions. Understand that no solution will work for your company straight out of the box; it will involve careful deployment and customization. However, Tableau is designed to work with your native architecture and is flexible as your data environment and business needs to grow and evolve.
  • Identify your data governance team that will work with your data stewards.
  • Plan your training program to get everyone on board.

Data governance best practices

Governing your enterprise data prepares your business for more robust analytics and business intelligence. We’ve reviewed the pillars, challenges, and benefits of data governance. However, this guide is just the beginning. We’ve created an in-depth Blueprint for enterprise data governance and included resources for you to get started. Don’t feel alone embarking on data governance. Our Tableau Community meets to empower each other and drive change with data.