Implementing a data governance program and building the framework to execute it can seem very time-consuming and complicated. But it doesn’t have to be. Although data governance means a lot of things and covers a lot of areas, it doesn’t have to be complex. In fact, a data governance program and data governance framework that integrate more naturally in business processes will get better user adoption and lead to better decision-making across the organization. In this blog, we explain what a democratized approach to data governance is and why it’s our preferred approach, we outline eight steps to get started with your data governance program, and we go over how to build a framework to execute your data governance program.

What Is a Data Governance Program?

A data governance program is a collection of practices and processes that form an approach to manage the data assets of an organization. But more importantly, data governance is how you add accountability and follow-through to ensure data management is being executed true to the plan defined in your data strategy.

If your data strategy is properly aligned with business strategy then data governance will be the mechanism to ensure data and analytics initiatives are aligned to driving business value.

Learn where your data governance program fits into your data strategy.

You can have a more centralized approach to your data governance program, which is more traditional and has strict rules in place about who has access to what data and how they can use it. Or, you can have a more democratized approach—that still requires security—where all necessary business users are granted access to data and enabled to self-serve analytics, and ultimately lead to faster decision-making.

Although there isn’t a one-size-fits all approach—and it usually depends on the size of your organization and the resources you have available to commit to a data governance program—a democratized approach will provide “just enough” data governance to see quick wins and measurable ROI, as well as allow you to start small and scale as you learn more.

Why Is a Data Governance Program Important?

Data governance maximizes the investment in data and analytics initiatives by promoting the proper use of analytics in business processes, ensuring accurate insights based on quality data, reducing risks with security, and guiding the prioritization of projects so the right information is available at the right time.

As businesses create vastly more data than they know how to process—coming in from hundreds of sources—they need a data governance program in place that will ensure a consistent approach to the valuation, creation, consumption, and control of data. The data governance program holds people, processes, and technology accountable by answering questions such as:

  • What constitutes data?
  • Where and how is data collected, extracted, transformed, delivered, and used?
  • Who cares for and maintains which data?
  • Who owns systems, who owns data, and who stewards data?

A data governance program is required in a modern and cross-functional data and analytics environment because it ensures that data is usable, high quality, trustworthy, accessible, and secure—and by extension, brings value to everything else that the organization is trying to do with the data.

What Are the 8 Steps to Start Your Data Governance Program?

When it comes to data governance, every organization is different—but all can benefit from following simple steps to building a data governance program.

Illustration showing the 8 steps to data governance, including Identify Leadership, Establish the Vision, Assess Existing Data Governance Components, Determine the “Right” Level of Data Governance, Automate Data Governance, Focus on people and process, Establish a Data Governance Committee, Roll-out and Sustain.

Every data governance program is different and ultimately depends on the needs and resources of an organization. Consider these eight steps to get you on the right track to successful data governance.

1.) Identify Leadership:

A top-down approach to data governance will garner the most support across the organization. Start with getting executive buy-in, then identify leaders throughout the organization that can steer a committee dedicated to seeing the program through; you want leaders who recognize data as an asset, not a technical responsibility.

Leadership should be prepared to provide the vision for how data will be used across the organization and set the precedent for what is expected from everyone—tech, business, and the stakeholders. They should be an advocate for change, celebrate successes, and help remove any barriers to success.

Blue illustration of enterprise data governance program detailing corporate organizational leadership.

No matter who leads the program, data governance should be a top-down approach and have buy-in across the organization.

2.) Establish the Vision:

Many people don’t comprehend how data governance affects their position or work environment. As you create and communicate your vision, demonstrate to each team how solid data governance can specifically help them do their job better and how their actions contribute to overall success. This will help them realize that their work, such as creating an Excel spreadsheet that several other people use on a shared drive, is a data asset and helps the company. When people know that they are a key component to your overall business goals, they will be more likely to adopt the data governance program you put in place.

3.) Assess Existing Data Governance Components:

Whether you know it or not, there are likely a lot of data governance activities already going on at your organization. Look at what you’re already doing—things like “master data management” and “information management”— to figure out where the holes are, where there are contradictions between departments, and what you need to have a stronger data governance program.

When we work with clients on data governance initiatives, we ask them to answer a series of questions to start their assessment, including:

  • How would you assess your data quality? Are there known data quality concerns?
  • Do different business units have different definitions/values for the same metric/KPI?
  • Is data lineage known and documented? By being transparent and defining data lineage, everyone who uses the data knows the who, what, and where behind it. This will ensure the integrity and trustworthiness of your data across the organization.
  • Does the business have ownership of data, or does IT?
  • How do discrepancies with KPIs and metric definitions get resolved today?

4.) Determine the “Right” Level of Data Governance:

There isn’t a one-size-fits-all approach to data governance and there are several factors that will determine the right level for your organization. The program you implement will vary depending on the:

  • Size of your organization and complexity of your data.
  • Resources—time, staff, and funds—you can make available.
  • Level of regulatory requirements within your industry.

Start small and iterate. Don’t attempt to govern every aspect of your data. Start with two to four business areas that are most interested in helping with data governance and begin by looking at the data subject areas that they’re most interested in. Form working groups right from the start so that the main work of data governance can be accomplished in small groups and the larger committee meetings can be focused on holistic progress and setting direction for the working groups. Keep in mind that data governance needs to be practical, maintainable, and proportional:

  • The level and intensity of data governance should scale with your organization’s size, needs, risks, maturity, and capabilities.
  • It should align with your data strategy roadmap.
  • It should set realistic expectations of what business stakeholders will be able to contribute to the program.

Your data governance program should blend into everyday operations—increasing adoption and reducing maintenance.

5.) Automate Data Governance:

This is where technology can help, and you may already have the tools to get started. Look into how you can automate data lineage in an existing ETL tool—for example, dbt includes a baked-in lineage feature that you can use. It also allows for embedding metric definitions and auditing tests. You can automate back-end monitoring by adding extra coding in your existing data transformation tools. Or you can automate KPIs and scorecards right within your existing BI platform. There are also specialized data governance tools on the market that can help with front-end business rules automation such as Atlan, Collibra, and Alation just to name a few.

Learn more about “What is dbt and What It Can Do For Your Data Pipeline

6.) Focus on People and Process:

Although there are technologies that will enable—or make easier—your data governance efforts, technology alone cannot lead to a successful data governance program. People and processes drive its success. Start by treating data governance like an actual program and not a project with start and end dates.

  • Determine the core list of what data governance will accomplish.
  • Identify processes that will be required to achieve those goals.
  • Identify where holes may exist in those processes.
  • Determine the organizational functions required to carry out data governance activities.

Make sure there is on-going support as your program expands and remember that each new component will require initial effort and then be folded into existing processes. Whether you are ready to invest in a data governance technology yet or not, it is a requirement to identify and plan for the people and processes which will govern your data assets. This foundational work will also speed up the implementation of a data governance tool if that is planned for a later date.

7.) Establish a Data Governance Committee:

Data governance activities are likely already happening within your organization in a less formal or undocumented fashion. Identify the people in the business who are doing this kind of work and form a data governance committee or add data governance responsibilities to an existing committee. The committee should include staff from a variety of business units whose purpose is to establish clear data definitions, develop comprehensive policies, oversee documentation by which internal business units collect, steward, disseminate, and integrate data on behalf of the organization. Your data governance committee should include data owners and data stewards.

8.) Roll-out and Sustain:

The data governance team should be dedicated to upholding the goals of the data governance program. Until data governance is totally internalized and ingrained, you must manage the transformation from non-governed data assets to governed data assets. Once the plan is rolled out and the change management plan is executed, the data governance framework should be continually evaluated for effectiveness. Use the metrics you defined in your data governance framework to assess performance on a scheduled basis.

Data governance is not self-sustaining; your organizational goals, the marketplace in which you do business, and your sources of data are going to constantly evolve, so be willing to adapt your data governance program as needed.

And finally, for your data governance program to be successful, you need to drive user adoption and that can only happen if you treat change management as part of the program. Attach the program to business objectives and on-going projects, clearly communicate the changes and how it affects the business users, explain their roles throughout the process, provide appropriate orientation and training around data literacy, and most importantly, celebrate wins—big or small.

Implementing a data governance program is a cultural shift within the organization. And for the implementation to be successful, you need a data governance framework that is easy to follow, easy to understand, and easy to communicate.

What Is a Data Governance Framework?

A data governance framework defines how you will implement a data governance program. It creates a single set of rules and processes around data management—and in turn makes it easier to execute your data governance program.

Talk to an expert about your data governance needs.

How Do You Build a Data Governance Framework to Support Your Data Governance Program?

Here are a few steps you can take in creating a data governance framework when approaching a more democratized data governance program:

Illustration of what to consider when building a data governance framework, including develop policies and standards, create a business glossary, establish data governance processes, and apply governance metrics.

When building a data governance framework, remember to develop policies and standards, create a business glossary, establish data governance processes, and apply governance metrics.

Develop Policies and Standards

To effectively communicate your data governance program and get buy-in from the organization, you need to establish policies and standards for data management from the beginning and assign data stewards whose role it is to ensure that the policies and standards turn into practice across the organization. Often two of the most urgent things to address are data quality and data security.

Establish and maintain thresholds for data quality metrics and identify and create rules for sensitive data and regulatory compliance. Democratizing your data still requires security, it’s just not as governed. Take inventory of policies and standards already in place, identify gaps, and add initiatives to a backlog to address the gaps. Define user roles and responsibilities for who will manage and monitor data quality and data security. While data quality and data security are critical, setting the prioritization and direction of ongoing business initiatives is an often-over-looked area of data governance.

Data governance can better ensure the long-term success of data and analytics initiatives by establishing the stakeholders and data stewards to approve priorities and own the definition of metrics and SLAs.

Learn more about “Why Data Quality Can’t be Fixed with Technology

Create A Business Glossary

A business glossary provides an agreed upon standard set of definitions across an organization and in many cases may force the business to collaborate on the definitions with stakeholders from multiple departments—sometimes to even re-think or improve the way they were defining a key metric. Start with a business glossary of your organization’s core KPIs and expand in iterations to provide maximum value. Avoid any kind of manually documented column-level technical metadata—required with many traditional data dictionaries—as it takes effort to build, is not a good use of time to maintain, quickly becomes outdated, and is usually too technical for business users.

Although there is value in data dictionaries and data catalogs if implemented incrementally with appropriate automation and scope, you will still need to first get agreement on the business definitions, which is why your first step should be to start with a business glossary.

Learn more about “The Difference Between a Business Glossary, a Data Dictionary, and A Data Catalog. How Do They Play a Role In Modern Data Management?

Establish Data Governance Processes

It’s critical to establish new data governance processes that can be integrated within existing business processes. Establishing data governance processes will allow you to scale and mature your data governance program over time. Examples include:

  • Processes to govern self-service: Even though business users are empowered to use self-serve analytics, you should still put in place guidelines on how to proceed. Processes may include a level of approval required before the business user can publish their analytics or use the results in an external facing presentation. It should also include supporting business users through training and a community of practice, so they are enabled to use the tool properly and maximize the investment in data and analytics.
  • Processes to govern how new data sources are integrated: Consuming more data from new data sources can help you meet your business objectives, but you should have some guardrails in how that data is integrated. You can dictate that new data acquired must have input from the business on how it will be used before back-end decisions are made on how it is modeled. A process can require that your data governance committee agree on definitions for any new business metrics.
  • Processes to govern new report or dashboard requests: When business users want to create a new dashboard, you should have processes in place that can confirm that there isn’t overlap with existing content or projects. Other processes can include, for example, documenting and standardizing if something like data validation is required.

Apply Governance Metrics

Defining and applying metrics to track and measure various aspects of your data governance program is crucial to understanding the impact, value, and relevance of the program across the organization. Some governance metrics examples include:

  • People metrics: An indicator of a successful data governance program is an increase in user adoption. You can track how often business users are logging in to view and refreshing their reports and dashboards.
  • Data quality metrics: An indicator of good data quality is decreased reliance on IT support and improved productivity among business users.

Remember, data governance doesn’t need to be complex, it just needs to be intentional. Have a plan, outline your expectations, and get everyone on the same page—your data governance program will lead to better quality data in the long run and improved analytics across the organization.

Christina Salmi Christina leads the Data Strategy Service Line, helping our customers to think and act strategically about data and analytics.
Subscribe to

The Insider

Sign up to receive our monthly newsletter, and get the latest insights, tips, advice.

Thank You!