What Is Data Governance?

Data governance is the set of processes, procedures, policies, standards and metrics that organizations implement to understand what their data is, where their data is and how best to use it. It also dictates who gets access to what data, in what situations, and using what methods.

Data governance typically requires a dedicated governance team, including a steering committee to act as the governing body and a group of data stewards to enforce everything. Ideally, data governance also involves executives and other stakeholders, along with the IT and data management teams.

“Data governance is an enterprise-wide undertaking. It’s something that is designed to be supporting all data across the entire company,” Malcolm Hawker, head of data strategy at data management company Profisee, told Built In. “It’s the rules of the road when it comes to data.”

Data governance is the set of internal standards an organization follows to ensure its data is accurate, usable, secure and available to the right people and under the right circumstances.

Data governance is often closely associated with data management. And while the two are related, they are not synonymous.

Data management refers to the process of creating, collecting, storing, maintaining, securing and using all the data in an entire company. It covers the entire data lifecycle.

Data governance is a component of data management, along with data quality, data security, data warehousing and so on.

Another way to look at it is the “what” versus the “how,” Hawker said.

“Governance is the what, management is the how. Management is how you actually support the policies and procedures,” he explained. For example, putting controls on data regarding who can create, update and access it. “Data governance is policies and procedures, management is doing it. It’s putting a shovel in the ground and actually implementing the policies and procedures.”

To ensure compliance with government regulations and data privacy legislation, companies have to know what kind of data they’re collecting, where it is being stored, how it is being used, and by whom.

And as governments continue to tighten regulations on data usage, data privacy and even companies’ environmental, social and governance metrics, data governance is becoming even more important in making sure businesses’ data collection, usage and security stay within the bounds of the law.

“If you’re a large company, and you fall under one of the global privacy protection acts, and you don’t have data governance, then you’re going to be in trouble,” Peggy Tsai, chief data officer at data security company BigID, told Built In. “Or, if you’re any company, and you don’t want to be hacked or have any kind of cybersecurity breach, you better have a data governance program.”

Data governance has also shown to be more than just a means of securing one’s data and ensuring it is compliant with the law. It can also be a business enabler — “rocket fuel,” as Hawker put it — that allows an organization to run more efficiently and try new things with its data. Particularly when it comes to the implementation of artificial intelligence, which is something most companies are looking to do more of.

“No organization can be data-driven, or can do great analytics, or can do any type of AI or predictive analytics, without having a basic data governance framework in place.”

Indeed, artificial intelligence — especially generative AI, where machines automatically produce their own content — is all the rage. Companies are using AI to streamline a variety of business processes, from marketing to people management. Executing all of this well requires an immense amount of data. And to mitigate any bias, ethical or legal issues, data governance should be an important piece of any company’s AI execution strategy.

“A good data governance program really enables you to take advantage of that kind of technology rapidly, safely, legally, ethically and in a commercially valuable way,” Jay Militscher, VP of data analytics at data intelligence company Collibra, told Built In. “Not having a good data governance program means you can do really bad things without even intending to.”

So whether a company is trying to reduce its carbon footprint, or experiment with things like chatbots and AI content generators, data governance is a must-have. “No organization can be data-driven, or can do great analytics, or can do any type of AI or predictive analytics, without having a basic data governance framework in place,” Tsai said.

More on Data PrivacyData Privacy Laws Every Company Should Know

When it comes to implementing a data governance program, there is no one-size-fits-all method. All companies are different. Still, all data governance programs have some guiding principles — a set of fundamental goals to achieve success.

Companies are often structured into isolated departments — each with very different ways that they create, manage and define their data. High quality data for the marketing department isn’t the same as what it would be for the finance department. And people working on the sales side might quantify customers differently than those working on the product side. This inevitably makes it difficult for a company to have a consistent understanding of the data it owns, where it is stored and what to do with it.

A well-executed data governance framework ensures that all of this disparate data is consistently defined, organized and protected so that it can serve as a shared resource. It covers all the strategic, tactical and operational aspects of managing data so that everyone — no matter what department they’re in — is on the same page.

“A great data governance program is going to break down those silos and get the whole organization connected through data,” Militscher said. “Everybody can have a common understanding of the organization’s information.”

Without an effective data governance program, any data inconsistencies across an organization might not get resolved or even noticed because companies are so often siloed.

For example, customer names may be listed differently depending on the department, which could complicate data integration efforts and create data integrity issues, ultimately affecting the accuracy of business intelligence, analytics and more if it is not detected and resolved.

Food For ThoughtAre You Wasting Money on Big Data?

Perhaps the most important guiding principle of all is to focus on value — not just how data affects the value of the business, but the value of the data itself. It’s important for companies to have a firm understanding of how much it costs to create a new piece of data, or how much revenue they generate from having good data

Valuing is vital because it gives data leaders a quantifiable justification for placing more controls on the way different teams use and access their data. It looks beyond the vague idea of data governance as a means of achieving “better data,” and instead offers commensurate proof. After all, it’s a lot easier to tell people what to do if you have data showing it will save them money, or increase their revenue.

“It’s like world peace. Everybody knows it makes sense. ‘Better data’ — intuitively, we know it makes sense,” Hawker said. “And that’s not just kind of a feeling. It’s actually true if you take the time to quantify it, if you take the time to model it.”

No two data governance programs are created equal. But there are some general best practices organizations are encouraged to follow when creating their own data governance framework.

A big part of data governance is about building a glossary of definitions in order to establish what specific business terms and concepts mean. For example, what constitutes an active customer. By establishing a common vocabulary for business data, all members of the company are on the same page, making it easy to move forward in their data governance program.

Another common best practice is to build a data catalog, which collects metadata from company databases, data warehouses, data lakes, business intelligence systems and other sources, and uses it to create a searchable inventory of data assets. It also includes information such as how the data has been collected, where it came from, how it is being stored, what it is being used for, and who in the company has access to it. Tracking how data flows through an organization, and classifying them based on factors such as whether it contains personal information, helps determine how data governance policies can be applied to individual data assets.

Dive DeeperData Lake vs. Data Warehouse: Will the Already-Blurred Line Between Them Disappear?

Doing data governance in such a broad and comprehensive way isn’t necessarily the most popular or efficient way to do it, though.

“If you’re starting from zero, and you’re saying ‘Let’s implement data governance!’ And you follow the standard best practices approach. It’s inevitably a multi-year thing for most organizations,” Hawker said. Not to mention an expensive one. “And if you’re perceived as slowing things down, if you’re perceived as getting in the way of business processes, you’re going to fail.”

Instead, Hawker suggests turning the process upside down and starting with an outcome. Identify a measurable KPI that has already been established and that leaders in the company already care a lot about — reducing the time to onboard a new customer, for example, or identifying 20 percent more qualified leads. And then build a data governance program around that outcome to ensure that the goal can be reached.

Once that one outcome is delivered, do it again for a different one, and so on, until a full data governance program is in place.

No matter how an organization chooses to approach it, data governance isn’t easy. But it can certainly be rewarding. Tsai likens data governance to demolishing a house, where the company is the house and its data is the infrastructure that keeps the house standing.

“Once you start ripping the house apart — the actual plumbing and the framework — that is, I would say, the dirty work that comes with doing data governance. And most companies don’t want to invest in it because it takes so much time and effort to do it,” she explained. “But there’s no risk. There’s no harm in just trying something.”

Get Some Expert AdviceMake Company Leadership Stop Ignoring You Analytics

Data governance has a variety of benefits that make it worth the effort. Beyond more accurate data analytics and stronger regulatory compliance, data governance can improve data quality and even help them identify new market or product opportunities.

Data governance also allows companies to intricately catalog their data, and closely measure and monitor the usage of it, potentially identifying any dark data — data that is going unused, which is a waste of opportunity and money — that may be lurking around.

“While it’s cheaper to store, manage and use data these days with great cloud technologies that are out there, it isn’t free. There is a cost associated with it. And data volumes are growing, and the use is growing,” Collibra’s Militscher said. “If data doesn’t need to be collected or stored because it’s never going to get used, then get rid of it. You’re saving yourself some money.”

“Great data governance is completely invisible and completely silent. You know you are doing it well when the business is humming.”

In the end, having consistent, uniform data across an organization, as well as clear rules regarding who can access that data, where and when, can help businesses run with more efficiency. They no longer have to be bogged down by the chaos that so often comes with disorganized data and unclear requirements. Instead, they can move forward with confidence and agility, without having to work too hard.

“Great governance is completely invisible and completely silent. You know when you are doing it well when the business is humming,” Hawker said. “When data is at the very core of how you operate, but it's not viewed as something different from how you operate. It’s just viewed as what flour is to a cake.”