8 Best Practices for a Robust Data Governance Program

March 19th, 2022

header image for 8 Best Practices for a Robust Data Governance Program

What are data governance best practices?

Governance best practices are guard rails and policies that help make your data accurate, complete, consistent, relevant and help stay compliant with data protection regulations.

List of 8 time-tested best practices for building a robust data governance program:

  1. Have a well-established, measurable, organization-wide goal
  2. Understand the rules and regulations for compliance
  3. Involve C-suite from the beginning
  4. Adopt a solid data governance framework
  5. Identify all the data domains
  6. Educate everyone on the importance and benefits of data governance
  7. Track progress
  8. Invest in the right data governance technology

We’ll take a deeper look into each of these data governance best practices in detail.


Understanding and implementing data governance best practices

Implementing governance across the span of a data life cycle requires taking stock and understanding the ways people, process, and technology interacts with data in the present organaisational setup. This broadly includes:

  • Assessing your current data governance program
  • Understanding challenges to implementing data governance
  • Benchmarking and listing down how data governance can be improved?


Assessing your current data governance program

If you’re wondering whether your data governance program is doing what it’s supposed to, here’s a list of questions to ask that will either ease your mind or add to the anxiety.

Either way, these questions can help you gauge the effectiveness of your data governance program. So, here we go:

  1. What data does your organization have?
  2. Where does this data live?
  3. Where does it flow?
  4. What is it used for?
  5. Who owns that data? Is it IT, business, or both?
  6. Who defines, modifies, and uses that data?

Now, this list can go on and on until kingdom come.

But, the key here is to know what data you have, where it lives, who owns it or handles it, and how it gets used. If you’re unsure about your current data governance program addressing these questions, then what you need are a few data governance best practices to tweak your program and improve its chances of success.



Challenges to data governance

However, achieving these objectives and demonstrating the impact of data governance on revenue or efficiency is a Herculean task.

A top reason — organizations don’t assess, monitor, or measure their data governance systems.

Another reason — the outcomes focus on data and analytics rather than a significant business goal.

So, it’s tough to convince everyone, especially senior leadership, as they don’t see the impact of the data governance investments on business.

It’s common for data governance programs to be fragmented and, as a result, not adopted by the organization as a whole. Here’s how OECD puts it:

Data governance elements are often in place as part of broader digital transformation policies. However, these components can be fragmented, thus reducing their value. A holistic data governance (program) can help.

Lastly, organizations seek a quick fix for their governance woes rather than focusing on a long-term, continuous program. According to Mozilla:

Data governance is the equivalent of a ‘diet and exercise’ prescription for how to solve a digital rights problem - it’s not easy or convenient, but it’s a proven, obvious, and effective way to build legitimacy. By contrast, most of the data economy wants a ‘magic pill’ approach - a box to check to validate whatever their intended use of data is.



How can data governance be improved?

Every data governance program is ultimately about:

  • Creating value
  • Cutting costs
  • Increasing efficiency
  • Thwarting cyber attacks (to preserve the integrity and privacy of data)
  1. Get buy-in from everyone
  2. Map governance to significant business goals
  3. Ensure collaboration and transparency in everything from data definitions to transformations
  4. Adopt a well-designed data governance framework encompassing the rules, standards, roles, responsibilities, and more

That’s where following a few data governance best practices can help.


8 data governance best practices to follow


1. Have a well-established, measurable, organization-wide goal

As mentioned earlier, a strategic goal must direct the data governance program for it to work. This goal has to be unique to each organization. Otherwise, it becomes a lot like this shepherd and his sheep pointing in different directions.

Have a well-established, measurable, organization-wide goal

Every data governance program needs a direction. Source.

That means having a goal that impacts revenue, growth, efficiency, or risks. So, it all begins by defining your “why” for the program and mapping out ways to measure it.

A good rule of thumb here is to start with the end in mind — identify the data problems that must be removed for the entire organization to derive any value from it, or the business goals that must be met and how data can help meet them.

Examples of data governance goals

  1. Business goals:
    • Increase operational efficiency
    • Improve customer satisfaction
    • Increase sales for a certain region
    • Streamline data from mergers and acquisitions and get it ready to use for insights
  2. IT goals:
    • Improve data quality and accuracy so that it’s trustworthy
    • Support business integrations
    • Simplify data migrations

Here’s another (mini) data governance best practice — after mapping the goals, share them with all the stakeholders across the organization and get their support.


2. Understand the rules and regulations for compliance

Different industries come under varying levels of scrutiny.

For instance, banks and pharmaceutical companies have to wade through a sea of regulations, compared to a retail business or a media agency. Similarly, the geographies where you operate also influence the regulatory mandates.

That’s why identifying the regulatory requirements is essential to designing a data governance program that helps meet your goals and complies with regulations.

Here’s a handy graphic from McKinsey explaining the level of sophistication required from a data governance program.

Chart explaining data-governance archetypes

A matrix to decide the level of sophistication needed from a data governance program. Source.

This goes without saying, but you shouldn’t proceed with the nitty-gritties of your governance program without familiarizing yourself with the compliance requirements.

That’s why yet another good practice here is to set up a steering committee —with people from business, IT, legal, and finance — in charge of research and move on to the next stage after they present their findings.


3. Involve C-suite from the beginning

Here’s an excerpt from a McKinsey report on building data governance programs that offer value:

A leading global retailer struggled to extract value from their data. So they embarked on an enterprise-wide analytics transformation journey. A huge part of the journey was investing in educating and involving the entire senior leadership team in data governance. Each executive leader was made in charge of either data domains or business-data subject areas across multiple lines of business.

The result? The C-suite became the champions of the data governance program, speeding up the process of establishing data domains and reducing the time spent cleaning up data assets.

Within months, the retailer could derive value from their data and rely on analytics-driven decision-making.


Conversation starters for the C-suite

If you’re still struggling to capture the leadership’s interest, then here’s a great way to begin is by addressing questions such as:

  • What data do you need to meet your KPIs/goals? Do you have a “wish list” of sorts?
  • What challenges do you face with accessing that data seamlessly?
  • How does that impact our product or customer service?
  • What about the operational challenges faced?
  • Are there any significant risks (from a compliance, legal, or security standpoint) to the company because of bad data?

Referring to the business value of data from the beginning, using questions like the ones mentioned above, gets the leadership's attention and helps you design your data governance initiatives more efficiently.


Messaging for the C-Suite

One more thing that matters when dealing with the C-suite is messaging. So, Dataversity recommends best practices for communicating with the management regarding data governance programs and updates.

Involve C-suite from the beginning-1

Getting the senior management on board with the right messaging. Source.


Data governance communication management

Now, keeping everyone in the loop at all times is also an arduous task. That’s where technology can help. Modern data governance platforms like Atlan are equipped with centralized dashboards to manage access, along with in-built chats and automated reporting.

Screenshot from Atlan platform's 'Manage Users' view

A single dashboard to manage and monitor data access rights. Image by Atlan


Screenshot from Atlan platform's 'Chat' function

In-built chats to discuss data assets. Image by Atlan


4. Adopt a solid data governance framework

A data governance framework must ensure the proper management of data through its entire life cycle.

OECD

Every organization should have a set of rules — policies, standards, controls — describing:

  • How data gets used
  • Who monitors data usage
  • Processes governing data to ensure quality and compliance

That’s why a key data governance best practice is to adopt a comprehensive data governance framework that’s best suited for your organization.

While there are several well-established frameworks out there, you should pick one that matches the needs of your organization. Here’s a handy set of rules to remember while crafting your data governance framework.

Adopt a solid data governance framework

Tips to pick a data governance framework for your organization. Source.

Read more on data governance frameworks.


5. Identify all the data domains

This (identifying data domains) is a big step -- corralling all the data sets and access points to start thinking about systemic data control points. This step includes mapping out a soup-to-nuts plan to define automated workflow processes, approval thresholds, reviews, issue resolution, and a whole lot more.

Snowflake

Examples of data domains include sales, finance and accounting, purchasing, and manufacturing.

First, however, it's crucial to spot the data domains that will affect your business goals and identify critical elements within each data domain. That's because critical data typically represents only 10 to 20 percent of total data in most organizations.

Here’s how a North American retailer did it, according to McKinsey:

The company wanted to transform itself with advanced analytics. However, they realized that the current data wasn’t enough. So, they set up a Data Management Office (DMO) and identified ten domains across the enterprise that would affect their governance program. Of these, the company prioritized the deployment of transactional data (logging in-store purchases) and product data (establishing a clear hierarchy of products and their details).

As a result, the retailer fast-tracked the implementation of their top data governance use cases. Once the data domains are mapped, it’s time to chart out the relevant data definitions, standards, usage rules, and more, as provided in the framework.

Once these definitions and rules have been established, technology can help automate several processes, such as cataloging, quality checks, PII classification, and data lineage mapping.

Screenshot showing Atlan's Atlan’s algorithms auto classify PII

Auto-classified PII and sensitive data. Image by Atlan


Screenshot from Atlan's data lineage flowchart

Automatic parsing of SQL queries from the warehouse to map data lineage. Image by Atlan


6. Educate everyone on the importance and benefits of data governance

Good data governance does not happen in isolation. [Instead], it benefits from the adoption of open, inclusive, iterative, collective, and value-based approaches to its definition, implementation, evaluation, and change.

It is also not the responsibility of a small group of people. [On the contrary], it should reflect the needs of a globalized, fast-paced, diverse, digitized and interconnected world.

OECD

That applies to businesses too. A data governance program should reflect the needs and objectives of the entire organization. And, it shouldn’t be the responsibility of just one group (like IT).


Encourage collaboration and participation

As we’ve mentioned earlier, the overall goal should tie into the business objectives rather than purely solving the problems of data and analytics teams. That’s why the next step is to encourage collaboration and participation so that data governance becomes an organization-wide mandate.

That requires educating everyone on data governance and getting them on board is vital for the program’s success.

Here’s how McKinsey puts it:

When people are excited and committed to the vision of data enablement, they’re more likely to help ensure that data is high quality and safe. Leading organizations invest in change management to build data supporters and convert the skeptics.

Lastly, since data governance programs mature and evolve as the organization’s data management becomes streamlined, another data governance best practice is to ensure that the process of educating and updating everyone must be continuous.


7. Track progress

At the risk of sounding like Captain Obvious, you can’t manage what you can’t measure. That’s why your data governance program should have a way of reporting that tracks its progress towards the program goals.

While this depends on your governance program’s KPIs, it could include anything from updates on lineage mapping to final decisions on the data owners for specific domains and their responsibilities.


How will you track data governance KPIs?

Naturally, the next question is — what should we use to track these KPIs?

Several companies still use good ol’ spreadsheets to track progress and while it works for smaller use cases, it’s not practical for organization-wide governance initiatives. That’s why it’s a better idea to invest in technology that automates and simplifies monitoring your data governance efforts.

For instance, you could use a centralized dashboard that:

  • Automates reporting
  • Displays who the data owners are and tag them or ask questions
  • Traces lineage to verify the origins of data sets
  • Maintains a business glossary to add context to your data

8. Invest in the right data governance technology

As mentioned earlier, technology is fundamental to the success of your governance programs. However, it’s crucial to invest in the right kind of technology that meets your requirements.

Any good governance platform should:

  • Democratize access to data
  • Simplify regulatory compliance
  • Ensure impeccable data quality
  • Automate processes such as profiling, cataloging, and PII classification

For example, Atlan lets users create granular access controls and manage requests, policies, and access logs from a single dashboard, making regulating and granting access a breeze.


Screenshot from Atlan platform's 'Manage Request' functions

Manage all access requests in one place. Image by Atlan


Atlan also parses through SQL queries to auto-construct lineage and auto-detect PII data. Additionally, it helps you maintain granular audit logs that track actions such as users accessing data assets or the frequency of access.

Screenshot from Atlan's Access Logs

Granular access logs to track who’s accessing which data set and at what frequency. Image by Atlan


Atlan also lets users run automated quality checks to look for anomalies and guarantee the integrity of data. Users can automatically spot bad data using algorithms for duplicates, outliers, cross-column categorical patterns, and more.

Screenshot from Atlan's 'Metrics Schedule' function

Schedule automated checks to monitor data quality. Image by Atlan


Moreover, Atlan automates cataloging, and auto-populates metadata descriptions and data dictionaries, which makes your organization’s data relevant, useful and valuable.

Lastly, the platform should be straightforward, easy to set up and use, without requiring any technical knowledge or interventions from IT — the reason gaining access to the right data takes forever in the first place.

Invest in the right technology-4

All humans of data benefit from using Atlan for everything from data exploration to governance. Image by Atlan


Data governance best practices: Next Steps

Implementing data governance programs is a monumental undertaking. That’s why a solid plan, impactful goals, relevant and real-time metrics, and an emphasis on constant communication and collaboration are essential data governance best practices to embrace.

Ready to make data governance effortless?
Try Atlan — Auto-construct data lineage and deploy best-in-class data access governance without compromising on data democratization.


Ebook cover - metadata catalog primer

Everything you need to know about modern data catalogs

Adopting a modern data catalog is the first step towards data discovery. In this guide, we explore the evolution of the data management ecosystem, the challenges created by traditional data catalog solutions, and what an ideal, modern-day data catalog should look like. Download now!