Data governance is an essential component of data management. It describes how data is processed within an organization. Who owns what data asset? Who can access it? What security measures are in place? How does it comply with regulations?
In this article, you will learn how to use a data governance framework to set guidelines around data access. You'll also learn how to create a data governance program to gain a maximize the return on your data investments and quantify the impact of these measures with data governance metrics.
Considerations for adopting a data governance framework
A data governance framework should be your blueprint for implementing data governance programs. It creates a set of rules and processes for collecting, storing, and using data in an organization. It outlines how processes, people, and technology interact to:
- Make data accessible and comprehensible
- Ensure a high degree of data quality
- Guarantee data security and compliance
But it’s important to note that accessibility, quality, and compliance aren’t goals by themselves. Yes, each of these three benefits can positively impact the day-to-day work of many people within an organization, but the true purpose is to maximize the return on your data investments.
On the one hand, data governance ensures that an organization can make the most of the opportunities that lie within its data assets: data should be readily available for improved decision making or new data products. On the other hand, it helps an organization avoid the associated risks.
Most data governance programs focus on either one or the other.
- Build repeatable processes that improve decision-making processes
- Improve the resulting decisions
- Reduce operational friction, often by enhancing communication between teams or departments
- Ensure regulatory compliance
- Assure the privacy of data stakeholders
The final product of a data governance program is often the creation of data catalogs or definition glossaries, but it is essential to understand that implementing these things doesn’t necessarily mean that the program has been successful. Technology is only one piece of the puzzle. Success is defined by how you embed these technologies into your processes—the way people actually work in your organization.
Success factors for creating and implementing a data governance framework
To successfully develop and implement a data governance framework, take these first steps to add clarity:
1. Craft a business case and an accompanying vision
Your business case should clearly state your top goals for the data governance program and your methods for achieving these goals. A concrete vision should be actionable, ideally something reducible to a two to three-sentence pitch or a series of bullet points.
For example, let’s take the business case: “reduce time-to-decision by standardizing metrics, producing less conflicting analyses, dashboards, and reports.” This vision answers the goal that you are trying to achieve, why you want to achieve it, and your intended end-state.
2. Outline your first program
To make your first data governance program count, your primary criterion should be return-on-investment (ROI). Start with a concrete issue that demonstrates value to the business. This helps you get buy-in from across the business and the company as a whole. Next, choose a program that produces measurable results. If you have the numbers to back you up, you will improve the chances of expanding the program to a company-wide endeavor.
3. Clarify roles within your team
As with many change management programs, roles should be clear from the start. Typical roles within a data governance program include:
- The program owner dedicates a large percentage of their time to the program and is responsible for propelling it toward the finish line.
- The (executive) program sponsor is accountable for removing roadblocks and resolving issues that the other roles cannot resolve. The sponsor also provides the required resources.
- Program members are business and IT subject matter experts who will be impacted by the program, both downstream and upstream. They inform and facilitate the work of the program owner.
- The informed involves anyone interested in the program, including those who consume the data that is, or will be, processed according to the rules set forward by the data governance framework.
4. Structure a program road map
A data governance framework contains the rules that govern how data is processed, accessed, and validated. Defining these rules is often done in four separate steps:
- Discovery: First, the data must be identified, understood, and mapped to the business processes that consume or produce the data. Usually, this is done manually and is a very cumbersome task. This means navigating through several hierarchical layers of the company. But it also means you have to be familiar with a variety of legacy technologies.
- Definition: All findings from the discovery phase need to be codified into definitions, policies, standards, and processes. This phase is also the right time for establishing proper key performance indicators to evaluate the data governance program’s impact.
- Implementation: Everything that has been codified within the definition phase eventually needs to be implemented. And this is not a one-person job. Everybody who interacts with the data, in any form, is responsible for adhering to the policies and processes put forward. By clearly defining the processes and publishing them, it’s easier to communicate them and hold people accountable for following them.
- Evaluation: You can evaluate data governance efforts through measurement and monitoring. Although this is no easy feat, many technologies that facilitate data governance can be used to quantify how many people interact with the policies and processes.
5. Adopt fit-for-purpose technology
Your organization probably already has some tools that can facilitate proper data governance. However, it’s worth discovering best-of-breed vendors that produce data governance software that might be a fit for your data governance framework.
Good data governance tools integrate closely with your data management ecosystem, encourage collaboration, and scale with your organization’s needs.
Measuring the success of your data governance framework with metrics
Quantifying the impact of your data governance efforts is a convenient way to demonstrate its success to stakeholders.
If you’re working for a large organization, you should realize that governing all data assets all at once is probably impossible. For this reason, it makes sense to prioritize your data sources into what is often referred to as *key data elements*. These are the data elements with material impact on your organization’s operations, decisions, and other data requirements of regulatory nature.
Secondly, by breaking down data governance into various focus areas, a framework makes developing data governance metrics a lot more graspable. Below, learn more about key metrics related to data accessibility and security.
Finally, the number of available metrics can quickly grow. For this reason, many organizations have a metrics store on top of their data warehouse. A metrics store is a centralized and governed place for organizations to store key metrics. It is a way of centralizing organizational knowledge, creating a repository for stakeholders to access key metrics in a repeatable way, regardless of how they plan to consume them downstream. Why not start with your data governance metrics?
Creating a data-informed organization requires that data is easily accessible to all who need it. In the age of big data, this usually means making data sources available in an enterprise data warehouse or the supported business intelligence tools. The following are examples of data accessibility metrics:
- Total Data Usage: How many times have key data elements been accessed?
- Individual Data Usage: How many different people have accessed them?
- Report Dependencies: In how many reports are key data elements used?
Security, compliance, and privacy
Many data security metrics have to be tracked due to industry-specific regulations, though that isn’t always the case. Nevertheless, security, compliance, and privacy are integral to many data governance frameworks, so setting up metrics is always a good idea. Here are some examples:
- Processing time of data-related requests
- Timely deletion of data sets with regard to their maximum retention period
- Number of data sources for which data privacy controls have been implemented
Creating and implementing a data governance framework are much more manageable by breaking them down into data governance programs with an achievable scope: data quality, accessibility, and compliance. Making a business case for your first program, explicitly defining roles and responsibilities, and presenting a clear road map can drastically increase your chances of success. Finally, data governance isn’t a one-off effort—it requires constant monitoring. Data governance metrics help you evaluate the ongoing impact of your data governance framework.
To create organization-wide alignment across key metrics that integrate into other enterprise tools, consider Transform. Complement your data governance framework with Transform’s Metrics Catalog to identify a common set of metrics definitions and create alignment across your organization.
This post is guest authored by Roel Peters. Roel is a data generalist with a special interest in making business cases using structured data and simulation techniques.