Businesses live and die by their data. It’s the information that drives critical decisions, informs creative pursuits, and confirms the value of executed ideas. With so much at stake, a company must do more than hope their data is relevant and accurate, they must take strategic steps to guarantee it. That’s why strong guidelines around data governance are so significant. Data governance is the process of documenting the acquisition, reporting, and disposal of information. It provides rules on combining data to develop information and knowledge.
Data governance sets a precedent for the processes, policies, standards, and metrics that inform the most effective use of a business’ information. It's a function that supports the company’s overall data management strategy to minimize risks, increase data value, cut costs, and make sure compliance requirements are met.
As an organization begins that journey, there are a few considerations to keep in mind: What’s the most effective way to approach this? Are there best practices around data governance we should be aware of? Let’s answer these dilemmas.
What’s the Most Effective Way to Approach Data Governance?
Data governance is a multi-pronged undertaking that involves participants on various levels. It's not the sole responsibility of one party (like IT or management) but rather the entire organization at different points. Data governance approaches can be centralized with one authority managing everything or decentralized with multiple groups having authority towards the governance planning and execution. Which one you choose is dependent on your business model and processes.
One of the easiest and most recognizable ways to approach data governance is through a pyramid model. Governance responsibilities extend from the top levels down and involve different stages of authority and participation.
Top of the Pyramid: Data Governance Office
A data governance office—also referred to as master data council or data governance team—defines the policies and standards of the data governance plan and is responsible for communicating and administering those standards. A company’s chief data officer (CDO) is usually the ideal candidate to head this team and bring other data leaders into the conversation. A Data Governance Lead (DGL), IT reps, and a coordinator are also part of this team.
The DGL is usually someone who is a step below a VP and is committed to their governance role full-time. The DGL enforces policies, proposes new projects, and coordinates between business and tech groups.
All of the key players in the data governance office work together to put out training, playbooks, and other tools that help others in the organization educate themselves on what’s expected and why. When the data governance council (which we’ll talk about next) needs to meet, the data governance office provides an agenda for approval beforehand.
The office is also responsible for linking the data-governance efforts to the company’s technology, coordinating the application of the policies and standards across key roles, and supporting data quality analysis and remediation.
Top-Middle of the Pyramid: Data Governance Council (aka Steering Committee)
The data governance council—also called a steering committee, data council, data stewardship committee, data owner’s council, or data governance advisory council, depending on who you ask—is made up of senior management members like VPs, and Directors of departments like Marketing and Finance. One person from each business unit is represented in the group (plus an alternate) along with key stakeholders in IT. The head of the data governance office chairs this committee and reports to the CEO.
This committee takes ownership of the strategic direction of the data governance, confirming that standards and policies are being adhered to and reviewing the progress of initiatives around the data governance plans. They decide how data is maintained and what it should contain. They also resolve or escalate any issues that may arise. The council approves any funding necessary to accomplish data governance initiatives and prioritizes what those initiatives should be.
It’s also the data governance council’s responsibility to assign data stewards and owners (the next tier down) to hold data users accountable, champion the data governance plan, and communicate/coordinate the data governance policies at an operational level.
Middle-Bottom of the Pyramid: Data Stewards & Owners
Both data owners and stewards participate in the data governance council and are responsible for overseeing and protecting data as it relates to the data governance plan. Unsurprisingly, data owners “own” data quality for their data domains, which are sometimes referred to as “subject areas”.
What are data domains in data governance? We can think of them as high-level data categories used to assign accountability and responsibility for that data. For example, in the healthcare sector, data domains might be for patients, facilities, and medical procedures. For education, data domains may be defined as students, faculty, alumni, and research.
Data owners decide who has the right to access and edit data related to their domain and how they can use it. They also act as a contact for data users who are experiencing any issues with their data.
Data stewards act as ambassadors between the data team and the user community. As subject matter experts, they work with other data stewards across the lines of business to ensure each domain’s data is being properly managed and understood. Data stewards can act within specific subject areas as data domain stewards, or coordinate across the business and functional units as data steward coordinators. Overall, they work to enforce data governance policies and procedures for data users.
Data governance support teams might also include roles for data architects, modelers, and analysts. Architects map out the structure and organization of the data relevant to business functions, modelers are in charge of obtaining and documenting business rules for data quality, and analysts research problems for data owners and work to resolve data quality concerns.
Bottom of the Pyramid: Data Users
Data users may be at the bottom of the pyramid, but even they play a significant role in data governance. On a daily basis, these users are responsible for entering and utilizing data according to the data governance policies. As they go about their tasks, they are charged with conforming to the standards expected of them and informing data owners and stewards of any issues they encounter.
What are some Best Practices Around Data Governance?
To execute data governance effectively, you need a plan. What are your goals? What’s your strategy? How will this all work? Here are a few best practices to keep in mind as you explore data governance.
#1: People + Processes + Technology
Data governance requires small steps to begin. The first thing you must do is think about the people you want involved, the processes you plan to focus on, and the technologies that will support your goals. Once you’ve identified these main pieces, you can start to strategize your governance approach.
#2 Determine Governance Strategy & Team Structure
Before defining a team structure for governance, it’s important to create a strategy to work from. A data governance charter that includes a mission statement and overall program goals can be created with a team of stakeholders. Not only can this charter define the team structure and explain who has authority with data, but it can also explain how wide-reaching the proposed standards should be so teams will have something solid to work from.
#3 Make Data Governance Part of the Larger Whole
Maximize the value of data governance efforts by integrating governance plans with other ongoing efforts at your organization. For example, if the company is focusing on digitization, tie-in the functions of data governance with those projects. Product owners focused on driving process improvements also take responsibility for data governance within those new domains so, as they build out new capabilities, they are not just creating and consuming data, but also actively shepherding it.
#3 Pick and Choose
Trying to tackle all data assets at once creates a sizable project scope that can slow down the data governance progress. On top of that, the “big picture” approach can also end up hurting efforts to link governance to real business needs. Instead, prioritize the data assets by domains and by the data within each domain. The data council should review regulatory requirements, transformational efforts, and other factors to determine domain priority and choose two to three to roll out and make fully functional first.
Once those domains are chosen, it’s crucial to look at the assets within each to decide which data elements are most critical for data governance. For example, customer names and addresses would be critical elements within a domain. Plans for ongoing quality monitoring and tracking of this information would be crucial for analytics and reporting. Data assets used less often could be handled differently, perhaps with simple ad hoc quality monitoring and minimal tracking. Picking and choosing your data governance domains and associated data assets narrows the governance scope and helps teams prioritize the most important data.
#4 Make Sure you Can Measure It
How can you know if your data governance plan is working without being able to measure it? Any changes made to the data governance approach should be measured before to justify the results after. Track and collect metrics over time so you can see the overall changes and ensure the processes are effective and practical for your organization.
Putting governance in place ensures transparency, quality, and security for your businesses’ most important asset: data. With policies and procedures in place that everyone can follow, your organization can cut costs and extract more meaningful insights from your information. For help getting started with data governance, contact our expert team.