All organizations make decisions about data and seek to maximize their enterprise data assets. A formal data governance framework promotes active planning and management of data assets and provides multiple benefits to data users across an organization. These benefits include increased data consistency, formalizing and enforcing data access rules, defining privacy and security policies, monitoring data quality, and tracking data usage.
What is Data Governance?
“Data governance is a business program and bedrock that supports harmonized data activities across the organization. It accomplishes this goal as a formalized framework implemented to the specifications of a corporate Data Strategy. ” — Dataversity.
A data governance framework goes beyond the technical and security requirements of managing and protecting information. Data users need tools, defined processes, and effective education to grant the right people access to the information they need at the right time. A good data governance framework also seeks to bar nefarious actors from accessing data systems.
Goals of Data Governance
Determining goals for data governance will shape how you proceed with developing your framework. Common goals include:
- Improving data quality to increase user confidence in AI and analytics results
- Reducing costs and eliminating duplication
- Enabling data led decision making
- Ensuring compliance with quality and security standards
- Making data easier to locate and use
Key Components of a Successful Data Governance Strategy
Having a successful data governance strategy starts with incorporating the key components for your framework. Here are the key components we recommend based on our modern analytics and AI work. These components help strengthen data governance and security efforts within an organization by using proven practices, adopting Agile principles, and shifting left on security to build in these considerations from the start.
Identify Your Direction
Start with identifying the ‘why’ for the organization.
Teams should start by identifying shared goals that address existing challenges around trust, privacy, cost, compliance, and security. By starting with a shared understanding of the pain points on how data is managed, aggregated, and used, teams get to their “why” quicker and more aligned. A shared ‘why’ also increases the buy-in of the value of data governance. This will also be instrumental in designing operational metrics to demonstrate progress made towards those goals.
Kick off a Discovery Session
Begin with an assessment of the environment.
Each organization is on their own unique data governance journey. So, it’s important to understand the existing organizational structures, processes, documentation, flows, and technology available. The first step is to identify roles such as an executive sponsor (e.g., Chief Data Officer or Chief Information Officer), a data governance committee, security and privacy support teams, and data stewards. Having the support of these individuals is critical to creating a working governance structure, securing data and systems, and providing ongoing monitoring.
Document Your Data Assets
Develop a shared view of information assets.
To effectively govern and secure data, you need to know all the data and sources your organization has. Start with an initial list that inventories your existing data stores and take advantage of automation tools, such as intelligent data catalogs, to ease this process. For security and compliance purposes, it is critically important to identify the organizational standards and policies that apply to each piece of data. Standards and regulations that apply to the information system such as SOX, HIPAA, GDPR, FISMA, etc., can determine the process for your data governance approach.
Align Your Teams’ Data
Ensure a shared understanding of how data is collected, and metrics defined.
A common pain point is inconsistent use of metrics across teams, or different ways of calculating a metric. This inconsistency often leads to confusion, unhelpful analytics, and other problems. Integrating the use of an easily accessible data catalog and educating users to promote adoption are steps to alleviate this problem. Additionally, where there is a business need for different versions of a metric across organizational units, the data catalog can provide insight into the options and the differences between them.
Monitor Your Practice
Execute and monitor compliance with policies and strategies.
Example actions may include:
- Introducing processes to maintain and evolve the data catalog as the definitive knowledge source for data users to identify data assets, support search, and facilitate appropriate access.
- Working to seamlessly integrate organization security policies and tools into data access and use. For example, using role-based access controls to protect information from unauthorized access.
- Maintaining data integrity by taking steps to ensure that information cannot be changed without proper authorization.
- Investing in platforms and tools to ensure data availability aligns with user group needs.
- Auditing processes to ensure that governance and security approvals and workflows are followed.
Assess and Iterate
Continually review and improve.
It is important to include metrics in the process to validate how the system is meeting the goals set out when establishing the data governance program. Building out a robust data governance strategy is an iterative journey. By using an Agile approach, teams can build a foundation for their organization’s data governance strategy while balancing documentation with actions to deliver results quickly and evolve rapidly with changing business environments.
Avoid A Common Pitfall
In our experience, we’ve continually found that successful data governance program adoption must be business-led. Data governance adds process and additional work to daily operations across the organization. Most data users come from organizational departments outside of IT. These same departments must have responsibility and accountability for governance success—a stake in the game. When inter-departmental collaboration is achieved, the processes introduced are sensitive to data users’ needs and concerns, while meeting governance and security objectives.