What is Data Governance?
Data governance refers to the collection of practices, policies, and roles related to the effective acquisition, management, and utilization of data—ensuring that the data provides as much value as it can within an organization. Data governance confirms the quality and security of a business’s data across the entire organization, determining who can use what data and when. Often, data management and data governance are used interchangeably, but this is incorrect. Data management refers more to the technical management of data, whereas data governance refers to the policies of managing data within an organization such as who can use what data and when.
A company’s most important asset is data. This famous statement is a clear indication of the value of data. With increasing digitalization, every business has access to huge volumes of data. Rational use of this data could be a deciding factor in the success of an organization. For this, companies need to make their data clean and reliable. That’s precisely what data governance does.
A data governance strategy is a necessity for any organization that works with big data because it lays out how your business benefits from consistent, common processes and responsibilities. It also highlights the data that needs to be carefully controlled in your environment. Data governance will automatically ensure that retention requirements (e.g. history of who changed what information and when) are met to ensure compliance.
A successful data governance framework results in high-quality data that will help organizations to make smarter business decisions.
In a nutshell, implementing data governance within an organization involves many tasks:
- Defining the type of useful data and its quality standards
- Defining methods for acquisition, cleansing, storing, and retrieval of data
- Assigning roles and defining responsibilities for data management
- Creating policies and workflows for the usage of the data
- Continuously monitoring of data practices and collecting feedback
Data governance applies to a wide range of organizations. For example, an e-commerce company might own huge volumes of customer data. This data includes purchase history, customer preferences, and reviews. With an efficient data governance system, the e-commerce company can standardize the collection, storage, and retrieval of this valuable information. Then, various departments across the organization can utilize this data to make smarter decisions. For example, the marketing department can use this data to create highly-personalized ads and product recommendations to customers. With data governance, the e-commerce company can ensure the accountability of the data. It enhances data privacy and reduces data breaches.
Why Do Businesses Need Data Governance?
Even though most organizations do have volumes of data stored digitally or physically, most of the data is in a non-standardized format. Further, organization’s cannot always be sure of the reliability of data due to age, the source, etc. Employees or business leaders often hesitate to rely on this data for decision-making due to worries about data quality. Data governance is a process that makes an organization’s data reliable. It also makes sure that high-quality information is available across the organization. It empowers every department to make decisions based on this data. Data governance also drives the digital transformation of a business.
How Can Organizations Implement Data Governance?
Implementing data governance for a big organization might seem like a complicated task. The huge volume of data, the disparate systems, numerous people involved in the creation and consumption of data—all these make data governance a challenging task. It’s best to approach data governance one step at a time.
Phase 1: Do the Groundwork for Data Governance
As a groundwork for data governance, it’s essential to start from the very basics by answering the following questions:
An organization should first define the vision and mission of its data governance plan. An organization must also define the goals of the data governance program—increasing revenue, better decision-making, or transparency. Also, it should determine how to measure the success of the program. A clear vision helps employees and other stakeholders see how this data governance initiative is going to impact their day-to-day work life and how it is going to help them.
Assigning roles and responsibilities is a crucial step. This step defines who will be primarily responsible for different tasks involved in the implementation of the data governance framework. Often, organizations adopt a three-tier approach to set up data governance teams. The steering committee, data governance office, and data governance working group are three main components in this approach. Together, these groups decide the next steps in the implementation of the data governance framework.
The data governance teams need to begin with an analysis of the current data assets of the organization. Enormous volumes of data flow in and out of organizations every day. Trying to bring all this data under the purview of the data governance framework may not be a good idea. Hence the data governance teams must choose a few specific data assets to include in the framework. For example, an ecommerce company can choose to include only the purchase history into the data governance plan initially. The next step is to define acceptable data formats and draw up data workflows and policies for the entire organization. This is a blueprint for a phased data governance implementation.
Phase 2: Implementation of the Data Governance Plan
Step 1: Ensure the Availability of Data
The data governance teams should ensure the availability of specific data assets they want to standardize and control. In large organizations, data is spread across different information silos like customer care systems, enterprise management applications, sales records, and even partner systems. All this data should be readily available in one place. Organizations might need to design an integration mechanism for these distributed data assets.
Step 2: Ensure Data Integrity for Implementing Data Governance
Data assets that are clean, standardized, and reliable are the crucial component of the data governance framework. To find the definition of clean and reliable data, start by asking teams that consume the data on a daily basis. Ask them which data format makes the most sense to them. Based on their input, embark on a multi-step data enhancement workflow as below.
- Profiling: Only some parts of a data asset are useful for business decisions. For example, the location of a customer may be relevant, while gender isn’t. Start with defining the crucial components in a data asset. Then eliminate all unimportant ones.
- Parse and standardize data: One of the biggest challenges to data governance is the diversity in the data formats. Starting from the naming conventions, to the attributes of the data, there might be several disparities. The data governance framework should include technology to parse and standardize the data. It may consist of adding data tags, normalizing attributes, and standardizing the naming conventions.
- Enrich the data: The data governance teams should work towards enriching the data assets. This may involve combining two or more parts of the data in a single place. It also involves augmenting the data with complementary information and metadata.
Step 3: Enforcing Accountability and Adherence to Data Policies
Data governance efforts are not only limited to the members of the data governance teams. For the success of a data governance plan, the entire organization needs to contribute towards it. Each specific data asset should have an owner who is accountable for the integrity of that particular data. These owners, with the help of policies and workflows, should make sure their data asset maintains high quality at all times. This step also calls for a change in the organization’s data culture to embrace data governance. Data governance isn’t just a one-time project. It’s an ongoing process.
Step 4: Continuous Feedback and Monitoring
The systems and workflows for data governance need continuous monitoring and feedback. This is crucial as the data governance framework is a hybrid system that involves people and technology. While the technology needs upgrades and bug fixes, people need constant motivation and reminders. The feedback system is vital to assess if the data governance efforts meet the success criteria and goals. If not, it indicates that certain adjustments in the data governance framework are necessary.
Implementing a data governance framework is an iterative process. It can only be improved through continuous monitoring and feedback.
Data Governance Teams and Roles
Data governance is a highly people-centric concept. Multiple teams across the organization are essential for the implementation of a data governance framework. The teams include:
- Steering committee: This is a high-level team that drives and oversees the data governance efforts. The steering committee often comprises of senior executives from finance, marketing, sales, or production. The committee must include at least one stakeholder from all the top-level organizations within a company. The committee involves leadership who have the authority to allocate budget, make policies, and push projects up the priority list.
- Data governance office: This middle management committee gives guidance to the data governance efforts. The main roles within this team are digital governance lead, a coordinator, and technology experts. This team works together to draft the data standardization policies, data governance workflow, and processes. They also work with IT to address technological challenges related to the implementation of data governance.
- Data governance working group: This group works as per the recommendations of the data governance office. This group usually includes data owners, a data quality lead, data stewards, data architects, and analysts. Data governance is indeed a huge undertaking that demands cooperation between various teams and individuals.
What are the Benefits of Implementing a Data Governance System?
There are many benefits to implementing a data governance system:
- Better business decisions: Data governance gives decision-makers access to clean and reliable data. It enables them to make consistent and confident decisions.
- Increase efficiency of the workforce: With standardized data available throughout the organization, teams can avoid duplication of efforts. Everyone benefits from the data framework, and it increases overall efficiency.
- Data security: Data governance creates a higher level of accountability around data. Each data asset will have an owner and a definite life cycle. This reduces the chances of a data breach or data misuse.
- Data monetization: With data governance, businesses can unleash the power of valuable data they collect and generate. Reliable, standardized, and classified data can be used for new revenue streams.
- Avoid data-related violations: With stringent data protection regulations, many organizations find it challenging to control their data flow and often end up violating data protection regulations. With a highly mature data governance framework, every data asset is accounted for, managed, and owned.
What are The Challenges of Data Governance and How Should They Be Addressed?
Challenge 1: People and Perspectives
Often, in organizations, the true potential of high-quality data is underestimated. Employees are often busy with the operational activities and overlook the benefits of the data governance plan in the long run.
Solution: Have a clear vision and mission for data governance.
For people to appreciate the need and benefits of a data governance plan, they should have a thorough understanding of why it is done. The mission and goals of data governance should be practical and simple—rather than based on abstract terms.
Challenge 2: Over-dependence on Technology Teams
Often, organizations push the data governance efforts towards the technology-centric teams like IT. While it’s true that the IT team may be involved in the data governance efforts, they are not the only stakeholders in a data governance strategy.
Solution: Combine business and IT efforts towards data governance.
Take away the burden of implementing and maintaining the data governance plan from IT and involve business teams. While IT provides the technology infrastructure, the business needs to define and enforce processes and workflows to get the maximum business benefits out of the data.
Challenge 3: Data Silos and Inconsistent Implementations
Data silos are a big challenge to data governance. In many organizations, data might be owned by different teams and stored in various formats. Even with a data governance framework in action, some teams might fall behind and fail to adhere to the standard.
Solution: Data decentralization and a cultural shift.
The crucial step to achieve data governance is moving the data from silos and getting it into a centralized data governance framework. Also, data governance isn’t just a project but an ongoing activity. Hence, there should be a cultural shift in the organization that favors high-quality data.
Challenge 4: Data Governance Practices and Data Privacy
One of the benefits of the data governance framework is that across teams, high-quality data is available. However, if not properly managed, this universal access to information could lead to data privacy issues.
Solution: Strict ownership and accountability of data.
Data governance systems should segregate different levels of data with appropriate access rights. For example, the medical records of the customers in healthcare should be considered sensitive data. An organization that utilizes the customer data through a data governance system should properly disclose to their customers what kind of data they are storing and using. There shouldn’t be any secondary usage of data. Data that is collected for one purpose should not be used for an entirely different purpose.
Data Governance Resources
Data Governance, how to move from Strategy into Practice
Organizations are realizing that to drive business value from data, they need to understand the...
Five Ways Data Virtualization Improves Data Governance
Robust data governance is more than checking the data security and privacy boxes. Properly managed...
Data Catalog - Governed self-service for your analytics teams
Data scientists, analysts, and other data consumers want access to all data assets to support deeper...
Gaining Business Value from Governed Analytics and Discovery
A TDWI Checklist Report written by David Stodder, BI Research Director at TDWI Visual analytics and...