Compromised data can create problems across the enterprise, undermining business processes, raising regulatory and compliance concerns and generating adverse publicity. Inconsistent data can have a similar impact, hampering day-to-day operations and analytics efforts. Data management and protection efforts must be equally broad, with data governance roles that span the organization and are overseen by strong leadership.
"Data governance efforts must be a cross-functional, centralized effort, not a set of standalone projects that take place across the enterprise," said Yasmeen Ahmad, vice president of global business analytics at data platform vendor Teradata.
As part of data governance programs, organizations need to craft policies and procedures that consider legal issues, requirements around speed to market, data security imperatives, platform and tooling needs, and more. These programs are a team effort, headed up by a data governance council and involving input from not only data managers and data stewards, but also business leaders and regular business users.
More than just IT to govern data
The expansion of data governance roles in the enterprise reflects the reality that data consistency, accuracy, integrity and safety are no longer the province of IT alone.
Data governance principles and practices must be understood and implemented by both the data gatekeepers in IT and the users of the data on the business side. Data governance councils must include representatives from across business operations and IT, and they should consult regularly with the legal department for ongoing counsel on changing regulations and legal concerns related to data usage.
The broad spectrum of data governance roles prevents tunnel vision on how to govern data, according to Teradata's Ahmad.
Often, IT will focus on security, data modeling and compliance, whereas the business will bring considerations regarding data quality, access and agility, she explained.
For instance, IT teams are integral in managing the underlying infrastructure, tools and automation to ensure that all applications are supported and that data governance policies are implemented. Business users, on the other hand, are the subject matter experts and can assign value to how data is used to meet goals and solve specific business problems, an important consideration in framing data governance programs.
"This is why it's imperative that business users become key stakeholders in data ownership," Ahmad said. Their roles in data governance include being tasked with creating common data definitions and metrics for data quality, as well as having a hand in enforcing governance policies and processes.
Data governance roles: Leadership
Ana Maloberti, a big data architect at IT consultancy Globant, said a data governance program needs to start with an executive data governance sponsor who makes sure the program has enough resources and understands the overall vision of the program. Sponsors will align the council's data governance policy with higher-level business strategy objectives.
In large enterprises, the sponsorship role is sometimes filled by a chief data officer (CDO) or a data governance manager. In this capacity, a CDO or governance manager typically oversees communications on the data governance process and monitors the various initiatives to ensure the overall program stays on track and within budget. They also help coordinate the activities of the enterprise-wide data governance council.
Andy Hall, an application security engineer at FormAssembly, a web forms tool vendor, said this data governance role requires a marriage of managerial competency and technical knowledge. This person "sits at the intersection of C-suite strategic planning, departmental working groups and data stewards, similar to the way the editor of a publication sits between strategically-driven superiors and writers," Hall said. As such, data governance sponsors act as conduits between the strategic decision-makers at the top level, the working groups across the enterprise and the data stewards in the trenches. They can even serve as contact points for external entities concerned with data governance in their organizations.
Data governance roles: Data governance council
The data governance council is a steering committee within the organization that oversees the development of the organization's data governance program at a strategic level, Hall said. It can include technical, business and legal experts from across the company in order to address various elements of the program. Typical members also include data stewards from each domain to represent the interests of their domain and the interests of the governance program at the implementation level.
Additionally, one or more C-suite executives may be present to represent the strategic element of the program. The CDO, if there is one, helps facilitate development of a concrete plan for oversight and implementation based on the council's decisions. There may also be additional members, such as data analysts, to identify and report on trends within the organization and how those trends would shape decision-making at the strategic level.
Data governance council responsibilities
The data governance council regulates and gives strategic guidance for the overall care of data. Globant's Ana Maloberti said responsibilities for this data governance role include:
- Setting goals for the data governance program, identifying initiatives that need data governance, overseeing program implementation progress and assigning roles to data stewards and owners.
- Creating and approving data standards, policies, business rules and purchases of tools to support them.
- Resolving problems and issues that data stewards aren't able to resolve on their own.
- Communicating about the program and promoting the importance of data governance.
Data governance roles: Data stewards
The rubber meets the road in this data governance role. Data stewards tend to be domain-knowledgeable individuals who thoroughly understand how the organization's data governance policies apply to the data under their stewardship, Hall said. Data stewards can act as touchpoints for regular users on various practical data governance issues. However, their role mainly involves working behind the scenes to ensure data quality and engender trust in the data under their stewardship.
Additionally, data stewards ensure that the data under their watch is compliant with the ever-evolving landscape of data law, whether locally, regionally, nationally or internationally. As a result, data stewards are a resource for the compliance requirements in their domains.
Start with small victories on data governance
Maloberti said data governance roles and responsibilities should reflect the ways the company operates rather than be forced on the organization. She suggested that enterprises start by considering the existing business department structures and identifying who makes decisions, since data governance involves making decisions using data. It's important to recognize that many functional areas may cross several structural business units. Data comes from the execution of business, so following the business flow, identifying data owners and assigning stewards should be a common practice.
Although the goal is to extend a data governance program across the enterprise, it's important to start with incremental steps to gain momentum. "It's best to achieve small victories from the outset and build momentum rather than attempt huge, sweeping changes at once," Hall said. "In this sense, it's best to think of data governance as an ongoing practice rather than a project with a set deadline and deliverables."
It's also important to develop data quality metrics and other ways to quantify the impact of a data governance program in order to determine when it's succeeding or veering off track. "Effective data governance itself is data-driven," Hall noted.
4 questions to consider when building a data governance program
Kristina Bergman, founder and CEO of data privacy tools provider Integris Software, said the success of a data governance program depends on saying yes to the following four questions:
- Will the data governance council and team have the support and backing of the executive management team?
- Can the council develop clear strategies and goals that will impact business innovation, efficiency and customer support?
- Will the data governance team have the budget to acquire tools and services for the implementation and management of the program?
- Does the organization have subject matter experts who can support the tactical and operational challenges of a data governance program? If not, new staff or professional services will be required.