Rejuvenating data governance in the public sector

Contributors: Kristin Yuan, 
John Candeias, Partner, Data and Analytics, EY Canada

Public sector organizations should consider implementing these leading practices to advance their data governance efforts.

In Brief

• Data governance programs in the public sector are important for organizations to become data driven, but they do not always succeed.
• Organizations make three critical errors: they try to solve every problem at once, they fail to empower committees and stewards with practical authority, and they fail to practically define the roles and responsibilities of data stewards.
• Successful data governance programs are aligned with the organization's strategic business priorities and empower data professionals to work most efficiently.

There is no doubt that data governance¹ is foundational for public sector organizations to become data-driven organizations, as outlined in the 2019 Data Strategy Framework for the Federal Public Service in the Government of Canada². But the efforts to implement data governance have mostly failed to yield tangible business outcomes. Data access and data sharing remain a challenge. And poor data quality continues to impede organizations’ ability to adopt analytics in their decision-making processes.

Three common mistakes can hold public sector organizations back from optimizing their data governance programs. Here, we outline those mistakes and share our recommendations for leading practices that government agencies can follow to succeed with their data governance efforts.

Mistake 1: Taking a “big-bang” approach without linking directly to strategic initiatives.

Most public sector organizations are dealing with multiple data problems including poor data quality, lack of data accessibility, missing business glossaries, siloed applications, lack of data literacy, and challenges in establishing a single version of the truth. Moreover, the significant increase in volume, variety and velocity of data collected by public sector organizations have amplified data challenges.

Organizations often attempt to address all these challenges at once as they’re all connected. While it may seem feasible at first, once organizations start the implementation, they quickly become overwhelmed by the scope and scale of the challenges. Then, when data governance initiatives fail to show results after several months or a year, organizations begin to lose interest.

Recommendation: Bite sized approach - Support building one piece at a time delivering incremental value while building towards a strategic vision

Organizations should focus on linking data governance to strategic initiatives. For example, if ERP financial transformation is a key initiative, organizations could use it as a starting point to establish data governance, perhaps by focusing on data quality for financial data. Once the data governance initiative starts to yield tangible business outcomes, the organization can find another catalyst to sustain and expand the program to include additional data domains and data challenges that need to be addressed.

Mistake 2: Preventing data governance committees from making decisions.

A common misperception is that data governance committees must involve as many people as possible to be successful. During committees’ monthly or bi-weekly meetings, participants are invited to discuss their data challenges and their data wish lists. These large committees quickly become less effective as participants spend meeting time sharing their data concerns but don’t focus on making decisions that advance the data agenda.

Not only that, but data governance committees often don’t have sufficient authority to make important decisions because most decisions must be escalated to the next level. Over time, data governance committees are transformed from decision-making bodies that make real progress into communities of practice.

Recommendation: Organize for success - Give data governance committees decision-making authority.

Data governance committees should be limited to key individuals required for decision-making purposes. Leverage working groups and community of practices to discuss and find solutions to solve data problems. Organizations should also clearly define the mandate and scope of data governance committees. Committees should have decision-making authority on data matters at the committee level. And to keep the conversation focused, committees should set expectations on discussion items and decisions that need to be made prior to each meeting.

Mistake 3: Failing to practically define the roles and responsibilities of data stewards.

Data stewards are a critical component of data governance initiatives, as they are instrumental in identifying and solving data problems in the organization, leading to increased trust in data quality and accuracy. When establishing data governance committees, public sector organizations typically create a charter to define the roles and responsibilities of data stewards. Though charters are important documents, they often fail to provide practical examples of what data stewards should do in their day-to-day work, leaving the role poorly defined in practice.

Additionally, the role of data steward has not yet been formalized in the public sector. Most data stewards at public sector organizations have full-time roles, with the assumption that they can perform their duties as data stewards in their spare time.

Recommendation: Stop underestimating the role of data stewards. Leverage technology and establish a data literacy program to enable them to make a real impact.

It’s important to make data stewards’ roles and responsibilities as practical as possible and enable them to be successful. Implementing data governance technology is an effective way to operationalize data stewards’ responsibilities and enable their accountability and efficiency in the role. These tools use workflows to formalize data stewardship processes to empower data stewards to make day-to-day decisions on data. In addition, organizations should invest in data literacy to support data stewards so that they have the necessary knowledge and skillsets to get the job done.