How to make data governance work: a practical guide for utilities

From strategy to execution: what it really takes to implement data governance in utilities

Data governance rarely fails because of technology. The real challenge appears when organizations try to make it work in practice. What separates initiatives that stall from those that deliver impact? The answer lies in how decisions, priorities, and culture align once execution begins.

This second part continues the journey started in the previous article, where we explored the strategic relevance of data governance and its particular impact on the utilities sector.

The next step is effective implementation. This requires organizational readiness, a hands-on approach, and avoiding common mistakes that can compromise results.

Implementation: practical steps and recommended approaches

With the right preparation, organizations can move toward effective implementation. Key elements include:

Defining clear objectives aligned with business priorities and regulatory frameworks.
Establishing roles and responsibilities, including a data governance committee.
Selecting appropriate tools and processes, based on technological and organizational maturity.
Continuous monitoring and adjustment to prevent stagnation or misalignment.

Gustavo Mesa emphasizes that each organization must tailor its strategy to its own context:

“There’s no universal formula; every organizational culture requires a specific approach”.

That’s why he recommends starting with a maturity assessment and prioritizing quick wins, even before formalizing extensive policies.

Common mistakes and how to avoid them

Challenges in data governance are not only technical. Among the most common pitfalls are:

Lack of committed leadership
Overambitious plans at the outset, without practical solutions
Overreliance on technology, without aligned processes or people
No clear metrics to evaluate impact
Weak communication on the purpose of governance policies
Treating all data with the same level of priority

Mesa stresses:

“Test before you formalize. Implement before you document”.

Nicole Halm adds:

“Governance is a dynamic process that must evolve alongside data generation”.

Key indicators to measure governance success

Success in data governance initiatives varies by organizational goals, but some common indicators include:

Data quality: Are the data accurate, complete, consistent, and up-to-date?
Operational efficiency: Is governance reducing redundancies and optimizing processes?
Business impact: Does it contribute to strategic objectives, like increased revenue, better decision-making, or higher customer satisfaction?
Regulatory compliance: Does it ensure adherence to local standards and regulations?
Accessibility and usability: Are data available to authorized users at the right time, without compromising security?
Adoption and organizational culture: Is governance accepted and embedded across teams and collaborators?

When can a data governance policy be considered successful?

According to Gustavo Mesa, success is achieved “when the organization meets its goals, acknowledges governance as a process enabler, and addresses previously overlooked issues”.

He points out that, commonly, “80% of project time is spent on data acquisition”, often by non-technical staff, which “increases time and errors”.

Therefore, he proposes to measure success by reducing that time, allowing teams to focus on their core responsibilities. In other words, the strategy should enable organizations to “shift the conversation from data to decisions”.

Practical tips to get started

Define a clear objective
Assess the current state of data and processes
Secure active support from senior leadership
Form a dedicated data governance team
Start with a limited-scope pilot project
Prioritize simple, practical policies at the outset
Communicate benefits and train teams
Choose technologies aligned with your needs
Measure results and adjust continuously

Data governance at Quanam: our experience

Gustavo Mesa shared that Quanam implemented a metadata catalog to classify reports and define business terms, responding to frequent queries from the Chief Financial Officer (CFO).

He added that this catalog, which maps data lineage, helps streamline knowledge transfer and system maintenance.

“One might assume a CFO wouldn’t typically lead a data governance strategy. But there are always opportunities to deliver value in unexpected areas, responding to real needs that optimize processes”, he concluded.

A success story: utility in Florida, USA

Martín Cal, Data & Analytics Project Manager at Quanam, shared the experience of working with a public utilities company in Orlando, Florida:

“In our work with the client, we implemented KeeDATA, a data catalog based on CKAN. This catalog registers all data assets daily, including 6,363 objects in Oracle databases, 912 Control-M job objects, 637 objects in the Cognos production environment, 17 Talend jobs, and 17 external files used in jobs as CSVs”.

Cal highlighted that:

“The catalog enables users to search concepts, measures, and tables, check information availability, understand how data is loaded and distributed across the organization, and identify data owners and contact points to request access”.

*Control-M is OUC’s corporate tool for scheduling and automating script execution.
*Talend is an ETL tool whose processes are scheduled and executed from Control-M

Beyond frameworks and tools, data governance is something that evolves over time through practical decisions and continuous learning. As Halm and Mesa point out, it is not about achieving a perfect model from the start, but about making steady progress, understanding each organization’s context, and delivering value early on.

If you are navigating this challenge or exploring where to begin, we invite you to connect with us. Real conversations are often the starting point for building a sustainable path.