Thoughts on making data governance more than just a checkbox

Everyone talks about data governance. Frameworks, policies, stewardship roles, governance boards, you name it. Yet, time and again, organisations struggle with poor data quality, inconsistent definitions, and slow decision-making.

The harsh truth is that governance seldom works by itself. Policies do not fix bad data. Tools do not fix misaligned incentives. Governance fails when it is treated as a checkbox exercise, rather than a practical enabler of business value.

Start with real problems, not ideal frameworks

I dislike the approach where governance begins with a model or a policy. It does not. Governance begins where data actually hurts the business:

• Decisions delayed by missing or inconsistent data
• Reports full of errors that nobody trusts
• Processes blocked by unclear definitions or missing responsibility

Instead of asking “How many policies do we need?” ask:
Where do data issues actually hurt our business, and who feels it?

Governance should solve tangible problems, not exist for compliance theater.

Make roles meaningful

Data Owners, Stewards, Custodians, they sound great on org charts, but too often they are symbolic titles.

To make governance work:

• Give roles real authority to make decisions and enforce rules
• Empower them to resolve conflicts, escalate issues, and remove bottlenecks
• Recognize that different units have different agendas, and governance should bridge those agendas, not impose a single view

Keep it lightweight and iterative

Big, rigid frameworks kill momentum. Instead:

• Start small with specific, measurable actions
• Show immediate benefit, a few key processes corrected, reports trusted, decisions faster
• Expand as adoption grows, do not wait for perfection before acting

Governance should be a living tool, not a static document buried on a shared drive.

Measure outcomes, not activity

A common trap is tracking activity, like number of policies published or meetings held. Real governance measures:

• Reduction in errors and duplicates
• Faster access to reliable data
• Increased trust and usage of data in decisions

Focus on impact, not appearances.

Make governance collaborative

Finally, governance works best when it is a team sport:

• Involve IT, business units, compliance, and leadership
• Facilitate dialogue, acknowledge different priorities and agendas
• Build shared definitions of success rather than imposing one-size-fits-all rules

Remember, a governance board is only useful if it solves problems for the people using the data, not just creates more documents.

Ownership is key

Clear ownership is what makes governance actionable. Data owners are accountable for the quality, accuracy, and usability of the data. They make decisions, approve changes, and resolve conflicts. Without ownership, issues linger because everyone assumes someone else is responsible.

Ownership also helps bridge different agendas. In shared projects, owners act as the anchor point, balancing IT, business, compliance, and leadership priorities. Combined with data stewards, ownership ensures governance is applied consistently and improves over time, turning frameworks and policies from theory into practice.

At the end of the day

Data governance is not about frameworks, org charts, or meetings. It is about enabling better decisions, smoother processes, and shared accountability.

Treat governance as pragmatic, outcome-driven, and collaborative, and it stops being a checkbox exercise. Only then does it become the backbone of reliable data and informed business decisions.

Governance fails when it is theoretical. It works when it addresses real-world pain, respects different agendas, and produces measurable impact.