Approachable and Impactful Data Governance: Insights from DIR Discover
Last week, Applied Curiosity team members Sid Atkinson and Renee Veloz had the opportunity to present at DIR Discover 2025 on a topic that’s close to our hearts: making data governance more approachable and impactful (please try to tell us we’re not data nerds, I dare you).
A Challenge We’re All Facing
Data governance often feels overwhelming; there is so much we could be doing, so how do we filter through the noise to pick a place start? And just for fun, let’s add the pressure to build AI and ML capabilities while also balancing privacy, security, and explainability, and many times we find ourselves stuck, wanting to innovate but uncertain how to do so.
Our presentation, “Approachable and Impactful Data Governance,” aimed to reframe this challenge and offer a practical path forward.
The Core Message
Traditional approaches to data governance starts with framework and scoring the organization across many different data systems, teams, and functional silos. This produces an aggregate score, with an ensuing list of recommended activities and practices to perform across all teams, functions, and silos all in pursuit of a better score on the framework. Instead, we advocated for a fundamentally different mindset: start with outcomes that people experience, ideally outcomes that your team members or constituents are greatly impacted by.
Rather than feeling guilty about what you’re not doing, focus on what would make the biggest difference for your right now. Data governance should enable your mission, not slow it down.
The key pivot? Sustainable governance isn’t about perfection, it’s about consistent, incremental progress aligned with real business needs. At the heart, approachable data governance starts with change management, and recognizing the need to consistently and constantly introduce change.
Frameworks are critical, but they are the legend on your map, not the map itself.
Key Insights Shared
Finding an Impactful Starting Place
The most common question we hear is: “Where do we even begin?” Our answer: start with pain points, not policies. As humans, we engage best when the outcome and benefit is clear.
We recommend organizations ask themselves three questions:
What is causing the most friction right now? (e.g., teams can’t find the data they need, regulatory concerns are blocking new initiatives, or data quality issues are undermining trust)
Where is there already momentum or interest? (e.g., a department eager to pilot new analytics, leadership asking data-driven questions)
What would demonstrate quick, visible value? (e.g., solving a recurring data issue, enabling a stalled project, etc)
The intersection of these three questions often reveals much, and in a separate blog post, we’ll provide an incredibly useful tool and explainer to help rank and prioritize multiple opportunities.
Iterative Governance: Thinking Big, Starting Small
We introduced the concept of “governance MVPs” (Minimum Viable Products) starting with the smallest possible business use case that still delivers meaningful value.
The approach follows a simple cycle:
Define a narrow scope (one use case, create one team)
Implement lightweight processes (simple documentation, basic data quality checks, clear ownership)
Measure impact (in our Transit example, it was route optimization,% ridership, customer sat.)
Learn and adjust (what worked, what didn’t, what’s needed next)
Expand incrementally (add more use cases, another team, more sophisticated practices)
We emphasized the importance of maintaining a long-term vision while resisting the urge to solve everything at once. Create a roadmap that shows where you’re going, but focus execution on your current iteration. This builds credibility and momentum while avoiding the overwhelm that kills many governance initiatives.
Securing Buy-In Across the Organization
Getting cross-functional support requires meeting different stakeholders where they are and speaking their language.
For executive leadership, we emphasized framing governance in terms of strategic enablement. Show how governance supports their priorities—whether that’s digital transformation, AI adoption, operational efficiency, or regulatory compliance. Use their language and metrics, not governance jargon.
For technical teams, focus on how governance reduces technical debt and firefighting. Highlight how clear data ownership and documentation reduces the time they spend answering questions or troubleshooting data issues. Position governance as a way to work smarter, not a bureaucratic burden.
For business users, demonstrate how governance helps them do their jobs better—finding data faster, trusting it more, and spending less time on manual data wrangling. Share concrete examples of pain points that governance would alleviate.
One key tip: identify and cultivate champions in different departments. These advocates can help you understand local needs, pilot initiatives, and spread adoption in ways that top-down mandates never can.
Addressing AI/ML Governance Concerns
The rise of AI and ML has made data governance more urgent but also more complex. Organizations want to innovate but are rightfully concerned about doing so responsibly.
We discussed how strong foundational data governance practices directly support AI/ML initiatives:
Data lineage and documentation enable you to explain where AI training data came from and how it was processed, critical for both technical debugging and regulatory compliance
Clear data ownership and sensitivity classification help you identify which data is appropriate for different ML use cases and who needs to approve its use
Data quality processes ensure AI models are trained on reliable data, reducing bias and improving model performance
Usage logging and monitoring create the audit trails necessary to demonstrate responsible AI practices
Rather than treating AI governance as a separate initiative, we advocated for extending existing governance practices to explicitly address AI/ML considerations. For example, if you have a data sharing approval process, add a checkbox for “Will this be used for AI/ML?” that triggers additional considerations around bias, explainability, and monitoring.
The key message: you don’t need perfect governance before you can start experimenting with AI, but you do need enough governance to understand and manage the risks. Start with low-risk use cases while you build out more robust practices.
Audience Response and Questions
The presentation generated excellent discussion, with several themes emerging from audience questions:
Many attendees resonated with our observation that frameworks are useful, but the framework is not the goal. Serving your customers and constituents is, and you have to align all change efforts and governance activities to how they positively impact people.
Several people asked about how to prioritize use cases and initiatives. We will add a new post soon that provides a practical and effective ranking and prioritization strategy.
Key Takeaways
If you take nothing else from this presentation, remember these points:
Start with outcomes, not frameworks: Focus on solving real problems and demonstrating value rather than implementing comprehensive governance from day one.
Permission to start small: You don’t need full organizational buy-in, a big budget, or a dedicated team to begin. Start with what you can control and let success build momentum.
Governance enables innovation: Position data governance as what makes AI/ML and other data initiatives possible, not what slows them down.
Iterate and learn: Build governance practices incrementally, learning and adjusting as you go. Sustainable governance is a journey, not a destination.
Meet stakeholders where they are: Different audiences need different messages. Tailor your communication to what each group cares about.
What’s Next
The conversation around approachable data governance doesn’t end with this presentation. We’re continuing to develop resources and case studies that demonstrate these principles in action.
If you’re working on making data governance more approachable in your organization, we’d love to hear about your experiences—what’s working, what’s challenging, and what quick wins you’ve discovered. These real-world insights help refine our thinking and provide valuable examples for others facing similar challenges.
Interested in discussing these approaches further or exploring how they might apply to your organization? Feel free to reach out, we’re always happy to connect with others working to make data governance more practical and impactful.