ADG 5: How to start with DG

Alex Leigh

September 01 2025

4 min read

At the start of 2025, we ran the Agile Data Governance (ADG) webinar series.  In this collection of short articles, we’re going to look back at each one to summarise the most important takeaways. It’s been fun going back after a few months to see if there’s anything we missed or even forgotten!

ADG 5 - Starting small with Data Governance

Overview:

Practical steps to initiate data governance on a small scale.

Key Points:

• Identifying quick wins.

• Building a pilot project.

Starting small before smartly scaling: A practical Data Governance (DG) implementation

In previous blogs, we’ve made the case for introducing a DG capability using an operating model to implement and evolve that capability. This time round, we’re going to talk about one of the most critical phases – how to successfully launch.

Firstly, what not to do! Attempting to launch though a massive programme of work across many data domains is doomed to failure. If you think of introducing DG as “re-tooling an aeroplane in flight”, then attempting to rebuild that plane with a single attempt is not a good idea. So don’t do that!

The rest of this blog sets out pragmatic steps focussing on agility, people first and sustainability

Step 1: Start where you are.

DG implementation begins with understanding that this capability can be introduced in a more manageable, impactful way. Institutions don’t need to activate every component or data domain of a DG operating model from day one. Instead, you can choose from three practical starting points:

• Quick Wins: Target areas like data returns, funding sources or strategic reporting where improvements in data quality can deliver immediate value. However, we are always a little reticent about this option as quick wins are rarely as simple as they may seem.

• Pilot Projects: Support new initiatives or reboot existing ones by embedding governance practices that ensure clean, usable data. Right now, this is what we see happening most often especially with the introduction of and migration to cloud based solutions.

• Process-Focused Improvements: Enhance high-volume processes such as enrolment or timetabling by improving data quality before and after the process. This is what we wish we see more of, but it’s a tough environment to make the case to do so right now.

• Quick note: when working within a big project or programme, It’s vital to avoid becoming a “project resource” that disappears when the initiative ends. DG needs to stand on its own, with clear value and the ability to scale outside of the project.

Step 2: Make it reusable.

Wherever you start, your governance capability must be implemented in a way it can be reused across the institution. This means:

• Assigning, embedding and supporting data ownership and stewardship roles. Creating a community that bypasses hierarchies is the key here.

• Creating data quality issue logs to prioritise and track issues. Theming multiple issues into “one fix” can bring new impetus to problems previously “too hard to solve”

• Implementing sustainable solutions—not one-off fixes. With a bit of thought, everything is reusable! Think of building up a library or “toolkit” of solutions that can be deployed anywhere.

Step 3: People first.

I know we keep saying this, but we are in the people business! DG Success depends on strong relationships, trust, and communication. One of the most effective ways to begin is by encouraging individuals to take responsibility for data—without overwhelming them with formal roles or jargon. It has to be about what’s in it for them.

So language matters. Terms like “data ownership” can feel intimidating. Reframing the conversation around data responsibilitieshelps reduce resistance and make DG activities more accessible. The aim is to shift culture gradually, using inclusive language and building partnerships rather than enforcing compliance. Repeat after us: “We are not the data police”!

Wide engagement is critical. Whether working with your sponsors, university leadership teams, or project managers, you need to be speaking specifically to that audience. So be flexible, empathetic, and willing to adapt.

We were working at a university where two project managers responded very differently to governance—one embraced it, the other saw it as a threat. Success came from adjusting the approach to align to different needs.

Step 4: Now the foundations

Regardless of your starting point, we’re going to have to lay some foundations:

• Define Scope and Domains: Identify which areas (e.g., Admissions, Finance, HR) are involved. We recommend more than one and less than three! You need enough to show the benefits of cross silo working, but nothing so big it swamps your early capability.

• Map Roles to Needs: Determine which governance roles are required and who can fill them. Be flexible- it will change over the first year, so it doesn’t need to be perfect but it does need to start! You absolutely will need the role of the DG manager though – more on this in a later blog.

• Tailor the Operating Model: Start with the basics—owners, stewards, and the DG manager. You don’t need, for example, a full steering group or even a full complement of stewards from day one.

These activities help establish governance as a distinct capability that can be reused and scaled. They also support ongoing operations, such as data quality management and business glossary development which can follow shortly after. There also needs to be a “landing point” for DG which for most universities can be a simple SharePoint of Teams site.

Step 5: Scale and evolve

No two implementations of DG are the same. The priorities and culture of your institution must be reflected in your approach. This means identifying where you could most successfully pilot DG and which parts of the operating model might be the simplest to implement informally. It’s always a balance between making DG a “real thing” and not over-reaching.

It won’t be a linear journey! You must be prepared for setbacks and react appropriately. We’re attempting to change the university’s relationship with data one person at a time. That’s not something you can achieve overnight. But if you start small and be very clear you’re building a sustainable capability not a project resource, you’re heading in the right direction.

What’s Next? Scaling and sustaining

With these foundations in place, the next phase is significant scaling. This can happen in two ways:

• Deepening governance within a single data domain

• Expanding governance across multiple areas

  
  

Future blogs will explore proven approaches and common pitfalls when scaling your DG capability out or up.

Final thoughts: It’s a marathon not a sprint.

Data governance isn’t about launching a formal programme and expecting overnight results. It’s about starting small, building relationships, and creating reusable capability. It’s about helping people see the value of doing things differently—and supporting them in that journey.

On that journey, it’s useful to remember to:

• Explain the benefits to the individual before explaining why the university needs it!

• Use language that feels authentic and non threatening.

• Build trust with people before introducing a framework/operating model

• Be flexible and open to change and accept there are going to be setbacks.

• Focus on helping, not policing

DG is definitely a journey, sometimes without an obvious destination. But with the right approach, it’s one that can transform how institutions manage, use, and value their data.