Data governance is a topic that can make even the most bright-eyed individual immediately glaze over. It sounds immensely bureaucratic, stuffy, and boring—yuck. Fair. In some organisations, it is all of those things.
That was exactly how I felt. I firmly believed it was just another layer of unnecessary bureaucracy for big organisations. I’d turn my nose up at the mention of data governance. And yet—now I am writing an article about it. Over time, I’ve come to realise that data governance is essential, regardless of the size of your organisation.
In this data abundant era, having a shared language and approach for data is critical for organisational success. It doesn’t have to be bureaucratic, stuffy, and boring.
How can an organisation embrace data driven/informed decision-making if the reliability of its data is constantly in question?
Without a shared language and understanding, reports and analyses are easily undermined, questioned, and their validity compromised. People must trust the underlying data to confidently rely on it for insights.
I’ve seen a number of organisations attempt to implement data governance and here are a few of my reflections through living it:
Channel your inner Simon Sinek: Start with the ‘why’.
As established in the opening of this article—data governance is essential. The organisation needs to understand the value data governance will provide and why it is important.
People will be much more willing to go along for the ride if they understand why it is important.
The most important thing: Clear roles and responsibilities.
At the end of the day someone will need to make the decision and do the work. I’ve seen groups talk in circles for all that should be done but have no ability or resource to implement it. Or even better, no one is responsible for ensuring the work is completed.
Roles and responsibilities can be divided among multiple people, as long as they are clearly defined. The data governance group will have multiple people within it, however most will be subject matter experts (SMEs). The SMEs will be there for their input, not for their permission or approval.
From my experience organisations can be a bit reluctant to make decisions (especially for purpose ones), so meetings end up going nowhere as consensus cannot be made. Especially if there is a difficult person in the mix (there always is).
Clear roles and responsibilities will remedy this. You just need to ensure whoever is the decision maker has the organisational authority to make it.
Organisational authority is required
Data Governance teams need to be cross functional and cross departmental. One of the challenges with this is gazumping (when another manager overrides a decision) from a manager of another department/team. If a SME does not like a decision that is being made they can go up the management chain and get it stopped.
Gazumping will kill the momentum of the group. Why make decisions when they will get overridden by management? Organisational authority anticipates and mitigates this.
Organisational politics exist and need to be anticipated.
A few other thoughts for successful implementation:
This is a pragmatic guide to implementing data governance. I approach this through the lens that people in other teams are well meaning but time poor. Data governance often lacks a sense of urgency, making it easy for other stakeholders to delay or deprioritise.
Additionally there are layers of change management that need to be considered, especially around teams perceiving they are losing control of their data. For some people this will be untenable.
But it needs to happen—data governance needs to be implemented, and these are my pointers to navigate through that environment.
Scope: start small then grow
Start the scope of the group really tight and get momentum and expand out. Pulling all data under a new data governance umbrella may make stakeholders around the organisation defensive and protective as they don’t understand the value of it.
People are often reluctant to give up control—whether real or perceived—unless they see a compelling reason to do so.
If you start with a really focused scope (i.e. defining the account or contact records in Salesforce, or a less controversial object if that one is too spicy), you can build trust, buy-in and enthusiasm from across the organisation.
It is important to start with a scope that involves multiple departments/stakeholders. Building the muscle of making decisions together is what is important here, not necessarily the value of output.
Data dictionary, quality standards, shared language definitions
This work will fall on your team. The dictionary is typically an easy but time consuming project to outwork, however a great place for a governance group to start.
You and your team should put it together and bring it to a governance group for their input to refine and finalise it. It’s always easier for a committee to refine something that is 80 to 90% of the way there, than start something new. You can even add a few red herrings to stimulate the conversation.
Quality standards are a fun one to work through too. People often set high standards—until they receive reports filled with poor-quality data that they then have to fix.
Assume people are often too busy to prioritise work outside of meetings
I’m sure everyone in the organisation is really busy and has their own priorities. A governance group with lots of homework is a group that will not be successful. When it comes to spending time on a campaign to raise more money or provide feedback on a data definition, marketers and fundraisers will always prioritize the campaign.
Giving reasonable timelines for feedback is fair; however, I think it is worth preparing people that decisions will be made without their input to keep things moving. Manufacturing urgency in this way will help data governance work become a priority.
I’ve seen groups die because they were effectively held hostage by subject matter experts that wouldn’t prioritise feedback for the data governance group.
However, you always need to make it clear and easy what you need from people.
Communicate, communicate and communicate
Make it really easy for the rest of the organisation to access the data dictionary and quality standards.
Communicate data quality problems as they arise. One of the data governance groups goals should be to foster a more data valuing and data literate culture within the organisation.
Communication is key to this. Make it fun and exciting!
———
These are my thoughts—more cultural than technical—because data governance is as much a cultural shift as it is a technological one.
If there is any takeaway other than roles and responsibilities is that starting with a small scope though will really help you test and refine what Data Governance looks like for your organisation. It is less pressure to get it right the first go, and a smaller amount of people to test and learn with.
In the end, a well-implemented data governance framework doesn’t just ensure cleaner data—it empowers an organisation to trust its insights and make informed decisions with confidence.
Good luck! I hope this has helped.
