The duality of data governance 

By Stefan Wijers

The topic of data governance does not seem to excite many when you want to invite them for a conversation about it. And frankly, at Anderson MacGyver we kind of agree, especially our pragmatic, passionate MacGyver heart disengages when we hear ‘data governance’.  

The topic has gradually moved into a corner that interests few, and even with the hype of AI, data governance seems to not be the topic that reaches the board room or truly gets things moving. And yet, our Anderson brains, desperately want to get the governance in place to make sure organizations are compliant and can get their data foundations ready for creating business value with AI.  

Curious what we see happening on the topic of data governance; why we believe the topic of data governance is often more of a headache to management than an exciting enabler; and how we combine both our heart and brains to solve this? Read on! 

The negative perspective on data governance 

In one of our recent engagements with a client in the energy sector, we were asked to work with a central data governance team to help them create more impact within the organization. We started interviewing close to 30 people in different roles about their perspective on the data governance in their organization. The responses were not particularly positive:  

When we asked a data steward how he perceived his role, he replied: “What is a data steward?”; and after our introductions in an interview with a business unit manager, she asked us “What makes that you will fix this issue? You are the 4^th consultancy firm in 2 years that I am talking to about this topic”. The overall conclusion after the interviews: ‘There is a theoretical truth and people are tired of talking about it’. 

Unfortunately, this was not the only organization where we encountered this. The above is exactly what any pragmatic oriented manager or organization finds so frustrating and gives them a headache. Apparently, management hired a central team of people to write things such as policies, standards, and guidelines that people need to act on. However, the documents are often not understood by those people or they do not want to spend their time on it, resulting in zero impact.  

The value of data governance and its’ risks 

You may wonder what the main issue is with a central team that sets the direction and creates policies that people can adhere to. And our Anderson brains agrees that there is value in setting the direction:  

• A data engineer or product owner should not spend their time on thinking whether someone is allowed to access that type of data 
• It greatly helps if there is clarity on who decides on data definitions to prevent misinterpretations of data 
• It should be known what the escalation channel is when there are data issues to quickly resolve the most urgent issues 

Yet, what we often see is that teams or roles that set these policies are limited in their ability to also implement this in the organization and to make actual impact.  

Sometimes, this is due to a lack of capacity, as implementation takes quite some effort. However, more often implementation activities are deliberately excluded from the central role or the people working in these teams lack the skills to go out into the organization to do the implementation.  

If the latter is the case, a new risk appears: Navel-gazing. A potential toxic vicious circle of spending even more time on paper, rather than action. Often resulting in increased complex paper work, which in turn requires increased explanation and implementation efforts. 

How data governance can work 

All of this sounds somewhat hopeless, so what is needed to resolve the issue of non-effective data governance teams? 

One of the answers may be:  
Hire a consultancy company to create the basic documents and agreements. Make sure they keep it simple, stupid (KISS), and make sure that it is related to existing governance of the organization, so it does not fall of the radar completely. 

The above would actually work quite well for smaller organizations where the number of systems, processes, and teams is small enough that most people in the organization can understand an overview of these elements. When the organization is more complex, we do believe that there is legitimacy for a team that is more dedicated to data governance, but with three important notes:  

• Simplicity: Keep all documents, models, policies etc. lightweight and understandable (again KISS applies) 
• Non-invasive: Aim to search for connecting data governance to other governance mechanisms (e.g. connect data strategy to business strategy, connect data governance board to portfolio board etc.) 
• Value-adding: Ensure the team also does the implementation, and do this along use cases, rather than a big bang 

From theoretical truth to business value 

Are you interested to learn more about the topic of data governance and Anderson MacGyvers perspective, for example on how we view DAMA-DMBOK or what a use-case based implementation looks like? We offer consultancy services in this field, and also created a training specifically for data governance specialist teams that want to create more impact!