Part II Data Governance banana skins: Enterprise data governance – the final step must be the first

10 February 2017

by Kiran Gill Senior Associate

If you read the first part of this series, you’ll undoubtedly have been on the lookout for Data Governance banana skins and with any luck, will have avoided a few slips!

In this blog I’ll talk about the silver bullet that will (apparently) resolve all data-related issues in one go – the Enterprise-wide Data Governance initiative. This subject deserves all the airtime it can get – it’s where most programs fail before they even begin.

Enterprise-wide Data Governance cannot be achieved in one hit


I saw a TV series recently and (as I can’t get data off my mind) it got me thinking about the role people play in a successful Data Governance program. The series is centred on a theme park called Westworld, inhabited by pre-programmed, life-like “hosts”. The things they do and say are designed by the writers of the “narrative”. They react to people and situations as expected and the park runs smoothly. They follow a defined narrative… Don’t panic - there are no spoilers here but I immediately started to draw parallels between this park and a typical data dependent organisation. “How?” I hear you ask…

In a business with no clear or adopted governance, there are groups of people (much like hosts); the data producers, modifiers and consumers. They manage data in the best way they know how. Their responses are shaped by legacy processes, misinformation, workarounds and impulsive reactions. I see this as an improvised narrative. Unlike the hosts in the park, there is no real definition on the way the stakeholders respond to badly managed, missing, incorrect, duplicated or corrupt data, introducing excessive risk to the business. Governance is not successfully established and people do what they think is best.

Turning improvised narrative into a defined one requires a well-planned, robust and continuously improving Data Governance program. Many businesses see this as a one-time effort, where a policy is designed and communicated to the business and letting them understand it themselves. All staff, including data stakeholders are expected to (without guidance, context or training) understand, adopt and follow the policy and the program is expected to be a success. This is where failure sets in. People don’t have time to read a policy. They don’t have time to figure it out for themselves. Most of all, they are in no way compelled to act – what is in it for them?

When designing a Data Governance program, the “so what?” needs to be addressed. The stakeholders need to know why this program is important, and why their behaviours and workflows need to be adjusted. This can only be achieved by proving the benefits and communicating them out to the business. Although laying the foundations and establishing the program at an enterprise level is critical to get the buy in, it is not something that can remain an isolated set of events. This is where many programs fail. The enterprise effort needs to be supported by a tactical effort.

PwC’s ‘Operational Data Governance’ is a methodology that embeds governance in a sustainable and iterative way. Our way of thinking turns Data Governance on its head, making the leap from theory to an operational reality. The methodology focuses on developing strategic objectives, policies, models and roadmaps with a view to applying the theory to resolve the priority data issues, defined workflows, roles and responsibilities, data management tools, processes, and enablement, ensure successful deployment and adoption. It allows the business to take the theory and test it through well-planned pilot implementations in prioritised areas.

I worked on a recent project to design and implement Data Governance across the enterprise.  A program and operating model were carefully designed around the organisation and its structure. The objectives of the program, the strategy and the policy were communicated to the business. However, this was not the final step – it was the first. The fundamental difference here is engaging the business from the outset so they knew how to apply this new process.

Writing a set of rules and telling people to follow them, without involving them in the creation of the rules, is not only unrealistic but also unfair. The focus needs to be on informing, enabling and empowering people through carefully planned and focused deployments, allowing continuous improvement of the Data Governance model. The make or break is the way in which this is introduced into the business. A “launch” isn’t the final step – it is merely the first of many.

In the next blog, I’ll talk about planning a route to success using PwC’s Operational Data Governance framework. Until then, if you have an immediate questions or queries, please do get in touch with our Data & Analytics team.

by Kiran Gill Senior Associate