Data governance can seem like a daunting task. Described generically, it has the potential to take on an impossibly large scope and a pervasive, enterprise-wide reach. But data governance doesn’t need to solve world hunger … at least not at first. Successful organizations embarking on the data governance journey have thrived by first starting small, proving the value of data governance, and capitalizing on those achievements to expand the scope. The question for those just starting out is: How do we set a manageable initial scope that can produce immediate business benefit while instituting a permanent organizational structure and processes?
There are five dimensions to data governance scope:
- Data. What subject area and entities do we govern? What attributes?
- System. What systems are under jurisdiction? These could include data warehouses, master data hubs, business applications, and the integration infrastructure.
- Business Process. What business processes benefit from, or are accountable for, our data policies?
- Organization. What segment of the enterprise should we focus on, defined by geography, line of business or functional area?
- Policy Type. Do we focus on data quality or data security? Or just a common model or data dictionary?
Of these five dimensions, the first three – data, system and business process – represent the three common starting points for narrowing scope.
A data-centric approach takes a slice based on one or more key data entities. A major oil company, for example, is starting their data governance initiative with five attributes of customer, but tackling them at a global level. Starting with data, they determine what systems store or manipulate these five attributes, and what business processes produce or consume these attributes. These systems and business processes would be “under governance.”
A system-centric approach takes on a single application or data repository such as a data warehouse and sets up a governance structure around it. The scope covers all the key data elements stored in the system and all the business processes serviced by the system. A regional healthcare provider, having recently implemented a central billing and patient scheduling system, is taking this approach and focusing on data governance for the new system.
A business-process-centric approach starts with a core business process that can be improved with better data. We determine the key data elements consumed by the business processes and the upstream business processes that produce the data. These data elements include both master data and transaction data. And we bring into governance the systems that support the core business process. For example, a large insurance carrier is tackling claims processing to improve its efficiency with their data governance initiative.
Which approach is the best? Of course, it depends. But in the absence of any other key deciding factors, a business-process-driven scope is better, because it is far easier to align the program’s objectives with business objectives and result in a tangible ROI. In addition, it would be far easier to gain and sustain engagement with the line of business.
To focus on data quality or data security? This is a common question. Without any additional information, data quality in general yields more pervasive and sustainable benefits. And, it gets more people into the mindset that data should be treated as a valuable shared asset. In other words, a good bet for scoping your first data governance initiative is to improve a core business process by providing it with better data.
In my next blog, I’ll talk about a framework for building a business case for data governance.
This blog is part 7 of a multi-part series of blogs on the topic of Enterprise Data Governance. To read other posts from this series, please see below.