3 Strategies to Create a Data Governance Democracy
Much like in a democracy, good data governance will allow different functional units within an organization to disagree. Without the ability to disagree, we’d end up with either a dictatorship or anarchy. When it comes to data, dictatorships don’t work because knowledge workers will simply start their own glossaries, mappings, or dictionaries in Excel or their tool of choice outside of the central data authority’s domain and control. On the other hand, letting everyone do their own thing will be easier at first, but will result in total chaos.
Finding the balance between dictatorship and anarchy is the one thing any CDO, Data Governance Lead, or Data Steward needs to get right from the start. As a key member of the data authority within your organization, you’re thrown in the middle of IT and the business, in between several departments and business units and projects, each with their own definitions. So how do you become a democratic data authority?
First, capture the world as it is and allow people to disagree. Place each unit’s definition within its silo, but allow them to collaborate to roll up to your enterprise-wide definition and standards. Each unit will have its own context defined, but can acknowledge that at the enterprise level, we’re all talking about the same asset.
Several strategies become possible once we allow for this. These are the top three:
- Adoption: Departments can adopt a term from the enterprise whose standard attributes are managed at the enterprise level, and any changes made in the local glossaries will be overridden by the enterprise
- Promotion: Terms in local business glossaries can be prompted to enterprise glossary once the majority of business areas have agreed on the common standardized properties (attributes)
- Specialization: Terms in business area glossaries can take the properties from enterprise and specialize those properties to their needs. Ex: Customer-Borrower, Mortgage owner, Lender, Asset Manager, Patient, Practitioner, Facility, etc.
This approach worked really well for many Collibra customers. Key data leaders worked for many years to improve the quality and organization of the data. When regulations such as Solvency II came along, they realized that it provided the perfect opportunity to get all the lines of business on the same page. They took the definitions from the regulation and added them to an enterprise glossary. Each line of business could subscribe to the glossary to ensure they were using the correct definitions to comply with the regulations. As a result, they were (finally!) able to establish data as an asset for the organization.
It’s clear that establishing a data democracy takes time and perseverance. What advice do you have for organizations struggling with data dictatorships or data anarchies?