It’s not surprising that to new comers, data governance seems very fuzzy and unwieldy. Terms like rules, policies, procedures, standards, process, data quality, security, decision rights, and accountability have been used commonly to describe various aspects of data governance. At first blush, they appear to be a collection of disconnected concepts.
When we set out to build a software application from the ground up to support data governance, we thought about this a great deal. To build great software, we needed an integrated and coherent framework; otherwise it’d just be a bunch of standalone forms. Over a two-year period, we consulted with over 40 companies around the world as well as respected practitioners and analysts, and we put in long hours of white boarding. A simple and elegant framework emerged, and it became the foundation of our new product, Kalido Data Governance Director.
Central to our framework is data policy. A data policy is a collection of measurable rules that serve a specific business objective, implemented and enforced as a single unit. There’re 4 kinds of policies: data model, data quality, data security, and data lifecycle. Data governance, then, is all about what an organization does around data policies. To add a little more meat, data governance is the business processes for defining, implementing, and enforcing data policies. Pretty simple, isn’t it? (See my blogs on data policy and data governance processes.)
Next, we thought about the context for data policies. Companies that are mature in data governance created data policy documents, and then periodically update and circulate them. The policies are in natural languages rather than technical code, which is a good thing, because you want the business to be able to author and interpret the policies. But they’re too free form, lacking structure and clarity. As a result, they’re difficult to navigate, communicate, measure and enforce. To provide context, we developed the Kalido Unified Business Model, which defines data model, business process model, systems, organizational scope, people, and the necessary relationships among them. This model became the fabric for supporting the processes of data governance.
For example, there’re two relationships between data and business processes. Business processes produce data, and business processes consume data. Knowing who produces a data element makes it possible to enforce data quality policies: those data producers would be accountable for a level of quality. And knowing who consume a data element makes it possible to communicate policies around usage and security. We also define people and their responsibility. Data stewards are responsible for data. Process owners are responsible for business processes, and IT system owners are responsible for systems. These relationships make it possible to know whom to communicate to, and to assign accountability.
The Unified Business Model is the foundation of the Kalido Data Governance Director. It provides clarity for creating and understanding data policies. On top of that, it allows us to automate the workflow for many data governance processes, which significantly improves the efficiency.
I invite you to find out more about our product; however, if you’re at an early stage in establishing a data governance program, this framework can help you structure your thinking and communicate it to the rest of your organization.
- Please Stop Trying to Come Up with a Single Enterprise Definition of Customer
- The Biggest Philosophical Debate in Data Management
- Physics of Information Management: Work Done by a Spring
- What’s the Root Cause of Bad Data?
- A Brief History of Data Governance
- Managing Master Data Using Federalist Principles