Tuesday, March 24, 2009

How to Implement Data Governance

By Michael Remelsti

There are many domains in data management, and one of the fastest growing is Data Governance. It is obvious that Executives are realizing that they need Data Governance in their organization since Sarbanes-Oxley provides stiff punishments should they have bad data that they sign as valid. Data Governance allows an organization to implement tight controls and build strong business rules around critical data in a strategic, prioritized, and organized way.

It is important to note that Data Governance should be a program within an organization, not a project. Projects have a defined beginning and end, and data governance is not something that you do and then you are finished and can leave it alone. DG is a recursive and iterative process, where you are continually refining your data, rules, and controls.

The first step in implementing Data Governance is to identify the people that will be working on the program. This will be your core team, often referred to as your Data Governance Office. The Data Governance director should be chosen by management as someone with influence across the organization as well as a strong knowledge of the various data sets used across the organization. The leader should also have the backing and support of management, as executive buy-in is critical to success.

Once the Data Governance Director is identified, a team must be build around this leader. The team will first do the work of identifying what should be in scope. Could all of your organizational data be in scope? Sure, but that is likely a lot of data. It must be broken down into more manageable chunks. For instance, you could choose to focus just on customer demographic data to begin with.

At the same time as the scope is being developed, you can also be working on building the data governance council. The data governance council should be made of of mid or upper management representatives from each department or business area in the organization. They will need to have the power to delegate work to analysts in their areas, but must understand the day-to-day work and be able to raise any data issues they have identified to the rest of the council.

The council will be responsible for assigning their resources with work and research that must be completed for data governance. For instance, if we want to see how customer name is used across the organization, you may assign that to Marketing. A marketing analyst would then need to review how customer name is used from when a customer is a prospect, to when a sale is made, to when warranty work is done, etc. Is it all stored in the same place? In the same format? What business rules are placed around it? This is just one example of what would be completed by the data governance council.

Once the council is in place, regularly scheduled meeting should take place so that the work defined as necessary to complete everything in scope is completed. Standard project management skills apply here; keeping track of tasks and the resources assigned to them, preventing scope creep- working towards the goal, staying on schedule, solid communication, etc. There are many things that must be done, and done well, to start an effective Data Governance Program. - 15246

About the Author: