If you are in information technology, you have no doubt heard the term “data governance” bantered about more and more. The very sound of “data governance” conjures up images of layers of bureaucracy and regulatory minutiae. If you are in the ranks of development or change control – all that can be imagined from this term is just one more layer of regulations to navigate through for a simple data related change request to be approved and released.
Maybe your industry mandates the existence of said “data governance” to satisfy some compliance requirements, or maybe the mere fact that other companies are creating their “data governance” programs has caused the powers that be at your organization to determine that yours has got to have one too. Whatever the reason for its inception at your organization, what is the point of “data governance”? What does it really mean?
At its best, data governance provides a means to identify, share, and capitalize on the company’s data assets. Ideally – barring any privacy and legal restrictions, of course – all the data that a company holds and processes should be accessible by stakeholders having valid use reasons across divisions and departments. The data is then freed to provide actionable insights into all aspects of a company’s business operations. For example – perhaps by analyzing shipping cost data it is clear that the costs may be reduced by batching shipment pickups at certain times. Or – by mining Client Support activity data, trends in client behaviors can be identified by Sales that can be used to increase client retention via sales intervention. Or – in a large conglomerate organization having many different product lines across product categories, the ability to combine the ingredient data for food products with overall sales data across all product lines, including cleaning products, it is discovered that there is a positive relationship with sales levels of a cleaning product and the fat content levels of a food product. The results of the analysis is data itself, and can be fed back into the system to fuel further discoveries. Once the data has been unleashed, the ways in which it can be put to work for a company become almost unlimited.
However, the business relevance and ongoing success of your data governance program will hinge largely on these critical factors, and how they are managed:
Partnering/Engaging with DG Stakeholders – What kinds of information and insights would help them most in making critical business directives? Have there been prior attempts to access the needed information. What were the obstacles and reasons that they may have failed in the past? How can these issues be avoided? What level of funding is available to make the initial information available and get the Program started? What kind of Metrics can be defined to determine the success of the Pilot?
Bottom Line/Business Benefit – What are the biggest blind spots in your organization and biggest potential wins that can be made from employing information that has been inaccessible in the past? Which one of these would it make most sense to use as a DG Pilot? How can the positive impact of the initiative be highlighted and quantified to justify the investment in the data governance infrastructure for future initiatives? What metrics can be established to do so? How can word be spread and a buzz created?
Identifying What You Have for Data : Shared Data and Analytics Inventory – In order for stake holders to know what data they have available to them, as an initial step, the assets need to be centrally catalogued. This will serve to provide a kind of menu to stake holders of information available for analysis. An inventory of what analytics has already been conducted is also helpful. Ideally the resulting analytical data would also be added to the Inventory as secondary/derived data so it too can be shared, and built upon.
Data Stewards / Owners : Onboarding – Addressing their needs and concerns. Identifying Reasons to share data and reasons not to share data. Consider each reason not to share data and identify how that concern can be effectively addressed. For example – Human Resources may claim reluctance to share their data due to the sensitive nature of the data. This valid fear needs to be adequately addressed. Can incentives be provided to share data? Maybe there is data in other departments that this department would like to have access to?
Making it Safe, Secure, and Compliant – Implement the ability to limit access to only individuals who need to have access, and to mask sensitive fields when required. Partner with, and involve your organization’s Compliance, Risk, and Security departments to ensure that their oversight is integrated into the data governance program. The data in the data inventory needs to be tagged with subject area and topic information so that universal entitlements to the information can be assigned. Any legal and compliance regulations that apply must be embodied in the tagging of information. A set of roles also need to be identified that can be individually entitled to the data by subject area and topic. Users can then be assigned these roles, based on their job descriptions or project assignments, which will provide them with access to whatever data that their role(s) require.
Implementation of Staffing, Infrastructure, and Procedures to Support the Above – A set of data governance roles needs to be identified along with a staffing plan to fulfill them as the program matures. Initially, existing employees with other primary roles might also be fulfilling the DG roles for the data their business unit owns, or needs to use as a stakeholder. Eventually, as data use initiatives become more prevalent, more dedicated staffing may be required to provide the direction, oversight, and guidance needed to maximize on the company’s data assets. An infrastructure to facilitate and oversee data sharing must be put in place. This infrastructure could include a meta data dictionary, role-based data entitlements management, the ability to request access or an extract of a data set, identification of current data use initiatives and contacts, requests for new data sets, and a bulletin board/blog of data issues/findings/stories/needs.
Principles of Your Program – It is important to identify and state the Principles that your Data Governance Program will embody. Will the program be concerned more about Oversight or direct Management of Initiatives, or both? What corporate values and culture, in terms of sharing of information, will the Program reflect and attempt to enforce? What methods of enforcement and oversight will be applied?
If you are able to address the above factors in the initiation of your Data Governance Program you will be well on your way to establishing the foundation for its long term success, and to satisfying the point of its inception: to, at long last, provide a means to identify, share, and capitalize on your organization’s data assets.