Brandon Hall Group defines data governance as a set of rules and standardized processes designed to ensure data security, privacy and validity for use in analytics and decision-making. Basically, data governance is the structure of the data in your organization; the who, what, where and why of data are answered by your data governance efforts. How is data governance used right now?
Brandon Hall Group research shows that most organizations (56%) use a team of data security professionals and/or have some form of document management. Those are good and necessary parts of a true data governance effort. However, having a data council is what differentiates high-performers from other organizations when it comes to data governance.
As with any process involving people data, the biggest concerns are data security, data privacy and ensuring there are human interventions in the data-collecting and analysis chain. Data security and privacy are key to the data governance model, so be certain that the rules of access and storage are prioritized with an eye toward people data privacy and security beyond local regulations. Finding the strictest regulatory environment and using that as the base level of security and privacy is a best practice with people data.
As for human intervention, any decision that affects people (e.g., hiring, succession, compensation) should have at least one stage where an actual person reviews and approves that decision based on their knowledge of the data, the circumstances and their own judgment. Taken together, these steps will alleviate the major challenges surrounding data governance improvement.
Improper data governance means that your organization will be unable to rely on the data outputs created by your organization. Without knowing that the data going in is correct, secure and complete, there is no way to trust the types of insights that are needed to make business decisions. Brandon Hall Group shows that everything from trend analyses to skills gap analyses is affected by this, effectively making the analytical arm of your HR department weak and untrustworthy.
On the other hand, having a true data governance model in effect means that trust in the data can be verified and people using the data understand the inputs and outputs of that data. This allows for more secure decision-making and faster action as data governance lessens the need for continuous oversight and double- (and triple-) checking of outputs.
To create or improve an existing data governance model, it is imperative to understand how data is used at your organization to achieve business goals. Organizations should address that question, along with the following:
- Which stakeholders are fully invested in improving data governance and what is their role in your organization’s data governance efforts?
- What metrics are used to measure the effectiveness of your data governance model?
- Who or what group is responsible for creating and maintaining a data governance model?
- What decisions benefit most from an improved data governance system at your organization?
- What tools and technology are available to help collect, scrub and distribute data to relevant stakeholders throughout the organization?
A Data Council is the Best Way Forward for Data Governance
If your organization is serious about data governance, one of the best steps to creating a lasting data governance structure is to form a data council; a group that sets parameters for data access, common definitions, centralized storage, standardized procedure, emergency planning and organization-wide data policies.
Of all of these aspects of data governance, the first one to focus on is shared/common definitions, as that will affect all incoming data and ensure that anything done with that data will look and work the same for anyone in the organization, regardless of department or business unit.
Focus on Policies Based on the Types of Decisions that Need to be Made
The allure of simply copying another organization’s data governance model is strong, but that way is not recommended. Each organization has different requirements for data governance based on their people data needs. Some organizations want to make decisions based on current information or don’t have the resources to plan too far into the future, so collecting and storing the type of data needed for more advanced, predictive analytical models is not a good use of resources.
One practice that should be standard policy for data governance is being public and transparent about how data is collected, stored and used within the organization, and why that data is collected and used. Remember, the biggest issue surrounding people data when it comes to decision-making is trust, so the more you can build that trust, the better.
Devote Resources to Scrubbing Organizational Data
Data scrubbing is the act of improving data reliability, validity and consistency through checking and changing any data that is incomplete, incorrect, inconsistent, or duplicated throughout a system or database. Many organizations (77% according to Brandon Hall Group research), report that their data scrubbing is less than perfect.
Having clean data is the most critical step in any data governance effort. As the old saying goes, “garbage in, garbage out” and in the world of HCM, the types of outputs that could be made from un-scrubbed data will almost certainly affect real people, raising the stakes that much higher.
For any organization seeking to build a truly data-driven, evidence-based organizational culture, the foundation of that must start with data governance. It is the rule by which all data must work within your organization, without exception. Think of it as the grammar for your organization’s data language. Without data governance, there will not be trust in the data and without that trust, no other efforts will move your organization forward into the future of business.
