7 Data Governance Best Practices to Avoid Common Missteps
Director of Content Marketing, Sigma
A data governance strategy has a big job to do. It needs to promote compliance and ensure data quality while providing enough flexibility that the people who need to use the data can do so efficiently. It’s a tightrope that can be tricky to walk. Slip off one side, and you risk privacy breaches and bad business decisions based on faulty information. Slip off the other, and you fall behind your competitors who are taking full advantage of their data.
How can you stay balanced? In this post, we look at the data governance best practices that will help you avoid the perils that come from leaning too far in either direction.
As with any large, complex project, trying to tackle everything at once will ultimately slow your progress. If you aim for quick wins instead, like cleaning up data assets, you’ll see benefits immediately.
Starting small has two major advantages. First, it allows progress to snowball. As you implement additional strategies, benefits build on themselves. For example, if you start with consolidating duplicate data sets, you’ll be able to more easily clean up data that’s formatted incorrectly because you can do it in one place. Second, company stakeholders will quickly see value, locking in the funding you need to build out your data governance strategy. Consolidated, clean data can immediately be put into action, informing strategic decisions.
Decide roles and responsibilities
Data governance, both building an initial framework and putting policies and procedures into practice on an ongoing basis, is a team effort. Selecting team members and identifying roles and responsibilities will ensure every task is owned by someone — giving you confidence that everything will actually get done. Let’s look at the typical roles in a data governance team.
Executive Sponsors — These are the folks you need to get everyone else on board with data governance. They serve as champions for the data governance program, ensure you have the budget you need to implement it, and communicate expectations and requirements to everyone in the company. They also offer input on which specific initiatives should get priority.
Data Governance Council — This team is more involved in the day-to-day than the sponsors are, but they operate with a high-level view. They facilitate approvals, promote policies and practices to the teams under them, and make strategic decisions.
Steering Committee — On the tactical level, the steering committee gets more specific with the strategy as it relates to their individual business units. Usually, the steering committee is made up of VPs of each department. They work together to ensure the data governance strategy is implemented cohesively.
Working Groups — Members of working groups are typically managers with the responsibility of leading communication between IT and the business units. They’re focused on specific areas relevant to their departments. Working groups may recommend new projects based on their needs and help inform strategic decisions with their on-the-ground perspective.
Data Stewards — Data steward is an informal title describing someone who is a subject matter expert with an understanding of the meaning of the data and how to use it effectively. Data stewards are instrumental in creating data definitions and providing input on data policies, as well as keeping the team updated on data governance activities in their realm. They’re the ones who manage data quality and duplication issues. Another key responsibility of data stewards is to work closely with the data owners. They represent the data owners in strategic discussions as well.
Data Owners — Data owners are the ones with the closest relationship to the data. They implement data quality initiatives and serve as contacts for resolving problems with data.
Develop standard definitions
Without standard definitions, no one knows what anyone else is referring to in data discussions. Standard data definitions get everyone on the same page. This is where a solid data model can help. With a data model, you will already have a standardized naming procedure in place that is scalable, and everyone understands.
Map infrastructure, architecture, and tools
For your data governance strategy to be effective, business users should have a firm sense of the lay of the land when it comes to your data. All analysts and business users will need to know where to look and what data is usable (what has already been prepped, endorsed, deprecated, etc.). Using Data Vault or another methodology, you’ll want to create a curated, searchable catalog of datasets that are clearly badged with status. Don’t leave people to stumble through raw data that isn’t reliable.
Set goals and track progress
Your data governance framework should be built around your goals as an organization. These goals could include:
- Smarter decision-making
- Improved compliance
- Better collaboration between business units
- Providing domain experts with access to data
- Training business users on data management skills
- Reduce costs around data management
Once you’ve set your goals, identify the incremental milestones you’ll target on your way toward achieving each goal. Finally, decide how you’ll track your progress.
For example, let’s say your organization struggles with GDPR compliance. Your goal is to improve compliance. Your milestones might be to implement a data discovery initiative, develop a data mapping process, and draft a GDPR-compliant breach response. The metrics you choose to measure would tie directly to your success with each milestone.
Jason Simon, Associate Vice President for Data, Analytics, and Institutional Research and Dan Hubbard is Director of Data Management at the University of North Texas emphasize the importance of using a reporting tool. They write, “At UNT, this has helped us communicate with administrators about outstanding documentation and needed term revisions and has allowed us to track and celebrate progress as we continue to document and define a greater portion of our institutional data resources.”
Want to give domain experts better data access without compromising security? Check out Sigma.
Communication is key for implementing any new initiative, and this includes data governance. This is one reason why it’s so helpful to have a strong data governance team with strategic, tactical, and operational roles built out — it helps ensure communication across business units and throughout the organization’s hierarchical structure.
Craft a communication strategy that ensures everyone is using the same language, understands expectations, and is kept up-to-date on progress.
Remember it’s an ongoing effort
When you’re first building out your data governance framework, it’s easy to lose sight of the forest for the trees. Remember that the framework is only the first step. Governance is an ongoing effort that requires keeping your goals top of mind, sticking to your milestone deadlines as much as possible, and tracking ongoing adherence to policies and practices. It will be easier to keep people motivated for the long term if you regularly communicate progress and the benefits the organization is seeing as a result (again, the importance of communication).
The balancing act that is data governance is challenging, but these best practices will provide a firmer footing. As you implement these practices, you’ll become more comfortable in your data governance strategy, and reap the benefits as a result.
To learn more, check out our Definitive Guide to Data Governance.