Self-service business intelligence empowers analysts and their end-users with dashboards and insights. Anyone in the organisation can find the answers they need using common datasets without waiting for custom solutions. A solid data governance strategy is necessary to ensure data is secure, and that the right controls are in place with so much freely available data.
To ensure data isn’t misused, altered, or shared too widely, there are several layers of governance you can implement. From governing user roles to data source level security, you should carefully consider who has access to what and why.
What is Data Governance?
Data Governance refers to the management, protection, and security of a company’s data assets. However, it’s not all about red tape and rules. Having data that is easy to understand, trustworthy, easily discoverable, and reliable ensures great data quality.
This strategy and the supporting tools should focus on creating opportunities to use data, rather than creating fear in making mistakes. Having the right guidelines in place helps organisations and teams:
- Reduce uncontrolled sharing of sensitive information
- Enhance discoverability and increase adoption of Power BI
- Comply with government regulations
- Automate manual checks of reports by publishing processes and controls
When a process is controlled, it is monitored, observed, and any violations are detected before they have consequences. Using Power BI’s data control capabilities, everyone in an organisation can discover insights and share them with others securely.
Strategies and tools:
- Sensitivity labels – with Power BI sensitivity labels, users can classify critical content without compromising productivity or collaboration. As both Power BI Desktop and Power BI Service support these features, sensitive data is protected from development to the point when it is viewed by the end user.
- Row level security – users of Power BI workspaces have access to all data stored within that workspace. Access can be restricted for specific users by using row-level security. For example, you can ensure users have access to only the information relevant to their department.
- Private embed codes – reports can be embedded in any internal website that accepts a URL or an iFrame. The website can be hosted on-premises or in the cloud, and data security is protected via row-level security.
- Auditing and logging – tracking who is accessing Power BI datasets and reports can help organisations meet regulatory compliance requirements and maintain records. Power BI administrators can analyse usage by running custom reports based on Power BI activity logs.
Data quality and data governance are two distinct disciplines, but they work in parallel. Successful governance initiatives can create opportunities to improve data accuracy, completeness, and consistency.
Strategies and tools:
- Certification – as part of the Power BI data governance framework, one of the most important decisions is whether to use certified datasets. Dataset certification signifies the quality of the data, the accuracy of the measures, and the agreement on the business logic.
- Limit workspace creation – workspace creation should be governed by a policy that stipulates when, why, and how workspaces are created. The best approach is to base your decisions on policy and to be transparent about how permissions are granted.
- Data lineage view – keeping track of how data flows from the source to the destination can be challenging. Reports that are built from multiple data sources and have dependencies are even more challenging. With Power BI’s lineage view, you can identify reports that haven’t been refreshed and initiate a refresh from a friendly interface.
- Data stewards with clearly defined roles – in order to make informed decisions about data, we must communicate its meaning and its use within the organisation. Data Stewards are responsible for managing their datasets and answering questions when they arise.
It is much easier to collaborate when teams have access to all the governed datasets and reports they require. Using Power BI workspaces, teams can work together on dashboards, reports, datasets, and workbooks.
It should be considered whether data can be shared externally. Depending on the organisation, guidelines may address how data is shared, the level of specificity in reporting, and when access to data should be revoked.
Strategies and tools:
- Collaboration through workspaces – workspaces should be set up for each team, project, or subject matter and collaborators can be invited in when required. There is no need to create a workspace for every report or dataset.
- Distribution through apps – Power BI content is best shared with end-users through apps. Apps can include dataflows, datasets, reports, paginated reports, and dashboards. End-users see what has been distributed rather than everything in a workspace.
- External sharing strategy – sharing is the simplest way to give outside users access to your reports and dashboards. Having a strategy in place is crucial to ensuring that everyone understands what can and cannot be shared, as well as where they can get advice.
- Power BI champions – a Power BI champion is a content creator recognised by their peers as the go-to Power BI expert. They share knowledge and help their colleagues by providing solutions, troubleshooting, and staying up to date with the data governance strategy.
Whether Power BI is used for self-service, enterprise BI, or both, governance is an important part of the implementation. Many organisations try to implement Power BI without thinking about governance and then face the challenge of trying to change users’ habits in order to force them to use processes they feel will impede their progress. It doesn’t matter if your organisation is just beginning its Power BI journey, or if it has already entered the “Wild West”, governance is an important and necessary component of all Power BI implementations.