Most IT leaders and even many users understand the importance of data governance. Data governance refers to management of the security, lineage, and quality of the data. It is the linchpin to credibility and trust when it comes to analyzing and sharing insights based on data. If people don’t trust the data, every analysis and every report will be called into question.
Leading companies realize that establishing a proper data governance program requires putting into practice the people, processes, and technologies to manage how data gets used. It’s certainly not an easy task. Fortunately, more and more governance processes are being integrated into the technologies of the BI platform.
When considering a BI solution, it’s incumbent upon IT and business leaders to understand the particular data governance shortfalls so that they know which processes they will still need to do manually. Although Tableau has been attractive to users for its visualizations and quick analysis, it lacks key data governance features to provide one version of the truth.
Tableau implementations typically begin with a few power users that download Tableau Desktop. It allows them to analyze data and create visualizations without the need for IT. This requires users to create their own data sets and models, propagating silos of desktop-stored data that are hard to control and secure. What’s more, data sets and models are locked forever in Tableau’s proprietary format. As more and more users do this, the equivalent of another Excel hell emerges.
Security, one of IT’s top concerns, is also a problem for Tableau. Users are connecting to a variety of data sources, often bypassing security protocols. Organizations face additional security risks when sensitive data is downloaded locally on the desktop or embedded in shared workbooks. Plus, users can override security measures when sharing their analyses.
The real problem is, there is no central way to manage the Tableau Desktop users. That means there is no role-based security and no way to easily manage deployment and upgrades. In fact, many organizations don’t let end users install their own applications, thus putting an additional burden on IT.
Tableau, wanting to appeal to the enterprise, created their Tableau Server product. While Tableau may say that this product addresses governance, it continues to perpetuate standalone analyses due to its difficulties in sharing business logic and content. Tableau has limited reuse of the business logic, calculated metrics, or content. Additionally, there is no versioning of the business logic or content which causes data quality issues and can leave users scrambling through reconciliation exercises.
Tableau recognizes they are lacking in their data governance capabilities and has begun working on adding new features to address this. However, critical deficiencies remain, leaving business users to make decisions on inaccurate data and IT to reconcile different versions of the truth and try to clean up the data mess.
When we built BI Office, we did so with the enterprise in mind—from the start. This includes data governance; security and management is centralized even when users aren’t. Unlike Tableau, BI Office manages all business logic making it easy to move, import, or update across the enterprise. It provides content migrators, data source changers, and missing data wizards to address data model change unlike Tableau in which users can’t edit their data sets or parameters. In BI Office, all transactions are logged; its metadata and data lineage can be managed and analyzed; and business logic, reports, dashboards, or any other user-created object can be edited and versioned.
All this serves as a reminder to do your due diligence when searching for a business intelligence tool that will fit the enterprise. Look under the hood of that pretty graphic and make sure what’s underneath is just as robust.
- Follow @PyramidAnalytics on Twitter
- Pyramid Analytics Home Page