Developers built a certain level of data governance into traditional business intelligence (BI) tools, where they created some sort of semantic layer to act as a liaison between business users and the database. Business-friendly labels and definitions replaced unreadable column names, and business rules coincided with database joins and filters.
These environments brought a level of governance and helped preserve a single source of truth, albeit at the cost of slow turnaround.
Enter self-service analytics. Users can pull data and create their own analytics world. Again, the problem of multiple versions of the truth became a concerning issue for organizations.
Organizations have taken a variety of approaches to resolve these issues, from banning certain analytics tools to implementing robust data governance initiatives. Other organizations have simply let a Wild West flourish, leaving each line of business or department responsible for the governance of their respective data and analytics.
But the future of visual analytics and data storytelling will require an emphasis on data governance.
Self-service tools
Natural language processing and elastic searches are becoming a basic feature of many self-service tools. Qlik Sense allows users to simply type into a search box to quickly apply filters. PowerBI allows users to type a question to automatically generate a visualization. Tableau offers an elastic search within geospatial visualizations and has hinted their roadmap offers more robust capabilities.
Artificial intelligence will take this even further. PowerBI already offers integration into Cortana, where users can query data through either typed or vocal commands to the virtual assistant. g2o even connected a graph database to Amazon’s Alexa, allowing vocal queries to quickly search complex networks.
These features stand to change the field of data visualization (in the case of Alexa, it wasn’t visual at all). But data governance is critical for these features to work properly. If users are going to search for a specific phrase in the data, it must appear the way they expect it to. Likewise, fields and terms need to have the meaning that users expect. With artificial intelligence, users need to be confident that the virtual assistant is interpreting their queries correctly.
Luckily, the most popular self-service tools are bringing governance components. Qlik Sense, for example, allows for application-specific analytic environments where users can quickly apply governance rules. PowerBI allows for synonyms, so users can provide alternate words that reference a value. Tableau, at their recent conference, promised a scalable platform with governance capabilities.
But while these tools provide the governance capability, you need a user-centered governance effort to implement those capabilities. In the future of data visualization, governance is no longer about just managing data — it’s becoming a critical component of user experience strategy.
With the continued growth in self-service analytics, how is your organization addressing data governance issues?
