Tableau Data Management

Organizations once captured and managed only relevant corporate data. Now, every department—from marketing and sales to finance—has diverse types of data and niche analytical needs to maximize its value. People are capturing data even before knowing what to do with it, and some departmental data and its applications may not even be on IT’s radar.

Contributing to this proliferation of data is also all the work done to get the data into the right shape and format for analysis. Often in traditional models, IT prepares data for a specific analytics use case, then continues to maintain the databases or files. While more data accumulates, IT seeks stability to support the data environment, but faces challenges in deprecating tables and data sources, unclear of the impact to users downstream and wary of breaking business-critical dashboards.

Modern analytics have helped shift IT’s role from building reports to maintaining and securing the systems that enable self-sufficiency for data-driven decisions across the organization. Business users are now bringing in diverse types of data that matter to them, but not yet in a scalable manner. Many processes are still manual and may involve moving data from governed to ungoverned systems for analysis. IT is embracing governance as guardrails to data and analytics exploration, rather than restricting access and causing a bottleneck. But IT still carries the responsibility of scaling self-service analytics to a broader audience, which demands operationalizing and automating processes, alongside a balance of control and agility.

As organizations increasingly rely on data-driven decision-making for organizational change, more people are asking to access and analyze data. While some business users are developing more sophisticated data skills, most don’t know what data they should be using or where to find it. From naming conventions to complex data structures, understanding databases, and which tables to join are challenging for everyday users. When data access was limited to fewer individuals in the company, it was easier to just ask a “go-to” data expert. Today, that doesn’t scale alongside growing use cases for data and the fast adoption of modern analytics.

Figuring out how data assets are being used across the organization has historically been either anecdotal—“who is using this and how?”—or a coding challenge, scraping content trying to find answers. Let’s say an ETL job fails, or there’s a table to be removed from a database. In a traditional, data management model, an administrator might update a wiki or send a mass email in hopes that the people to be impacted will get the message. Even if an organization has an enterprise data catalog with helpful data descriptions, documented lineage, and indicators of the data’s freshness, how often are end users logging in to this other system to see whether the data they are using is trusted and up to date before they begin their analysis?

Compliance requirements are also making it more challenging to ensure people are using the right data, which can hurt the perception of an organization to properly manage sensitive data. If IT wants end users to understand what data they should be using and whether there is a quality issue, people need information about its quality in the context of their analysis—not in a separate system or tool. Bringing this metadata directly to the users helps them know to trust their data.

Data curation & discovery

Data stewards and data engineers can add descriptions and metadata to databases, tables, and columns, as well as certify data assets to help users find trusted and recommended data. Developers can also use metadata methods in the Tableau Server REST API to programmatically update certain metadata. Data engineers and database administrators will benefit from usage metadata to inform changes in tables to optimize data sources.

Content authors can search existing data sources, databases and files, or tables in Tableau Server and Online to see if the data they need already exists, helping to minimize duplicate data sources. Content consumers can more easily find trusted data while browsing by seeing descriptions, usage, and certification for data assets. While looking at a viz, users can easily access field descriptions in the Data Details tab on the dashboard, giving them confidence that they are using the right data for analysis.

Lineage & impact analysis

IT can perform impact analysis to understand which data assets and users will be affected by changes—like a database administrator or data engineer changing a column or table, or a data steward modifying a calculation. IT can email affected owners of data assets from within the Tableau platform.

Business users can quickly see the lineage of data they are using for analysis, including the origin and owner of the data, whether or not it’s certified, and prep flow run history—even before opening a viz. More advanced users may find analytical content published on the server and can dive in to explore calculations in workbooks.

Quality warnings & alerts

Data engineers and data stewards can set quality warnings directly in the server environment or through an API. These indicators—Warning, Deprecated, Stale Data, or Under Maintenance—help alert users to the status of data assets and provide additional information, like if a refresh is delayed or there is missing data that might compromise analysis. Prep flow owners, like data stewards or analysts, directly receive alerts for failed prep runs—like in the event of a database timeout or missing column—so they can quickly take action.

Business users will see quality indicators on data assets in the context of their analysis in Tableau. This can help people of all skill levels feel confident in making business decisions with the data, or in the event of a warning, they can read information from IT to decide whether or not to proceed with analysis.

Metadata API integration

IT can utilize the Metadata API to extend the power of your existing metadata management solution. Pull metadata into Tableau that database administrators and engineers curated in metadata tables or an enterprise catalog, and extract metadata from Tableau to use in other data applications and business flows.

Business users can get visibility into the work IT has done with valuable metadata from the enterprise catalog where they’re most likely to see and use the information—in the context of their analysis in Tableau.