A guide to preparing clean, consistent, and user-friendly Tableau Published Data Sources.
To ensure high-quality, trustworthy data in Tableau, it’s essential to follow both metadata best practices and the core data quality principles. The Metadata Checklist provides guidance for curating a Published Data Source by establishing consistent data standards. Applying this checklist prior to creating an extract or Published Data Source enables governed, self-service data access that is user-friendly, easy to understand, and aligned with business needs.
Effective data curation begins with a focus on the five key characteristics of quality data:
- Validity – Data should conform to defined business rules or constraints.
- Accuracy – Data should be as close to the true values as possible.
- Completeness – All required data elements should be present.
- Consistency – Data should be consistent within a dataset and across multiple datasets.
- Uniformity – Data should use standardized units of measure.
By ensuring these qualities are addressed—ideally upstream in the data pipeline—you reduce the need for corrections during analysis and improve trust in the insights generated. Use the following checklist to review and apply these principles when preparing data for Tableau.
Quality Data Model Checklist
- Validate the data model – Ensure the structure of your data supports analysis. At this point, determine if the data source should be compatible with Tableau Prep.
- Filter and size to the analysis at hand – Remove unnecessary rows and limit scope to what is relevant. If most users are looking for data within the past 3 years, limit the data source – instead of pulling all data from the beginning.
- Use standard, user-friendly naming conventions – Rename fields at the data source level so they are understandable by business users.
- Hide unused fields – Simplify the data source by hiding fields not needed.
- Enter field descriptions as comments – Add context by documenting each field’s purpose and usage. Don’t forget to add a description to the data source itself.
- Add new calculations – Ideally, calculations will be in the database. But if that is not possible, include common business logical calculations in the data source.
- Remove duplicate or test calculations.
- Create hierarchies (drill paths) – Build logical hierarchies to support intuitive exploration. For example, create a hierarchy of all cost centers. Another idea is to create a hierarchy based on geographical areas.
- Set data types – Confirm that fields are assigned the correct data types (e.g., dates are dates, zip codes are geography, text is string).
- Apply formatting to dates and numbers – Format fields for consistency and readability.
- Set fiscal year start date.