How can columns in dbt models be documented?

The use of the description key in schema.yml is the correct method for documenting columns in dbt models. When working with dbt, schema.yml files are used to define models, sources, and their attributes, including documentation for each column. By leveraging the description key, analysts and data engineers can provide clear and concise explanations of the purpose and meaning of each column, which enhances the understandability and usability of the data. This built-in structuring promotes better collaboration among team members and ensures that the dataset's structure and content are well-documented for future use.

While comments in SQL files can provide some context, they do not systematically capture documentation in a way that can be easily compiled and viewed for all models and columns. Similarly, relying on metadata in the data warehouse assumes that the warehouse will have comprehensive documentation built-in, which is often not the case. Creating a separate documentation file could lead to discrepancies between the actual model and its documented state, making it less ideal for maintaining accurate and accessible data documentation. Thus, the description key in schema.yml is the preferred and most effective way to ensure all column documentation is properly structured and easily manageable within dbt.