For each table, a data dictionary usually records a description, the business area it belongs to, its owner, its source, how often it is refreshed and the level of detail each row represents. For each field, it records the name, the data type, a plain description, the allowed values or format, an example, whether it can be empty, whether it holds sensitive or personal data, and how it was derived.
The data dictionary is the most technical of the documentation tools. A business glossary defines business terms and metrics. A data catalog makes data assets searchable across an organization. The dictionary is where an analyst goes to understand exactly what is in a particular table and how to use it correctly.
Most data dictionaries begin life as a spreadsheet. That works for a single data set and a small team, and our free data dictionary template is designed for that stage. It becomes harder as the number of tables grows and as the data changes, because a dictionary that is not kept current is worse than none: people trust it and get the wrong answer.