Data Cataloging: Knowing Who Uses What
A data catalog records where data lives and what it means. We explain the metamodel behind it and how a small team can start.

If three reports call the same customer field by three different names, or nobody remembers what the 'net sales' column contains, you need a data catalog. A catalog is an inventory and a dictionary of your data.
What does a catalog record?
- Assets: which table and column, in which source system.
- Meaning: the business definition, unit and calculation method of each field.
- Lineage: which source and which transformations a number comes from.
- Owner and users: who is responsible, and which report or team uses it.
- Quality and sensitivity: freshness, known issues, whether it contains personal data.
What is a metamodel?
A metamodel defines not the data itself but the shape of the information that describes the data: which kinds of object (table, column, measure, report) exist in the catalog and how are they linked? A simple metamodel can be summed up like this: every report is built from measures, every measure from columns, every column from a source table; each one has an owner.
A starting point for a small team
- Pick the five most used reports.
- Write the measures of each report and their definitions into a table.
- Add the source and the owner for each measure.
- Set a rule to update the table with every report change; a catalog that is not updated does harm.
Where Floemdata stands
Floemdata Analytics has two catalog-like structures: a signed extraction catalog (the manifest) that defines the data pulled from the source system, and a Gold catalog that defines the measures and dimensions usable in reports. The second is the single source of truth for report definitions. These are not a data catalog product or a search interface; an enterprise data catalog feature is not verified (not verified).
To get started, pick the three most disputed measures and write a definition, a source and an owner for each. Questions: info@floemdata.com.