A data catalog holds information about data, often called metadata: what tables and files exist, what they contain, who owns them, where they come from, how fresh they are, and how they are used. Its main job is discovery. An analyst who needs data on, say, weekly prescriptions by territory should be able to search the catalog and find the right table, with enough context to use it correctly.
Most catalogs also connect to related tools: a business glossary for definitions, a data dictionary for field-level detail, and data lineage to show where data comes from and what depends on it.
Catalogs often fail for a simple reason: they are populated once and not maintained. Descriptions go stale, owners move on, and new tables are never added. Within a year, people stop trusting the catalog and go back to asking colleagues.
The catalogs that stay useful are kept current as part of normal work rather than as a separate documentation project, and are organized around the questions the business asks rather than around how the database happens to be laid out.