Metadata describes other data. A photo's metadata includes the camera model, exposure settings, date taken, and GPS coordinates. A document's metadata includes the author, creation date, and file size. A dataset's metadata includes the schema, row count, source system, and refresh schedule. The data itself is the content. Metadata is the context. Without it, the data is harder to find, trust, and use.

Metadata comes in three flavors. Descriptive metadata helps you find things: titles, keywords, authors. Structural metadata explains how data is organized: schemas, formats, relationships. Administrative metadata covers management: creation date, access permissions, retention rules. Each type serves a different purpose. A data catalog is a metadata repository. It indexes datasets and makes them searchable. An analyst looking for customer data can search the catalog instead of asking around. The catalog shows what exists, who owns it, how fresh it is, and what it contains. Without a catalog, data discovery is word of mouth. Metadata also supports governance. You cannot enforce retention policies on data you cannot identify. You cannot protect sensitive data you do not know exists. Metadata makes the invisible visible. It is not glamorous work. It is the foundation of everything else.

Metadata types

• Descriptive — titles, keywords, authors
• Structural — schemas, formats, relationships
• Administrative — creation date, permissions, retention
• Technical — file size, encoding, compression
• Business — definitions, ownership, quality rules

Metadata is data about data. It is how you find, understand, and manage the data you have.