What Is a Data Mart?

By on

A data mart is a subset of a data warehouse designed to service a specific business line or purpose.

Data warehousing pioneer Ralph Kimball conceived of data marts to “begin with the most important business aspects or departments.” This bottom-up dimensional approach creates a user-friendly, flexible data scheme that delivers reports rapidly, without having to drill down different screens. Data processed from a staging area pipes into the data marts. As multiple data marts are constructed, an extract, transform, load (ETL) process uploads them to a data warehouse. A data mart often benefits a specific user group.

Other Definitions of a Data Mart Include:

  • “A subject-oriented data repository that serves a specific line of business, such as finance or sales.” (Gilad David Maayan)
  •  “Contains time-variant and subject-oriented data, but with relationships implying dimensional use of data wherein facts are distinctly separate from dimensional data, thus making them more appropriate for single categories of analysis.” (Gartner IT Glossary)
  • “A subset of organized data that focuses on a specific subject area.” (George Mason University)
  • A database “designed to meet the demands of a specific group of knowledge workers and have a comparatively narrow subject area: a single department, operating area, or perhaps a specific, nagging business pain.” (TechRepublic)
  • “Built on the foundation of the dimensional model,” it “serves the individual need.” (O’Reilly)

Data Mart Use Cases Include:

  • Providing Business Intelligence for a university, including data marts for accounts payable,  admissions, finance, human resources, and student accounts
  • Storing master data for a marketing campaign
  • Sourcing “a dashboard to support analytics and data mining”
  • Breaking down a data lake’s massive volume into specific patterns

Businesses Use Data Marts to:

  • Provide “full data context and analytic tools on the same platform”
  • Run particular reports for a business function or department
  • Process data quickly
  • Build a relational database quickly
  • Save money

Image used under license from

Leave a Reply