Data mart vs Data warehouse

Data mart is often interchangingly used to mean a data warehouse. If you are not proficient, you may be one of those who said data mart while it is supposed to be a data warehouse. This article will explain the concept of data mart and how it is different from a data warehouse.

Data mart definition

In simple terms, a data mart is a repository of data that is designed to serve a particular community of knowledge workers. Data is gathered from operational data and other sources that is designed to serve a particular community of knowledge workers.

In other words, data mart is a simple form of a data warehouse that is focused on a single subject (or functional area), such as Sales, Finance, or Marketing. It is important to note that definition of data warehouse does not mentions "to server a particular community". That is infact the major distinction between a data warehouse and data mart.

How data mart is different from data warehouse?

A data warehouse is a central repository for all an organization's data. The goal of a data mart, however, is to meet the particular demands of a specific group of users within the organization, such as human resource management (HRM). Generally, an organization's data marts are subsets of the organization's data warehouse.

A data mart is basically a condensed and more focused version of a data warehouse that reflects the regulations and process specifications of each business unit within an organization. Following bullet points should further clarify the difference:

Data warehouse Data mart
Holds multiple subject areas Often holds only one subject area- for example, Finance, or Sales
Holds very detailed information May hold more summarized data (although many hold full detail)
Works to integrate all data sources Concentrates on integrating information from a given subject area or set of source systems
Does not necessarily use a dimensional model but feeds dimensional models. Is built focused on a dimensional model using a star schema.

Why we need data marts?

While many of the reasons to create a data warehouse very much apply for creating a data mart, let us focus on why any organization has to create a data mart instead of a data warehouse.

  • To gain easy access to frequently needed data
  • To create collective view by a group of users
  • To improve end-user response time
  • It is easy to create data mart compared to creating a data warehouse
  • It will cost less compared to cost of implementing a full data warehouse
  • Potential users of the data mart are more clearly defined than in a full data warehouse
  • Data mart will contain only business essential data and is less cluttered

An important take away from the article is the fact that a data warehouse tends to be a strategic while a data mart tends to be tactical and aimed at meeting an immediate need.