Although the terms “data warehouse” and “data mart” sound similar, they are quite different. It is important to first understand how they differ in order to define some characteristics and practical applications for each.

Serra (2012) has a great explanation of data warehouses as being “a single organizational repository of enterprise-wide data across many or all subject areas”; while Inmon classifies the data mart as “a logical view or a physical extract of a larger data warehouse, usually isolated for the need to have a special data model or schema” (n.d.).

An even more succinctly as “a simple section of the data warehouse that delivers a single functional data set” (Gibbs, n.d.).

Data Warehouses Data Marts
Multiple subject areas Specific subject area (Finance, Sales, etc)
Holds very detailed information Usually stores summarised data (albeit many hold full points of interest)
Can take years to implement Can take weeks to months to implement
Often found to store over 100 GB/TB of data Often found to store below 100 GB of data
Multiple data sources Few data sources
Does not necessarily use a dimensional model but feeds dimensional models Is built focused on a dimensional model using a star schema
Works to integrate all data sources Focuses on coordinating data from a given branch of knowledge or set of sources
Heavy usage of ETL (extract, transform, load) More common use of ELT (extract, load, transform)

Applications of Data Warehouses

  1. Social Media:
    As social networks continue to grow, so does the data that is created by its users as well as the numerous connections between individuals on the network (Standen, 2008). Being able to map multiple sets of data to create a more fluid interconnected network understanding of users is vital. 
  2. Banking:
    Processing of high volume transactions 24 hours every single day along with many other metrics such as risk, fraud and credit checking (, n.d.); the banking industry is in need of ever more advanced data to backup history and usage requirements of its customers. 
  3. Trend Analysis:
    Many businesses have adopted the trend analysis pattern algorithms from historical data to determine how products and services will be used in the future as well as what trends perform better than others (Poolet, 2008).

Applications of Data Marts

  1. CRM (Customer Relationship Management) Systems:
    Customers are the heart of any business and storing accurate data to manage relationships thereof is vital to the success of any business. 
  2. Accounting Systems:
    Keeping track of income and expenses is not only good for reporting, but also indispensable to a continued business effort in any future endeavouring company. 
  3. Marketing Systems:
    Collecting information about products, services, how they perform and which customers are targeted is very useful to drive growth to a business.

Although data warehouses and data marts are two different things in the world of data and databases, they work hand in hand (, n.d.) by data marts coming together to form a data warehouse of a wide range of topics and subject matters.


Serra, J. (2012) Data Warehouse vs Data Mart [Online], Available from:

Inmon, B. (n.d.) Building the Data Warehouse: Third Edition (pg.142) [Online], Available from:

Gibbs, M. (n.d.) What is a Data Mart? – Design, Types & Example [Online], Available from:

Standen, J. (2008) Data Warehouse vs Data Mart [Online], Available from: (n.d.) Difference between Data Warehouse and Data Mart [Online], Available from:

Poolet, M, A. (2008) Data Warehouse Workloads and Use Cases [Online], Available from: (n.d.) Data Warehousing [Online], Available from: