flex-height
text-black

Server room in a data centre

What is a data warehouse?

A data warehouse (DW) is a digital storage system that connects and harmonises large amounts of data from many different sources.

default

{}

default

{}

primary

default

{}

secondary

Overview of data warehouse

A data warehouse (DW) is a centralised repository that collects, integrates, and stores large volumes of current and historical data from multiple sources. It supports business intelligence (BI), reporting, and advanced analytics by providing a single, consistent source of truth. By consolidating and standardising data, organisations can generate reliable insights, meet regulatory requirements, and make informed, data-driven decisions.

Data typically flows into a data warehouse from operational systems (such as ERP and CRM), internal databases, and external sources such as partner platforms, IoT devices, weather feeds, and social media. As cloud computing has matured, data storage has shifted from traditional on-premises environments to flexible multi-cloud and hybrid cloud architectures.

Modern data warehouses are built to manage both structured and unstructured data such as videos, images, and sensor streams. Many incorporate integrated analytics and in-memory processing to enable faster queries, real-time data access, and more efficient reporting and BI workflows. Without a data warehouse, organisations struggle to combine heterogeneous data sources, prepare data properly for analytics, and maintain visibility across datasets.

Benefits of data warehousing

A well-designed data warehouse is the backbone of successful business intelligence, reporting, and analytics. By consolidating data into a single source of truth, it accelerates insights for better, more confident decision-making across the business. Key benefits include:

What types of data can a data warehouse store?

When data warehouses first appeared in the late 1980s, they were built to store structured data—well-organised information such as customer details, product lists, and transaction records. As business needs expanded, companies also wanted to work with unstructured data such as documents, images, videos, emails, social media posts, and sensor output from machines and IoT devices.

Modern data warehouses can handle both structured and unstructured data, bringing them together to give businesses a more complete, integrated view for stronger insights.

Key concepts and comparisons

There’s a lot to learn in the world of data warehousing. Here are some of the most important concepts. Explore further definitions and FAQs in our glossary.

Data warehouse vs. database

Databases and data warehouses both store data, but they serve different roles. A database manages real-time information for a specific business area, while a data warehouse combines current and historical data from across the organisation to support reporting and analytics. Although it runs on database technology, a data warehouse adds tools for integrating, modelling, and managing data over time.

Databases keep daily operations running by processing transactions and updating records quickly. Data warehouses support analytics, helping teams identify trends, compare performance, and make strategic decisions.

Data warehouse vs. data lake

Data warehouses and data lakes both store large amounts of data but have different purposes. A data warehouse holds structured, prepared data for reporting and analytics, whilst a data lake stores raw, unprocessed data that may be used later. They often work together: raw data resides in the lake and is transformed and moved into the warehouse when required for analysis.

Use a data lake for flexible, low-cost storage of raw data. Use a data warehouse for fast, reliable analytics on structured data. Most organisations benefit from both; the lake captures everything, and the warehouse turns it into insight.

Data warehouse vs. data mart

A data mart is a subsection of a data warehouse, partitioned specifically for a department or line of business such as sales, marketing, or finance. For example, a sales data mart might focus on leads, pipeline activity, and closed-won deals, while a finance data mart would centre on budgets, forecasts, and revenue metrics.

Some data marts are created for standalone operational purposes as well. While a data warehouse serves as the central data store for an entire company, a data mart provides relevant data to a select group of users. This simplifies data access, speeds up analysis, and gives them control over their own data. Multiple data marts are often deployed within a data warehouse.

Key components of a data warehouse

A modern data warehouse includes four key components: a central database, data integration and ingestion tools, metadata, and access tools. Together, they provide fast, reliable analytics at scale.

  1. Central database: The core storage engine for the warehouse, traditionally a relational database but increasingly an in-memory or cloud-native system for higher performance.
  2. Data integration and ingestion: Data is brought in from source systems using batch methods such as ETL and ELT, along with real-time options such as change-data-capture replication and streaming pipelines. These processes also handle transformation, quality checks, and enrichment.
  3. Metadata: Information that describes the data—its origin, structure, meaning, and how it should be used—spanning both business and technical context.
  4. Access tools: Tools that enable users to query, analyse, and interact with warehouse data, including reporting tools, dashboards, analytics platforms, and application development tools.

Data warehouse architecture

Historically, data warehouses were organised into layers that aligned with how data moved through the system. A typical data warehouse includes three layers. Modern platforms simplify the architecture to support faster data movement and analytics.

Data warehouses were traditionally built and managed by IT teams, but modern platforms increasingly empower business users to work directly with data. Key capabilities driving this change include:

How does a data warehouse work?

A data warehouse organises information from across your business so it can be easily explored, trusted, and analysed. The process usually follows four simple steps:

  1. Extract: Data is retrieved from source systems such as applications, databases, and cloud services. At this stage, the data is collected as it is.
  2. Transform: The data is cleaned, standardised, and shaped so it’s consistent and ready for use. This may involve removing errors, aligning formats, or applying business rules.
  3. Load: The prepared data is stored in the warehouse in a structured format optimised for fast reporting and analytics.
  4. Analyse: Once the data is loaded, teams can explore it using dashboards, reports, and advanced analytics to make informed decisions.

ETL vs. ELT: What’s the difference?

ETL (Extract → Transform → Load): Data is transformed before it enters the warehouse. This approach is common with traditional data warehouses that have limited processing power.

ELT (Extract → Load → Transform): Raw data is loaded into the warehouse first and transformed within the warehouse. Modern cloud platforms favour this method because they can efficiently handle large-scale transformations.

What are the four key characteristics of a data warehouse?

A data warehouse is built on a few core principles that ensure it delivers reliable, consistent, and analysable information across the business. The four main characteristics are:

  1. Subject-oriented: Organised around core business topics–such as customers or sales–to support analysis.
  2. Integrated: Data from different systems such as ERP and CRM is cleansed and standardised, so it fits together consistently.
  3. Time-variant: Stores historical data over long periods, enabling trend and performance analysis.
  4. Non-volatile: Data is stable once loaded—readable but not updated or deleted—ensuring a reliable source of truth.

Benefits of cloud data warehouses

Cloud data warehouses are increasingly popular because they offer significant advantages over traditional on-premises systems. Here are the top seven benefits of moving your data warehouse to the cloud:

  1. Quick to deploy: Set up storage, computing, and new environments such as data marts or sandboxes within minutes, from anywhere.
  2. Lower TCO: Pay only for the resources you use. Avoid hardware, facilities, and maintenance costs, and reduce expenditure by separating storage and compute.
  3. Elasticity: Scale up or down instantly to handle changing workloads and large data volumes without manual effort.
  4. Security and disaster recovery: Cloud platforms often provide stronger security controls, encryption, and automatic backups to protect against data loss.
  5. Real-time performance: In-memory and cloud-native engines provide rapid processing speeds for real-time insights.
  6. Access to new technologies: Easily integrate capabilities such as machine learning, automated insights, and advanced analytics.
  7. Empowers business users: Provides teams with a unified view of data as well as intuitive tools to analyse information and connect new sources without extensive IT involvement.

Data warehouse best practices

When building a new data warehouse or expanding an existing one, following proven practices helps you achieve your goals while saving time and money. Some practices focus on business needs, while others fall under broader IT guidance. The list below is a solid starting point, and you’ll refine it as you work with your technology and services partners.

Business best practices

IT best practices

Summary

Modern data warehouses—especially cloud-based ones—play a central role in digital transformation by unifying data from internal and external sources for a complete, timely view of the business. They power dashboards, KPIs, alerts, and reports across the organisation and support fast, complex analytics without impacting operational systems.

Because they can start small and scale easily, they help both corporate teams and business units make better decisions and improve performance.

FAQs

What is a data lake?
A data lake is a place to store all kinds of Big Data, whether it’s structured data from business applications or unstructured data from mobile apps, social media, or Internet of Things (IoT) devices. Because data is stored in its natural format–structured, unstructured, semi-structured, or binary–conversion, normalisation, or other processing may be needed to enable analytics across multiple data types. Most data lakes are cloud-based due to the large volumes of data they store, the need for high-speed connections to distributed sources, and the need for scalability. Their ability to store vast amounts of raw data makes them a flexible, low-cost complement to a data warehouse.
What are ETL and ELT?
ETL stands for “extract, transform, and load.” It refers to the process of extracting data from a source system, cleaning and shaping it into a usable format, and then loading it into a data warehouse or another data store. Many modern systems also use ELT—“extract, load, and transform”—where the data is loaded first and transformed afterwards. Both approaches help turn raw data into something that can be analysed, whether it comes from transactional systems or more complex, unstructured sources.
What is a data mart?
A data mart is a focused slice of a data warehouse designed for a specific business area or team, such as finance or marketing. It gives that group quick access to the data most relevant to its work and allows it to manage its own curated dataset within the larger warehouse. For example, a finance data mart might include budgets, forecasts, and revenue data tailored to the finance team’s reporting needs.
What is data modelling?
Data modelling is the process of defining how data is organised and connected so it can be stored and used effectively. A data model outlines what the data represents and how different pieces relate to one another, creating a blueprint for consistent structure across systems. For example, a sales data model might show how customers, orders, and products are linked together to support reporting and analysis.
What is an enterprise data warehouse (EDW)?
An enterprise data warehouse (EDW) is a centralised system that stores all of a company’s current and historical data in one place. It provides a single, consistent source of information for analytics, reporting, and companywide KPIs. Many EDWs operate in the cloud for easier access, scalability, and management.
What are the three types of data warehouses?
  1. Enterprise data warehouse: An EDW is a central, company-wide data warehouse that stores all current and historical data in one place. It provides a single, consistent source of truth for analytics, reporting, and KPIs across the entire organisation. Most modern EDWs are cloud-based for scalability and easier access.
  2. Operational data store: An ODS is a near–real-time data store used for operational reporting and day-to-day activities. It sits between transactional systems and the EDW, combining data from multiple sources in a more up-to-date, but not fully historical, form. It’s useful when data needs to be refreshed frequently for quick operational decisions.
  3. Data mart: A data mart is a smaller, subject-specific slice of a data warehouse, designed for a particular team or business unit, such as finance, sales, or marketing. It provides fast access to the data that matters most to that group without exposing the entire warehouse.
What are the four components of a data warehouse?
  1. Central database: The primary storage layer where structured, cleansed, and integrated data is housed. This is typically a relational, columnar, or cloud-native database optimised for analytics.
  2. Data integration / ingestion tools: Tools and processes—such as ETL (extract, transform, load), ELT (extract, load, transform), batch loads, and real-time replication—that bring data from source systems into the warehouse and prepare it for use.
  3. Metadata: Information that describes the data: where it came from, how it’s structured, what it means, and how it should be used. Metadata helps users understand and trust the data.
  4. Access tools: The applications and interfaces that enable users to query, visualise, explore, and analyse the data, such as reporting tools, dashboards, analytics platforms, and SQL query tools.
Is SQL a data warehouse?
No. SQL is a language used to query and manage data, whereas a data warehouse is a system that stores, organises, and processes large amounts of data for analysis. SQL is simply one of the main tools used to work with data inside a data warehouse.