What Is an Enterprise Data Warehouse (EDW)?

Enterprise Data Warehouse vs. Data Warehouse

An enterprise data warehouse is used to store and report all of a business’s data, regardless of where the data originates from and what team or department will use the information.

In comparison, the data in a data warehouse may be specific to a single department or a line of business.

Why Do Businesses Need EDWs?

Businesses rely on useful, accurate data to make informed decisions about products, employees, customers, and more. Without quality data, company leaders must rely on their gut feelings to make these crucial choices.

EDWs empower business leaders to evolve past gut feelings and integrate data from multiple, unstructured sources into business intelligence and data visualization tools, such as Tableau, PowerBI, and Qlik. The tools then provide teams with quick, data-driven answers to pressing questions.

Types of EDWs

Enterprise data warehouses fall broadly under two categories — on-premises or “traditional” data warehouses and cloud data warehouses. Some organizations use a third type: virtual data warehouses.

On-Premises or Traditional Data Warehouse

On-premises data warehouses are primarily used within the company’s firewall. These include Teradata, Netezza, and Exadata. The on-premises data warehouses provide full control; however, the control comes with more responsibility. A traditional data warehouse needs a full tech stack and has to be maintained by database administrators, system administrators and network engineers.

Cloud Data Warehouse

Organizations have started investing heavily on cloud data warehouses. Cloud data warehouses are designed to provide scalability, elasticity and cost efficiency. Cloud data warehouses include Amazon Redshift, Google BigQuery, and Snowflake. With cloud data warehouses, organizations can purchase compute power and storage as needed. Plus, cloud data warehouses don’t need additional tech resources and staff to manage the data.

Virtual Data Warehouse

Some organizations go with a third option called data virtualization. In this scenario, the data stays in the source systems and a virtual layer is created for data analytics and reporting.  This can appear to be an easier and faster technique for getting started. However, data virtualization causes major performance issues at scale and has to rely on source systems for querying the data.

Benefits of EDWs

• Enterprise data warehouses are powerful and useful for answering specific business questions.
• They serve as a central repository for organizational data.
• They maintain a history of transactions and aggregated information that is required for BI and analytics tools.
• They serve multiple users across the enterprise.

Challenges of EDWs

• Data must be copied, standardized, and moved from source systems into the EDW using ETL processes. Data is first extracted(E) from multiple source systems (ERP, CRM, HR, external files). Then the data is transformed(T), aggregated and loaded(L). For example, if customer data that includes first name, last name, address and gender comes from different sources, the gender, state, city, country has to be standardized before loaded into the data warehouse. These ETL processes take resources to build and maintain them.
• When business users have additional questions that are outside the scope of the current data warehouse, new data has to be brought into the warehouse. This can take weeks to months.
• BI and analytics users extract data outside the data warehouse and build cubes and imports to optimize performance. This creates multiple copies of the data. The result is often stale data with no governance.

Use Cases for EDWs

The data in enterprise data warehouses help companies answer specific business questions and make data-driven decisions. Enterprise data warehouses answer questions such as:

• What is the total company revenue for the last two fiscal years? How does it break down by regions?
• What is the average deal size in Canada vs. the United States?
• What is the YoY spend across sales, marketing, engineering, finance and HR departments?

An enterprise data warehouse helps businesses answer questions that involve cross-organization data and teams. They can help elevate data-driven decision-making across the entire enterprise.

Dremio & EDWs

Dremio’s forever-free lakehouse platform enables interactive BI and high-performing analytics directly on your cloud data lake. It opens up a broader set of data to a larger set of data consumers for diverse analytical needs. Learn more about how Dremio complements your data warehouse.