Glossary
What is ETL / ELT?
ETL (extract, transform, load) and ELT (extract, load, transform) are two ways to move data from source systems into a data warehouse. With ETL, you process the data before loading it; with ELT, you load first and then transform in the warehouse itself.
ETL / ELT explained
With ETL, data is processed on the way, in a separate tool, before it reaches the warehouse. That was the norm for a long time, because storage and compute in the warehouse were expensive. With ELT you load the raw data first and then write the transformations as SQL in the warehouse itself, often with a tool such as dbt.
ELT has the advantage that the raw source data is kept. If a calculation is wrong, you adjust the transformation and recalculate, without pulling everything from the source again. The loading itself, the "EL" part, is often done with standard connectors, for example with Airbyte or Fivetran.
Which variant you choose matters less than whether the process runs reliably. Watch for alerts when a load fails, for changes in the source and for the moment users expect the new figures.
Example
Every night, Acme Groothandel loads orders and invoices from its ERP into the data warehouse unchanged. An SQL transformation then calculates net revenue and links it to the customer and product dimensions, so the dashboards are ready in the morning.
ETL / ELT in ENABLE
Data is loaded outside the ENABLE portal by Data Analytics, for example with Airbyte. ENABLE then uses the result in the PostgreSQL data warehouse; the portal has no ready-made connectors to systems such as Exact Online or AFAS.
FAQ
Frequently asked questions
What is the difference between ETL and ELT?
With ETL you transform data before loading it into the warehouse; with ELT you load first and transform in the warehouse itself. ELT has become common with modern cloud warehouses.
Which tools are used for ETL or ELT?
For loading, tools such as Airbyte, Fivetran and Azure Data Factory are common. For transformations in the warehouse, dbt or plain SQL is often used.
How often should you load data?
It depends on how current the figures need to be. For most management reporting once a night is enough; operational dashboards may need more frequent loads.
Further reading
- SolutionData integration: connect data sources for insightConnect data sources for reporting and AI, via one central data warehouse.Read more
- ResourcesWhat is a data warehouse? Explained and comparedA data warehouse is a central database in which you bring together data from different source systems, such as your ERP, CRM and webshop, clean it up and organise it so that you can report on it and analyse it effectively. A source system is built to process transactions; a data warehouse is built for reading, comparing and keeping history. A data lake stores raw data in any format, and a data lakehouse combines the storage of a lake with the structure of a warehouse.Read more
- SolutionBuild a data platform: layers, choices and approachFrom sources to dashboards and AI: how a data platform is structured.Read more
- IntegrationsIntegrations: accounting, CRM, marketing and moreENABLE integrations: accounting, ERP, CRM, marketing, online shops, HR and databases. Power BI and PostgreSQL directly, everything else via your own data warehouse.Read more
Use AI on your own data. Securely, and without the headache.
In an online demo we show how ENABLE makes your data securely available and how you build on it: describing an app in the AI Builder, plain-language questions, dashboards and APIs. And how we set it up and manage it for you.