Definition

ETL (Extract, Transform, Load)

Also called: extract transform load, ELT.

ETL stands for Extract, Transform, Load — the three steps of moving data from a source, cleaning and reshaping it, and writing it to a destination.

ETL stands for Extract, Transform, Load. It describes the three steps of moving data from where it lives to where it is useful: extract it from the source, transform it (clean, type, de-duplicate, reshape), and load it into a destination like a database, spreadsheet or report.

The three steps

  • Extract — pull the raw data out of files, documents or systems.
  • Transform — fix types, remove duplicates, join sources, derive new columns.
  • Load — write the cleaned result somewhere it can be used.

ETL is the engine behind most reporting. When the three steps run on a schedule, you get a data pipeline. The free tools on this site are a lightweight ETL: they extract (PDF/CSV), transform (clean and structure), and load into a dashboard or report.

ETL vs ELT

ELT swaps the order — load the raw data first, then transform it in the destination. The choice depends on where you want the heavy processing to happen; the conceptual steps are the same.

Run a mini ETL — extract a CSV, transform it, load a dashboard: Financial Dashboard from CSV →

FAQ

Frequently asked questions

What is the difference between ETL and a data pipeline?
ETL describes the three steps (extract, transform, load). A data pipeline is the running system that performs those steps automatically, usually on a schedule and with monitoring.
Do I need ETL for a simple report?
You are already doing it informally whenever you export data, clean it and paste it into a report. Naming the steps just makes it easier to automate them so they run on their own.