Data
ETL and ELT
ETL (extract, transform, load) and ELT (extract, load, transform) are the two common patterns for copying data from business systems into a database or data warehouse for reporting.
How it works in practice
Both start by extracting data from sources such as the CRM, accounting system and ecommerce platform. In ETL the data is cleaned and reshaped before it is loaded. In ELT the raw data is loaded first and transformed inside the warehouse, often with a tool such as dbt. Cheap warehouse storage and compute have made ELT the default for most new builds, because the raw data stays available if a report needs changing.
Tools such as Fivetran and Airbyte handle the extract and load steps for common systems. The transform step is where business definitions live, such as what counts as revenue or an active customer, and it needs owning like any other piece of code.
A worked example
A 70-person retailer loads Shopify, Xero and Google Ads data into BigQuery each night. SQL models join them into one table of margin by product and channel, which feeds the Monday trading dashboard.
Examples in this glossary describe typical builds for companies of 1 to 1,000 people. They are not named client results.















