ETL (Extract, Transform, Load)
ETL is the data integration process of pulling data out of a source system, cleaning and reshaping it, and loading it into a destination warehouse or lakehouse. Sometimes flipped to ELT when the transformation happens after loading.
For dealerships, ETL covers everything from a nightly DMS extract to a real-time event stream from a modern API. The transformation step is where identity resolution and dealership-specific business rules are applied.
Also known as: ELT, data integration, data pipeline
What it looks like in a dealership
- A nightly job pulls repair orders from the DMS, standardizes VINs and op codes, and loads them into the warehouse.
- An ELT pipeline loads raw CRM leads first, then applies identity resolution inside the warehouse.
Common questions
What is the difference between ETL and ELT?+
ETL transforms data before loading it into the destination. ELT loads raw data first and transforms it inside the warehouse. Modern dealer platforms prefer ELT because the raw copy stays available for reprocessing.
How often should dealership pipelines run?+
Nightly is the minimum for reporting. For AI agents and same-day decisions, you want intraday or real-time sync on the tables that change constantly: inventory, leads, appointments, and open repair orders.