Reverse ETL
The practice of syncing modeled data from a data warehouse into operational tools such as CRMs, ad platforms and email software.
Reverse ETL lets sales and marketing use the scores, product usage and revenue figures already calculated in the warehouse, instead of rebuilding that logic inside each tool or exchanging CSV files by hand.
An analyst writes a SQL model in BigQuery that flags accounts with more than 20 active users and no paid plan. A Hightouch sync pushes the flag and user count to a HubSpot company property every hour, triggering an assigned sales task.
Destination APIs impose rate limits and their own schemas, and most business apps have no rollback, so a faulty model can overwrite thousands of CRM fields. Syncs also inherit any quality problems in the warehouse model.
It reverses the direction of Extract, Transform, Load (ETL): a SQL query or model selects records from the Data Warehouse, field mappings translate columns into destination objects, and scheduled syncs write changes through each tool's API. Change detection compares results with the previous run so only new or changed rows are sent. Tools such as Hightouch, Census and Segment offer it, overlapping with a warehouse-native Customer Data Platform (CDP).
Intelligent Automation, Revenue Operations, Systems Architecture
Twilio Segment Documentation — Reverse ETL — https://www.twilio.com/docs/segment/connections/reverse-etl; Hightouch — What is Reverse ETL? — https://hightouch.com/blog/reverse-etl