Advanced Usage of Snowplow dbt models

📄️ User Mapping

On this page, ` can be one of: web, mobile`

📄️ Duplicates

The e-commerce, web, and mobile packages perform de-duplication on both eventid's and page/screenviewid's, in the base and page/screen views modules respectively. The normalize package only de-dupes on eventid. The de-duplication method for Redshift & Postgres is different to BigQuery, Snowflake, & Databricks due to their shredded table design. See below for a detailed explanation and how to process potential duplicates in any custom models.

📄️ Incremental Logic

On this page, ` can be one of: web, mobile, ecommerce, normalize`

📄️ Snowplow Optimized Materialization

This package makes use of the standard dbt incremental materialization with an optimization applied for incremental models. Its key advantage is that it limits table scans on the target table when updating/inserting based on the new data. This improves performance and reduces cost. We do this by overriding the macro that generates the sql for the merge and insert_delete incremental methods.

📄️ Advanced Operation

Asynchronous Runs