SegmentStream layers cross-channel attribution and budget optimization on top of Snowplow — reading raw enriched event data loaded by Snowplow's warehouse loaders into your data warehouse.
Snowplow collects, enriches, and loads behavioral event data into your own data warehouse. Once Snowplow is loading data, point SegmentStream at the same warehouse and SegmentStream reads the events from there.
Snowplow supports BigQuery, Snowflake, Redshift, Databricks, and the open lake formats Delta Lake and Apache Iceberg. SegmentStream connects to whichever one you use.
Snowplow's job is to collect and land high-quality, enriched behavioral data in your warehouse. SegmentStream then builds identity graph, cross-channel attribution, and budget optimization on top of those events. Snowplow's schema is read natively — no separate event mapping or transformation is required.
Snowplow loads data forward from when the pipeline starts running, and any enrichment changes can be re-applied. Historical replay of raw collector data depends on how your pipeline retains it — see Snowplow's loading-process documentation for retention and replay options.