SaaS Must Provide Native Warehouse Exports
Conor McCarter, co-founder of Prequel, on Fivetran's existential risk
Native warehouse exports are becoming a core product feature, not a nice to have integration. Once a SaaS buyer wants to join product data with billing, support, and marketing data in Snowflake, BigQuery, Redshift, or Databricks, the app either ships a direct pipe or risks becoming a silo. That shifts value away from generic ETL vendors and toward software vendors that can package warehouse access as part of the product and monetize it directly.
-
Fivetran built a large business by maintaining roughly 150 to 200 connectors to popular SaaS apps, but that model is expensive because every connector sits on top of third party APIs that change constantly. Native exports remove that middle layer because the SaaS vendor already owns the source data, schema, and roadmap.
-
The buyer need is not basic dashboarding inside the app. Teams want raw data in their warehouse so they can join Stripe transactions with Salesforce opportunities, Zendesk tickets, ad spend, and internal tables, then run company specific SQL and models. That is why in app analytics and warehouse exports coexist rather than replace each other.
-
This is also becoming easier to sell inside SaaS. Prequel describes warehouse connectors as a profitable add on that can lift activation and retention, and prior work shows vendors like Stripe and Salesforce have already moved in this direction. More recently, AI has increased the pressure because SaaS vendors now need customer data flows both out to warehouses and back into apps to power AI features.
Over time, warehouse connectivity is likely to settle into the same enterprise readiness bucket as SSO and audit logs. The winners will be SaaS products that make customer data portable by default, while infrastructure vendors that once sold standalone syncing will keep moving up the stack into broader platforms, observability, and adjacent workflow products to defend growth.