Databricks Moving Into CDP
Charles Chretien, co-founder of Prequel, on the modern data stack’s ROI problem
This points to Databricks moving from being a place where customer data is stored and processed to being a place where marketers actually use that data to run campaigns. That is a natural expansion because a CDP is basically a packaged workflow on top of warehouse data, identity stitching, audience building, and syncing those audiences into tools like ad networks and email platforms. Databricks already has the data layer, and it has started packaging vertical products for marketers.
-
The market logic is already visible. Databricks launched Data Intelligence for Marketing in May 2025, which gives marketers a unified customer and campaign view with partner integrations. That is not a full CDP, but it is the same buying center and the same workflow entry point.
-
CDP has shifted from a standalone category into a warehouse native application layer. Internal research shows reverse ETL alone was too narrow, which pushed Hightouch toward CDP, Customer.io to bundle CDP with messaging, and mParticle to combine CDP with a larger commerce platform after merging with Rokt.
-
The broader warehouse battle is turning vertical. Snowflake closed its Observe acquisition in February 2026 to add observability as a workload specific product, which mirrors the same pattern. The winners increasingly sell opinionated solutions for a concrete team, not just generic compute and storage.
From here, the likely path is more warehouse vendors shipping team specific applications that sit directly on top of their core platform. For Databricks, marketing is one of the cleanest targets because CDP ties data work to obvious business outputs, like better audience targeting, higher conversion, and lower paid media waste.