Databricks CustomerLake Debuts as Agentic CDP
Key insights
- CustomerLake replaces batch campaigns with 'infinity campaigns': AI agent loops continuously analyzing and acting on every customer in real time.
- Early customers HP, Circle K, AB InBev, and Getnet by Santander joined the Private Preview, with pricing shifting to a consumption-based model.
- Launch partner Unity's global mobile audience of 2.85 billion monthly active users becomes addressable directly from the Databricks lakehouse.
Why this matters
Databricks is leveraging its position inside 70% of Fortune 500 data infrastructure to attack the CDP market from within, eliminating the data-movement tax that legacy CDPs require. The shift to consumption-based pricing disrupts the traditional seat-license model that has kept CDP margins high, creating direct pressure on incumbents at enterprise renewal cycles. Agentic identity resolution and native activation without copying data out of the lakehouse redefines where the system of record for customer data lives, with downstream implications for every martech vendor that profits from data portability.
Summary
Databricks launched CustomerLake at its Data + AI Summit on June 16, 2026, entering the customer data platform market with an agentic CDP built natively on the lakehouse.
The platform bets on "infinity campaigns": continuous agent loops that analyze and act on every customer in real time, replacing one-off batch campaigns. Data never leaves the lakehouse; Adobe, Meta, Braze, and The Trade Desk activate against it in place.
Essentially: (Databricks, HP, Circle K, AB InBev, Getnet by Santander) are the launch coalition.
- Currently in Private Preview with consumption-based pricing, not traditional licensing.
- Launch partner Unity brings 2.85B monthly active users globally into addressable inventory.
- Agentic identity resolution and a built-in identity marketplace cut the need to copy data to a separate CDP.
Databricks already counts 70% of Fortune 500 companies; CustomerLake converts that infrastructure footprint into a direct bid for marketing budgets.
Potential risks and opportunities
Risks
- Existing CDP incumbents could accelerate lakehouse-native integrations to neutralize the no-data-movement pitch before CustomerLake reaches general availability.
- HP, Circle K, and other preview customers bear unexpected compute cost risk if consumption-based pricing scales sharply under high-volume real-time personalization workloads.
- The open partner ecosystem depending on Adobe, Meta, and Braze introduces API compatibility risk that could break reverse-ETL pipelines silently with any partner-side update.
Opportunities
- Legacy CDP vendors face accelerated enterprise churn from the 20,000-plus organizations already using Databricks, creating pressure to develop lakehouse-native competitive responses.
- Activation partners already in the ecosystem, including The Trade Desk, Unity, Adobe, Meta, and Braze, gain early-mover advantage by deepening CustomerLake integrations before GA.
- Systems integrators and data consulting firms can capture implementation and migration revenue as enterprises shift from existing CDP contracts to the consumption-based CustomerLake model.
What we don't know yet
- General availability timeline: Private Preview launched June 16, 2026, but no GA date or public pricing tiers have been disclosed.
- Whether AB InBev has shared verifiable performance metrics from the Private Preview beyond endorsement language.
- How Databricks plans to handle agentic identity resolution across jurisdictions subject to GDPR and CCPA cross-vendor data-sharing restrictions.
Originally reported by databricks.com
Read the original article →Original headline: Databricks CustomerLake Enters Private Preview: Agentic CDP Replaces Campaign Waterfalls With Continuous AI-Driven Personalization Loops