The Agentic CDP Era Has Arrived

After years of consolidation across the marketing technology space, a lingering question emerged: was the Customer Data Platform more compelling as an idea than as an actual product? Many in the digital experience ecosystem watched as major players snapped up CDP vendors, often because customers already used both solutions together. Merging forces simply made practical sense.

That wave of acquisitions reshaped the landscape. Twilio acquired Segment, SAP bought Emarsys, Contentstack purchased Lytics, and Uniphore took over ActionIQ. These were not aging legacy tools. They were purpose-built for a specific data problem: collecting and unifying customer data, building detailed profiles, constructing audiences, and powering campaign activation. Their core value lay in identity resolution and customer profile unification.

But the conversation has shifted dramatically. Last week, Hightouch published its vision for an agentic CDP. The very next day, Databricks announced CustomerLake, its own take on the same concept. While the two visions share significant common ground, the core message is clear: the future is no longer about static customer profiles. It is about customer decisions.

Think of it as CDP 3.0. If CDP 1.0 focused on unifying profiles and CDP 2.0 introduced composability, then CDP 3.0 merges unified customer data with AI decisioning and autonomous execution. The old paradigm treated data as the primary obstacle. The new one identifies humans as the bottleneck. We are simply too slow to analyze data and act on insights. AI agents, by contrast, operate continuously and at machine speed.

As Forrester VP and Principal Analyst Joe Stanhope wrote, “Agentic AI offers the pathway to not only implement new capabilities that extend the CDP’s remit but also develop a new paradigm for generating insights, targeting audiences, decisioning, and orchestrating customer journeys.”

The key question now is who will own this agentic layer and where it will live. This is where Hightouch and Databricks diverge philosophically.

Hightouch’s Tejas Manihar and Alex Haase explained their perspective: “Five years ago, we thought we were building a better CDP architecture. In reality, we were building the foundation for intelligent agents. By moving audiences, journeys, and activation onto the warehouse, the Composable CDP connected marketing directly to a company’s richest customer and business context.”

For Hightouch, agents operate directly inside the data warehouse without copying data. This stays true to its composable roots. The agentic layer in CDP 3.0 sits within a marketing platform built on top of the data platform.

Databricks’ CustomerLake reflects a different philosophy. It argues that the data warehouse itself, specifically its data lakehouse technology, can serve as the application platform. Databricks already applied this logic to enterprise security with its Lakewatch launch in March 2026. Now it is bringing that same approach to marketing. CustomerLake can also ingest data from sources outside the data lakehouse, but the core idea remains: build your CDP directly on the data platform where governance, AI, and enterprise context already exist. Do not move or copy the data. Simply execute there.

At first glance, these two models seem headed for a collision. (Other players are also entering the space; BlueConic’s acquisition of Blueshift last week similarly adds AI agents and actions to customer data.) Yet in practice, Hightouch and Databricks can coexist because they typically target different types of organizations.

Databricks sells a data platform, so its primary buyer is often data, AI, and platform teams on the technology side. Hightouch sells more frequently to CRM, marketing, and lifecycle teams. This means each vendor enters the conversation from a different starting point. Hightouch usually finds an existing data warehouse and martech stack already in place. Databricks finds an existing lakehouse and an enterprise-wide AI strategy.

Databricks works at the enterprise level, seeking customers with strong data engineering and AI maturity. Hightouch looks for marketing operations maturity. The enterprise focus of Databricks also means its time to value is typically longer than Hightouch’s.

These differences mean the two companies often fish in separate ponds. Hightouch appeals to large DTC players, consumer financial services, retailers, subscription businesses, and travel and hospitality companies where the CMO plays a significant role. Databricks talks to global financial services firms, telecoms, large healthcare systems, and large enterprises with centralized data organizations where the CIO is heavily involved.

Neither approach is likely to dominate. Each can be the right solution for the right customer. The best outcome is that these platforms deliver on their promises, and both marketers and their customers come out ahead.