Your Customer Data Doesn't Have to Move Anymore: Optimove, Databricks and ZoomInfo All Went Warehouse-Native in Two Weeks
Your Customer Data Doesn't Have to Move Anymore: Optimove, Databricks and ZoomInfo All Went Warehouse-Native in Two Weeks
For years the deal with marketing software was simple: copy your customer data into the tool, and the tool does the work. Three announcements in the last two weeks point the other way. The tools are now coming to the data.
What shipped
Optimove launched Zero Copy Customer Lists on October 26, now in open beta. A marketer connects Optimove to Snowflake once, writes a SQL query in Optimove Data Studio that returns customer IDs, and runs it once or on a schedule. Optimove receives only the matching customer IDs; everything else stays in the warehouse[1][1], and each sync reflects the warehouse's current state rather than a copy taken when the audience was built. Optimove's CEO called the old approach "a governance gap" that "was never going to be" fixed by "another pipeline." Only Snowflake is named so far.
Databricks moved its CustomerLake agents forward on October 19. Campaign Agents went to Beta and Profile Agents to Private Preview[2][2], on top of an agentic CDP that runs inside the lakehouse under Unity Catalog governance. Databricks has not announced a general availability date or pricing beyond a consumption model. Customer-reported numbers are eye-catching (HP says audience-ready data in two hours versus two to three months on a traditional CDP), but they are vendor-published and not independently verified. Gartner predicts that by 2030, 80% of net-new enterprise CDP deployments will be embedded in or composable with data platforms[2].
ZoomInfo made its MCP server and API available inside Snowflake on October 30. Bulk enrichment runs at up to 10,000 records per request and 200,000 per query[3][3], with OAuth 2.0 and each user's role-based permissions. Pricing is undisclosed and no customers were named.
Why this matters for a lean team
Most of our readers do not run Databricks. But the direction of travel affects you anyway, for three reasons.
- Fewer copies means fewer places for data to go stale or leak. Every export to a marketing tool is another copy to secure, delete on request and reconcile.
- Agents need governed data. An AI agent that builds audiences or enriches records is only as safe as the permissions on the data it reads. Role-based access at the warehouse is easier to audit than five tool-specific settings.
- Someone has to own the SQL. Zero-copy audiences are defined as queries. That is powerful, and it moves a decision from a marketing UI to a data definition. Decide who signs off on it.
All three launches are beta, enterprise-leaning or vendor-reported. Treat them as a direction, not a purchase order.
A practical checklist before you follow
If your customer data lives in a CRM and a spreadsheet, warehouse-native tooling is a later step, not a first one.
Pick a single list you rebuild by hand every month, such as active trials with no activity in 14 days, and define it as a query.
Write down who approves changes to the query, and where the definition is documented.
Which warehouses are supported today, what leaves the warehouse (IDs only, or rows), and how are permissions enforced for agents?
Beta features change. Keep your current export path alive until the new one has run cleanly for a full cycle.
The bottom line
The shift is less about any one product and more about where the source of truth lives. If you have a warehouse, start asking your martech vendors how they plan to work with it instead of copying from it. If you do not, a clean, well-owned CRM is still the right foundation.
FAQ
What does warehouse-native or zero-copy mean in marketing?
The marketing tool queries customer data where it already lives, in your data warehouse, instead of importing a duplicate. In Optimove's case, only the IDs of matching customers come back.
Do I need Snowflake or Databricks to benefit?
Not today. The announcements target teams that already run a warehouse. Smaller teams should first get their CRM data clean and owned.
Are the Databricks results proven?
The HP and Wine.com figures are customer-reported and published by the vendor. Treat them as indicative, not independently verified.
Is this a replacement for a CDP?
Gartner expects most new enterprise CDP deployments to be embedded in or composable with data platforms by 2030. For a small team the practical step is to ask vendors how they integrate with your warehouse.
Sources: Optimove, CMSWire; Databricks CustomerLake, CMSWire; Marketing technology news, October 31, Agile Brand Guide.
Related
Google's 13-Day Spam Update and New AI-Content Guidance: A Quality Gate for B2B Teams Publishing With AI
Google's September spam update ran 13 days and 16 hours, and its AI-content guidance was refreshed in early October. A pre-publish quality gate for B2B teams that write with AI.
51% of Marketing Teams Run AI Agents. Only 6.5% Benchmark Them Against a Metric.
51.1% of marketing organisations use AI agents, but only 6.5% benchmark AI investments against defined metrics. A lean-team scorecard for proving what agents actually do.
AI Overviews Now Appear on 83% of Branded Searches. Your Brand SERP Needs a Defense Plan.
Ahrefs found AI Overviews on 82.91% of US desktop branded searches, and Semrush puts commercial-keyword coverage at 88%. Here is how a lean B2B team should defend its brand results page.