I’m on the ground at Databricks Knowledge + AI Summit this week. Yesterday, I watched CEO Ali Ghodsi take the stage and announce CustomerLake, an agentic CDP constructed natively contained in the Databricks lakehouse.
My preliminary response? Vindication.
Listening to it first-hand felt like lastly being affirmed for what MessageGears has been saying for years.
This announcement is a turning level for the martech business – however it’s not the turning level some folks suppose it’s.
The CDP as middleware is dying
The loudest takes since yesterday have been about CDPs vs. Databricks, composable distributors scrambling to reposition, and what this implies for the Brazes and Iterables of the world. That framing misses the larger story.
Tasso Argyros, who runs CustomerLake for Databricks, stated it plainly in AdWeek: “I believe the CDP, as middleware, goes to go away.” Alex Dean, CEO of Snowplow, put it this way: “the attention-grabbing frontier is now not nearly the place buyer knowledge lives. It’s about what AI brokers do with that knowledge — in actual time, within the second, on behalf of consumers and companies alike.”
They’re each proper. The query is: who’s truly positioned to execute on that future?
MessageGears is a Validated Partner of Databricks, and we’ve spent over a decade constructing the one cross-channel advertising and marketing platform that prompts and deploys messaging natively, proper from a buyer’s personal knowledge supply. We didn’t pivot to warehouse-native when it turned trendy. It’s been our structure from day one (though another distributors that got here alongside later have claimed to have invented the thought).
So when one of many largest knowledge and AI firms on the planet takes the stage and declares that advertising and marketing execution belongs within the warehouse, MessageGears is much from threatened. We’re excited.
What CustomerLake truly is (and what it isn’t)
CustomerLake is a genuinely spectacular product imaginative and prescient. Profile Brokers unify uncooked buyer knowledge into governed Customer 360 profiles natively inside Databricks. Marketing campaign Brokers construct audiences, advocate next-best actions, and optimize repeatedly round enterprise targets. Databricks calls the output “infinity campaigns” — always-on loops that analyze, resolve, and act in actual time.
Matthew Niederberger, one of many sharpest impartial analysts in martech, summed it up well: “CustomerLake is a advertising and marketing software that lives on the lakehouse as a substitute of beside it.” That’s the fitting body. Each CDP earlier than it was, in some sense, a replica of your knowledge with a advertising and marketing interface on high. CustomerLake is the advertising and marketing interface on the info itself.
However right here’s what the keynote glosses over: CustomerLake doesn’t ship messages. It doesn’t execute cross-channel campaigns. For precise supply, it fingers off to companions — Braze, Iterable, and others listed within the announcement.
Databricks constructed the intelligence layer. They intentionally left the execution layer to the ecosystem.
That’s not a knock. It’s a design resolution, and a wise one. Nevertheless it means CustomerLake is, by definition, incomplete with no cross-channel execution associate. And never all execution companions are created equal.
The hole that also exists (and why it issues)
Braze and Iterable are good platforms. They’re additionally constructed on a essentially completely different architectural assumption: that advertising and marketing knowledge must reside inside their system. To make use of them with CustomerLake, you continue to want to maneuver knowledge — or at minimal, sync profiles and audiences — out of the info lake and into their cloud.
That sync is the place intelligence goes stale. That’s the place the ML mannequin scores that reside in your central dataset change into the attribute snapshot that arrived in your ESP six hours in the past. That’s the place “warehouse-native decisioning” begins to look lots just like the structure it was supposed to interchange.
MessageGears doesn’t work that means. Our platform runs natively contained in the buyer’s knowledge supply — Databricks, Snowflake, BigQuery, and so on. — without having to maneuver, copy, or retailer a replica profile. When a journey step fires, it queries the warehouse immediately. The ML mannequin rating your knowledge science group up to date this morning is out there to the marketing campaign operating in quarter-hour. The transactional historical past, the behavioral alerts, the computed fields — all of it, at each step, from the supply.
And at present, we introduced one thing that makes the mix much more highly effective: MessageGears Journeys — a totally reimagined visible journey orchestration canvas constructed on data-native structure, designed for the AI-driven future Databricks simply described from the keynote stage.
We rebuilt orchestration as a result of the long run demanded it
Journey builders have existed for… a very long time now.
We didn’t construct a brand new one as a result of the class wanted one other drag-and-drop marketing campaign canvas. We rebuilt orchestration from the bottom up as a result of AI is essentially altering how buyer journeys are designed and managed.
Entrepreneurs want a instrument that may work with their rising military of brokers – a instrument that doesn’t make them need to rip their hair out at each step.
MessageGears architected data-native journeys in order that AI brokers (administered by advertising and marketing groups) can choose the fitting execution path for each marketing campaign movement primarily based on the precise trade-offs that matter: personalization depth, latency, and compute value. Warehouse-native journeys for high-value, data-rich flows that require the total depth of your buyer intelligence. Hybrid journeys when real-time event-triggered entry is genuinely required. Cloud-based journeys when sub-second latency is the laborious constraint.
The intelligence for these selections — the mannequin scores, the behavioral alerts, the identification decision — already lives in your Databricks atmosphere. MessageGears is the execution layer that turns it into motion, step-by-step, proper the place it lives.
That is what CustomerLake’s associate ecosystem is meant to supply. The distinction is that we do it with out the info needing to go away your lakehouse.
What this second means for enterprise leaders
Should you’re a advertising and marketing or knowledge chief, the CustomerLake announcement is a sign value taking significantly — not as a result of Databricks is your new advertising and marketing platform, however as a result of the most important knowledge firm on the planet simply advised you that the way forward for advertising and marketing execution is warehouse-native.
Gartner predicts that by 2030, 80% of net-new enterprise CDP deployments shall be embedded in or composable with knowledge platforms. Should you’re nonetheless evaluating a standalone CDP on a multi-year contract, CustomerLake simply modified that math.
However the resolution doesn’t cease on the intelligence layer. You continue to want an execution layer that may truly ship on the personalization CustomerLake is designed to unlock. If that execution layer requires copying your knowledge out of the lakehouse first, you’ve already compromised the structure you simply paid to construct.
The mix of Databricks CustomerLake and MessageGears offers enterprise groups one thing genuinely new: ruled, AI-ready buyer intelligence in a centralized dataset, and cross-channel execution that doesn’t pressure that intelligence to maneuver. No syncs. No stale profiles. No second system to reconcile.
That’s not a partnership pitch. It’s the logical conclusion of an structure that each firms have been constructing towards — from completely different instructions, for years.
Centralized knowledge has received – now the query is what you construct on it
I’ve spent 12 years at MessageGears watching the martech business slowly arrive at conclusions we constructed our firm round. Knowledge shouldn’t transfer to advertising and marketing. Advertising ought to transfer to the info.
Databricks stated that from one of many greatest levels in enterprise tech this week. The market is listening.
Should you’re at Knowledge + AI Summit, come discover us — MessageGears is a sponsor this yr and would love to speak by means of what this second means on your stack. Or in case you’re able to see what data-native journey orchestration seems like in follow, request a demo and we’ll get you on the books with an skilled.
Source link
