Traditionally, lifecycle marketers built their marketing stack around whichever tool held the most complete view of the customer profile. For 20+ years, that was the customer engagement platform. But after reading a post from Peter Oleson (he spent 5+ years at Iterable) and seeing all the recent developments with AI, I'm beginning to realize that’s changing.
The best metaphor I can think of for what's happening right now is Henry Ford's assembly line.
Before 1913, cars were built one at a time, with workers moving around a stationary chassis. Build time dropped from 12.5 hours to 93 minutes, and the price of a Model T fell from $850 to $260. Subsequently, nearly every car manufacturer adopted the “assembly line” model of work. And they didn’t do it because the old way of building cars stopped working, they did it because the new way of building cars simply exposed a better workflow.
Lifecycle marketing is facing a similar inflection point, and we're right at the beginning of a curve that's going to transform the industry exponentially. That's obviously a big claim, so let's start at the beginning: how the model came to be, why it worked for as long as it did, and what's actually changing.
How we got here
To understand why the model is shifting, you have to understand why it existed in the first place. To do that, we need to look at the history of lifecycle marketing.

The email era. In the early 2000s, as the internet and e-commerce took off, email became the first major direct-to-consumer digital channel. Platforms like Salesforce Marketing Cloud (SFMC) and Adobe rose to prominence, giving marketers a place to store customer data and run campaigns. But these platforms were limited to file uploads, relational databases, and scheduled refreshes.
The mobile era. In 2007 the mobile era began with the invention of the smartphone (specifically the iPhone). New channels like mobile push, in-app, and SMS made customer engagement more immediate and event-driven. Customer engagement platforms (CEPs) like Iterable and Braze emerged after 2010 to support API-first, real-time data flows and to coordinate experiences across channels.
The CDP era. Around 2013, customer data platforms (CDPs) began to gain traction. This technology brought marketers one step closer to the complete picture of the customer. It became a central layer that could aggregate events and coordinate customer experiences through downstream tools.
In the subsequent years, the category has expanded, but the lifecycle world has converged around the same three core components and interface.

The customer profile: the data layer and record of the customer
The orchestration layer: the interface where you build and manage audiences, customer journeys, messages, templates, etc.
The delivery layer: the pipes to send omnichannel messages
What changed
For 20+ years, the model built around the customer profile made sense: CEPs would store a version of the customer, help marketers build and orchestrate messages, and then deliver the message. But there are four macro changes starting to break the old model down.
1) Singular customer profiles are no longer enough: While a profile is a convenient model for sending, it’s too simple for marketing in 2026. A user profile almost always assumes one person, one identity, and one relationship to your brand. Real customer interactions are never that clean. For example, a household can share one login, or a loyalty member can also be a churned subscriber. On top of that, there’s data that’s critical for great campaigns (inventory, in-store activity, data science models, etc.), which don’t naturally conform to a user profile.
2) Cross-channel marketing for paid and owned channels is complicated: CEPs and ad platforms have no way of talking to each other, and they both need data in slightly different ways. A winback audience built for email needs an email address and a purchase history. The same audience for paid retargeting needs a completely different identifier, and it needs to refresh constantly. Otherwise, you end up paying to retarget someone who already converted, or serving an offer to someone your lifecycle team just suppressed.
3) The center of gravity shifted: The data warehouse became the single source of truth for all the customer data, not just the user profile. Orchestration moved upstream outside of CEPs because journeys, audience segmentation, and channel personalization require access to the full context of the business and the freshest data possible.
This isn’t a contrarian view either. Nearly every CEP is racing to build around the data warehouse. Braze now has Cloud Data Ingestion, and Iterable has a feature called Smart Ingest (fun fact — this is powered by Hightouch) to bring warehouse data into the CEP. But even with these advances, the profile still has to live inside the platform's own database to send a message. This means the version of the profile living in the CEP will always be a limited version of the one in the warehouse.
4) AI is rewriting the rules of marketing: The number one question on every lifecycle team's mind right now: How do I rewire my lifecycle program around AI? And the single most impactful factor that addresses this problem is context. Context means three things: your data, your brand, and your memory (e.g., campaign performance). Without access to all three of those at the same time, you can’t fully attain the benefits of AI for marketing. You’ll simply speed up singular points of work rather than change the shape of the work entirely. The underlying AI models are constantly changing and the the model matters far less than what’s around the model.
What’s next?
If the fullest and most accurate definition of the customer and all of your other data already lives upstream in the data warehouse, that might mean marketing teams no longer have to build around the limitations of a user profile. And that also might just mean that, for the first time, the customer profile and the delivery of the message no longer have to been bundled together in the CEP.
To be clear, a warehouse full of data and a world full of AI agents doesn’t mean your lifecycle program can magically run by itself. Something has to sit between that data and every decision, message, and send. At Hightouch, we call that something a “marketing harness.”And next week you’re going to learn all about it.


