Why Your AJO Data Foundation Determines Everything, Including How Well AI Works

I’ve been in enough Adobe Journey Optimizer (AJO) implementations to spot a pattern. A brand arrives excited about AI-powered personalization, real-time decisioning, and cross-channel orchestration. Six months later they’re wondering why the platform isn’t performing the way they envisioned. The issue is usually what sits behind it, and whether it was designed to support the journeys, decisions, and AI capabilities the team wants to use.

Building a solid foundation is genuinely hard and requires thoughtful decision-making. Organizational pressure to show ROI from platform investments quickly leads teams to prioritize speed, rather than taking time to make the critical architecture and planning decisions needed for long-term platform durability and scalability. And it’s something we see across almost every implementation, regardless of the team’s experience or intent.

What makes this especially important right now is AI. The more AI capabilities your team wants to leverage, the more your foundation needs to be able to support them. A strong data model makes your journey better today and unlocks the full potential of AI tomorrow.

You can’t automate your way out of a foundation problem. Going slow now is the only thing that lets you go fast later. 

The Real Issues with AJO Data Foundations

The most common issue is a data foundation for AJO that lacks a real understanding of what the data attributes are and when to use them. Teams aren’t always clear on the specifics around AJO-specific concepts including what an event should represent, what attributes should live on a profile, or how identities should be stitched together across channels. Journeys get built on top of that uncertainty, and the gaps compound quietly until they become very loud problems.

The second issue is treating foundational work as a time-intensive effort that doesn’t deliver immediate value. Data governance, schema design, and merge policies can feel like they slow things down. But in AJO, your data model is the most critical part of the product. Every journey you build is only as good as the profile it reads from and getting that wrong early creates compounding technical debt that’s expensive to unwind.

The third is the pressure to show results before the platform is ready to deliver them. The drive to demonstrate ROI early is real and completely understandable. But shortcutting key architecture and design decisions in pursuit of quick wins often delays the very outcomes the business is looking for. The teams that invest properly upfront are the ones that reach sustainable performance and real AI-driven results faster.

How to Build the Right AJO Foundation

The right implementation starts before you touch the journey canvas. Begin with the end in mind. What do you actually want to do downstream? What experiences do you want to deliver, and what decisions does the platform need to make?

Once you know that, you can work backwards. What data needs to be pulled in, from where, and how? What attributes need to live on a profile to support those decisions?

From there, define your identity strategy and how customers will be recognized across channels. Getting those sequencing decisions right sets everything else up for success.

“The end in mind” doesn’t just mean your next campaign but also means your vision for the next two to three years. The brands that set up their foundation well are thinking about what kind of experiences they want to be able to deliver as the platform and their team matures.

That long-term vision is what separates a data model built to last from one that gets rebuilt every 18 months. It’s also what makes AI meaningful down the line, because a foundation designed for where you want to go is one that AI can actually work with.

Your data foundation isn’t a project deliverable. It’s a long-term asset and it should be designed like one.

From there, governance matters as much as the technical setup. Who owns which journeys? What rules govern how a customer can be entered, exited, or re-entered? How do channels share frequency budgets, so you’re not overwhelming people? These are business decisions that need to be made before anyone opens a canvas, rather than platform-specific questions. Build the framework, document it, and make sure your team knows why the rules exist, not just what they are.  That understanding is what makes governance stick after go-live.

The teams that get this right end up with something powerful, a unified, trustworthy view of their customer that becomes the foundation for everything that follows, including AI.

Where AI is Already Adding Value

AI is already working inside AJO in meaningful ways, and understanding what exists, where to access it, and what it can do is something a lot of teams are still getting their heads around. Adobe’s documentation can be hard to follow, so here’s our plain-language breakdown of the AI capabilities available in AJO today and where they fit within the journey orchestration process.

AI Assistant is the main interface for AI inside AJO. Use it for product knowledge, troubleshooting, and operational questions without needing to raise support tickets or dig through documentation. It also powers content generation from natural language inputs across email, push, SMS, and web, including subject lines, body copy, image variations, and full message variants (available on AJO Prime and Ultimate tiers). Brand alignment scoring and content quality checks are built in too. It’s the interface through which Journey Agent capabilities are accessed (see below for more details), so it’s worth getting familiar with them early.

Built-in AI platform features are separate from the AI Assistant. AJO includes platform-level AI features like send-time optimization, which learns the best time to reach each individual customer based on their engagement history. These features run in the background once enabled and improve over time as your data accumulates. They’re some of the most immediately impactful AI capabilities available today, and they work best when the event data feeding them is clean and consistent. You will find these features embedded throughout the AJO product menus. Journey Agent accessed via AI Assistant (recently GA), powered by Adobe Experience Platform’s Agent Orchestrator has two skills, Analyze and Create.

Analyze, (GA), is a natural language performance analysis of existing journeys. You can ask questions about how a journey is performing and get clear answers without needing to build reports manually.

Create, currently in limited availability, lets teams build or optimize journeys from natural language prompts, with channel-specific content generation included. This capability is growing fast but is currently available to a limited set of organizations. Its quality is directly tied to how well-defined your underlying data is, which is exactly why foundation work matters.

AJO now provides an MCP (Model Context Protocol) server, currently in Public Beta, that surfaces campaign, channel configuration, and sandbox insights directly inside any MCP-compatible application such as Claude. The intended use is conversational inspection and troubleshooting. This means different personas can collaborate around and evaluate the same orchestration data conversationally, without switching between tools or writing API queries. It’s an area that’s evolving quickly and worth building awareness of now.

AI is already  accelerating knowledge, access, and opportunities. AI Assistant helps teams navigate the platform, troubleshoot issues, and keep documentation up to date without it becoming a manual burden. That means the foundation gets built right and stays that way.

For teams who have built the right foundation, the results are tangible. Send-time optimization moves beyond batch windows to a model that improves with every send because the event data feeding it is clean and consistent. Predictive suppression lets AI identify customers at risk of fatigue before they disengage, because the profile is rich enough to read the signals. And, generative content variants can be tested at scale, because when segmentation is trustworthy, experiments are too.

Journey performance insights are available in natural language through Journey Agent’s Analyze skill giving teams answers in seconds that would otherwise take an analyst hours to surface. And because teams are no longer firefighting data issues, AI-surfaced insights and automated capabilities can be used the way they were designed.

These outcomes are already happening for teams that invested in the foundation first, whether they’re mid-implementation or well into run and operate. The difference is groundwork.

Building for the Next Wave of AJO AI

AJO’s AI capabilities are advancing fast, and the roadmap is moving quickly. The teams best positioned to benefit are the ones who understand how to get the most out of the capabilities that exist today, while building the foundation that unlocks what’s next. Investing in the foundation now is what makes speed possible later. That is the competitive advantage.