Moving Beyond Rules-Based Workflows to Coordinated System of AI Agents
Agentic marketing is less dependent on third-party data than many of the approaches that came before it. Its strength comes from making better use of what organisations already have. That is first-party customer data combined with real-time enterprise signals like orders, inventory, and service interactions, says Jessica Keehn, CMO, SAP Customer Experience.
Topics
Artificial intelligence has already transformed how marketers analyse data, personalise campaigns, and automate repetitive tasks.
Yet for many organisations, AI remains largely confined to generating insights rather than driving meaningful business outcomes. Teams continue to struggle with disconnected systems, fragmented customer data, and manual processes that slow execution long after valuable insights have been identified.
As enterprises explore the next phase of AI adoption, attention is shifting toward agentic AI—systems capable of planning, reasoning, and acting autonomously against business objectives while remaining aligned with human-defined goals and governance frameworks.
According to Jessica Keehn, CMO, SAP Customer Experience, the real opportunity lies in connecting AI to the broader business context rather than limiting it to customer data alone. In her role, Jessica is responsible for strengthening the company’s CX go-to-market strategy, championing market requirements, and driving product marketing excellence.
In this interview, Keehn discusses what truly agentic marketing looks like in practice, why fragmented data remains one of the biggest barriers to AI scale, how CMOs should approach accountability in autonomous systems, and why first-party data is becoming increasingly critical in an AI-driven future.
Excerpts from the interview;
In your experience, what’s the most common internal bottleneck that prevents marketing teams from moving from experimentation to scaled AI deployment?
The biggest challenge we see is not ambition; it is fragmentation. Most organisations are already experimenting with AI, and many are generating useful insights. The issue is that their data, systems, and teams are still disconnected.
According to our Global Engagement Index Report 2026, 55% of enterprises say their data is too unstructured to use effectively, while 54% say they are unable to access and use data in real time.
We call this The Engagement Divide, where AI can see what should happen, but the organisation cannot act on it in a consistent or scalable way.
That shows up in a few ways. Data is often unstructured or not accessible in real time. Systems are not properly integrated. Governance is not always clear. So even when a pilot works, it stalls because execution still relies on manual handoffs or siloed processes.
From a customer perspective, that leads to inconsistent experiences across marketing, sales, and service. At Sapphire, we highlighted that only a small proportion of organisations have a truly consolidated approach to customer experience, which shows how far there is still to go.
Closing that gap is not about adding more AI. It is about connecting the entire CX landscape and grounding it in a shared business context so that insights can turn into action.
When businesses get it right, the impact is real. Jack Wolfskin’s challenge wasn’t a lack of channels—it was understanding how to activate each customer in the right way. SAP Engagement Cloud enabled the business to combine data insight and AI to personalise journeys in real time across all touchpoints.
That means it can respect channel preferences, trigger communication based on behaviour, and use AI-enhanced recommendations to suggest next-best actions and offers.
For customers, this means communication that feels relevant and personal; for the business, it means higher engagement, stronger loyalty, and more efficient use of every channel.
As AI systems become autonomous, how should CMOs rethink control and accountability in decision-making? Where should the human remain firmly in the loop?
At Sapphire, we introduced a simple principle that sits at the centre of the Autonomous Enterprise. People set the direction, and AI executes. That idea applies directly to marketing. Autonomy is not about removing control. If anything, it makes accountability more important.
CMOs and their teams still define the strategy, the outcomes, and the guardrails. That includes brand voice, compliance, data usage, and how decisions should be made and escalated. What changes is that execution can happen at a much greater speed and scale.
Because these systems are grounded in real business processes and governance, decisions can be both fast and reliable. But transparency is key. Teams need to understand what the system is doing, what data it is using, and what outcomes it is driving.
The human role does not go away. It becomes more strategic. It is about setting intent, defining value, and making sure everything is aligned to the business and the customer.
What does a truly agentic system look like in action, and how does it elevate the automation and orchestration capabilities CMOs are already using today?
At SAP, we think about agentic marketing as a shift from isolated automation to something much more connected and outcome-driven across the entire customer experience.
In practical terms, it means moving beyond rules-based workflows to a coordinated system of AI agents that can plan, reason, and act against a defined business goal. Instead of manually building campaigns step by step, marketers set the objective, whether that is increasing repeat purchases or improving conversion, and the system takes care of orchestration end to end.
What really changes the game is that these agents are grounded in a full business context. They are not just looking at customer data in isolation, but also signals like orders, inventory, service interactions, and financial data.
That means marketing moves from educated guesswork to execution based on what is actually happening in the business in real time. That depth of context is where SAP has a real advantage. We have been embedded in core business processes for over 50 years, so AI is working from a much fuller picture of how a business runs, not just a snapshot of customer behaviour.
For example, SAP Joule agents can interpret customer behaviour and business context, and with partners like Google Cloud, generate and adapt content dynamically. The system can activate campaigns across channels and continuously optimise them. But more importantly, it adjusts based on real conditions. If inventory changes or delivery timelines shift, campaigns adapt automatically.
At Sapphire, SAP’s annual conference, we talked about this as moving from a lightbulb to a laser. Many organisations already have AI insights, but they are often scattered and hard to act on. Agentic systems align that energy into something focused and coordinated that can actually drive outcomes across the full journey.
Ultimately, CMOs remain in control of strategy and guardrails, but with true agentic AI, execution happens faster and more consistently than any human-managed workflow could support.
With restrictions on third-party data, can agentic marketing truly deliver on its promise, or is it overly dependent on data access that’s becoming harder to secure?
Agentic marketing is less dependent on third-party data than many of the approaches that came before it. Its strength comes from making better use of what organisations already have. That is first-party customer data combined with real-time enterprise signals like orders, inventory, and service interactions.
This is why we talk about grounding every interaction in full business truth. When AI understands both the customer and the operational context, it can deliver experiences that are not only personalised but also realistic and fulfillable in the moment.
As third-party data becomes less reliable and more restricted, brands that can unify and activate their own trusted data will increasingly have an advantage.
AI has just accelerated that. There’s some good news from our Global Engagement Index: Over three-quarters (77%) of enterprises we talked to said they had a clearly defined first-party data strategy. Strong data governance and consent management aren’t constraints here; they’re essential enablers of scalable and trusted AI.
As AI takes over more execution and optimisation, how do you think the role of CMO will evolve over the next five years?
The role of CMO remains aligned with designing an ideal customer journey. As AI connects marketing more closely with commerce, service and sales, while continuing to shift these touchpoints into more AI-first and digital formats, the role of CMO focuses on growing customer relationships through every touchpoint in the journey.
The focus shifts toward defining customer value, building long-term loyalty, and making sure the entire experience works seamlessly from end to end.
I spoke about this recently at OMR, and one theme that really resonated was that marketing is no longer just about messaging. It is about orchestrating the entire customer experience in real time. Agentic AI enables that, but it also raises the bar for how marketing operates and how it connects to the rest of the business.
It is a very exciting moment. CMOs have more opportunity than ever to drive growth, shape customer experience and Customer Engagement by bringing the organisation together around the customer in a much more meaningful way.
At SAP, we believe that it’s critical for marketers to have the right AI, apps, and data foundation to build the customer journeys rooted in operational reality with full customer understanding – so every interaction is personalised, optimised and deepens the customer relationship.
ALSO READ: What Enterprise Marketers are Getting Wrong about GEO
