At Entirely in Dialogue with b.telligent and at DMEXCO 2026, the conversations moved beyond individual tools. Cornelsen, dm and the marketers in the room showed what connected marketing requires in practice.

The word cloud filled quickly: data silos, transparency, budgets, reporting, AI integration, end-to-end control. At Entirely in Dialogue, which we hosted in Cologne together with b.telligent, we asked guests about their biggest marketing challenges. The answers described different parts of the same problem. Marketing work crosses teams and systems, while the context needed to coordinate it often does not.

That question stayed with us through two days at DMEXCO 2026. How can marketing teams connect what they plan, create, deliver and learn, while giving AI a useful role in the process? Three lessons from Cologne offer a place to start.

 

1. Marketing orchestration starts with the work between systems

b.telligent brought its data and MarTech perspective to the discussion. Its work with Cornelsen made the value of connected marketing tangible. As Cornelsen shifted towards demand generation, its marketing organisation and system landscape were changing at the same time. Campaign information moved through different tools and manual handovers. The ambition was to create a clearer view of planning, budgets, tasks, costs and performance across channels.

 

Cornelsen introduced Marmind in phases and connected marketing planning with SAP ERP processes. Each system had a role: Marmind supported the overall view and flexible planning; SAP continued to support functions such as master data, cost calculation and purchasing.

 

As b.telligent showed in the case, integration depends on decisions made well before a connector goes live: defining target processes, ownership and interfaces, then testing, documenting and training for the way people will actually work.

 

The lesson: Marketing orchestration is the coordination of work, information and decisions across the tools a team already uses. Its value becomes visible when people can follow a campaign from the first plan to execution and cost control, with fewer gaps in between.

 

 

2. Relevant customer experiences need a shared understanding of context

A customer signal does not explain itself. In dm’s session, an interest in beauty could point to several situations: someone new to the dm ecosystem looking for orientation, an existing customer open to inspiration, or someone who simply wants to reorder a product. Each calls for a different response.

Katharina Endsuleit, Lisa Bekurdts and Živka Tešić showed how our Entirely customer dm thinks about personalisation across campaigns and always-on engagement. The sequence begins with understanding a situation using signals and its own customer knowledge. Then comes the decision about the most useful contact, moment and channel. At times, the best choice is no message at all.

That choice matters at scale. A shared logic can guide interactions across channels and markets, while data ownership, trust and transparency remain part of the design.

The lesson: Marketing orchestration should help teams act on customer context, rather than simply increase the number of automated interactions. Relevance is a decision about what helps someone now.


3. AI needs context and guardrails to move beyond isolated tasks

Our second Mentimeter question asked participants where they currently stand with AI. Active use of LLMs and AI tools, and even the creation of individual agents, featured prominently. Work with multiple orchestrated agents appeared much less often. It was a snapshot of the people in the room, not a representative survey, but it brought the practical challenge into focus.

An agent can draft content or analyse a dataset. To help run a marketing process across systems, it also needs to understand the campaign, the relevant data, the brand rules and the limits of its authority. Teams need to know which actions can proceed, where approval is required and how a decision was reached. The fireside conversation on From AI to Impact and our exchanges at DMEXCO kept returning to this question of how AI becomes useful in everyday marketing work.

The lesson: Effective AI orchestration connects agents to the marketing value chain and gives people clear control over consequential decisions. The first useful test is a specific workflow with an agreed outcome, rather than a broad promise of autonomy.

What connected marketing looks like from here

For Entirely, these lessons point in one direction. Existing technologies and specialised capabilities need to work together across planning, content, activation and analytics. Shared context makes information usable beyond the system where it originated. Governance gives teams and AI the rules for acting on it. This is how an open marketing ecosystem can turn a collection of capable tools into a more coherent way of working.

The most useful starting point is concrete: choose one process where work currently slows down, map the systems and people involved, and define what a better result would look like. It may be a campaign approval, a content handover or a customer interaction. Once the context, decisions and measures are clear, it becomes easier to see where orchestration and AI can make a real difference.

Where does work lose momentum in your marketing value chain? Let’s find the first connection worth making. Get in touch with us.