Sarah Probert on what New York agencies can tell us about the future of marketing. Here are some reflections from her recent trip to the Big Apple — and why the same themes kept showing up again and again.
After a series of meetings with agencies in New York, one thing quickly became clear to me: while every group had its own language, platform story and point of differentiation, many of them were circling the same big ideas. Some were polished, some were provocative, and some were a little more abstract than others, but taken together they offered a pretty strong read on where the market thinks marketing is heading next.
What stood out most was not just the volume of AI talk – that was inevitable – but the way the conversation has shifted. It is no longer about agencies sprinkling AI on top of existing services and calling it transformation. The stronger stories were about redesigning workflows, rethinking production, collapsing the distance between media and creative, and helping clients adapt to a world where both people and machines shape brand performance.
Infrastructure
Perhaps the biggest shift is that AI is no longer being presented as a clever efficiency layer sitting off to the side. Agency after agency talked about it as the infrastructure behind briefing, audience planning, creative development, production, deployment and optimisation. In other words, AI is becoming the operating layer of modern marketing. That is a much bigger claim — and a much more disruptive one — than simply saying it can save time on versioning or automate a workflow.
Systems
Creative, media and production are collapsing into one system. We identified this trend last year but in 2026, global agencies are making it happen. The old model — where strategy hands to creative, creative hands to production, and media comes in later to distribute it all — now looks far too slow and far too fragmented. The emerging view is that media signals should shape the work from the start, with some arguing that production should be part of the strategic conversation rather than the final execution layer. In that sense, the most interesting agencies are no longer talking about disciplines. They are talking about systems.
Content
The market is moving beyond ‘content at scale’. One of the most refreshing themes was a growing scepticism about the old ‘personalisation at scale’ promise. There was a clear sense that brands have spent years producing huge amounts of content that was technically efficient but creatively forgettable. A lot of digital landfill, to borrow one phrase. The more compelling alternative is not more content, but better content: work that is relevant to the audience, right for the platform, grounded in context and strong enough to actually make someone do something.
Context
Signals are replacing static segmentation. Another recurring thread was the move away from rigid audience definitions toward signals, behaviours, intent and context. That makes sense in a world where people move fluidly across platforms, interests and identities, and where algorithms respond to behaviour faster than marketers can update a segment. The implication is significant: agencies are increasingly designing for dynamic systems, not fixed audience boxes.
Tech ready
Technology is no longer the barrier to the integration of AI across all systems and processes – the technology is there but we humans are struggling to keep up. Leadership, trust, legacy processes, training, incentives and decision-making all came up repeatedly. In fact, some of the smartest points in the meetings were not about the technology at all, but about the uncomfortable reality that many organisations are still trying to force new capabilities into old structures. Ultimately workflows will need to be redesigned from the bottom up – AI is not a bolt-on.
Orchestrators
Open ecosystems are winning over closed models. Another interesting pattern was how many agencies now want to position themselves as orchestrators rather than owners of a closed stack. Whether they called it open architecture, modular tooling, orchestration or an open garden, the message was broadly the same: no single platform will be enough, and no one model will stay ‘best’ for long. The advantage will come from stitching the right tools, data sources, talent and governance together in a way that feels usable and adaptable.
Reframed
Brands now need to work for humans and machines. Maybe the most fascinating shift of all is that brand building is now being reframed for an algorithmic world. A few agencies talked explicitly about the need to earn trust twice: once with people, through emotion and meaning, and again with machines, through clarity, structure and recognisability. That tension feels important. The future is not just about being memorable to consumers; it is also about being legible to platforms, algorithms and large language models that increasingly shape what gets seen, surfaced and selected.
Shift
So, the biggest takeaway from New York? The conversation has moved on. The agencies that felt most convincing were not simply promising faster content, cheaper production or shinier tools. They were describing a more fundamental shift in how marketing gets done: more connected, more signal-led, more adaptive, and much more intertwined with technology. The opportunity is obvious, but so is the challenge. The winners will not just be the ones with the best AI story. They will be the ones that can turn all of this into a model that people can actually use.
These insights were gathered at the recent AdForum Summit in New York, an event where agency CEOs and leadership teams present their strategy and vision to the world’s leading search and management consultancies.
Sarah Probert, is a partner at Pt78, a consultancy specialising in marketing effectiveness. Our services include marketing ecosystem design and pitch consultancy; sarah@pt78.ie
