The AI era is now here, and marketing teams are facing into the biggest wave of change our discipline has ever needed to ride.
For the past two decades, scaling ‘marketing output’ almost certainly meant scaling marketing headcount, supporting more channels, more content, more events, more campaigns, more geographies.
That model is now changing as agentic AI gets woven into marketing organizations.
AI isn’t simply another layer in the tech stack: it’ll rewire how marketing work gets done. Executed well, it can drive step changes in both effectiveness and efficiency, and will certainly reshape the structure and makeup of our teams.
The shift isn’t cosmetic, it’s structural, and already happening at pace.
This article examines the likely impact of agentic AI adoption within B2B marketing teams, the implications for org design and skillsets we require, as well as some of the challenges we need to face into.
The structural change: from pyramid to ‘superhuman diamonds’
Today, most marketing org designs still resemble a pyramid. At the base sits a broad layer of executional resource – copywriters, designers, campaign managers, digital specialists, analysts, coordinators. Their role is to produce, activate, report.
Above that, a layer of middle management coordinates and oversees activity. Think campaign planning, stakeholder management, task allocation, performance oversight, and people management. These marketers sit between delivery and direction, in my experience often operating as player-coaches.
At the top, is marketing leadership. Setting strategy and direction, managing budget, and alignment to the wider business.
It’s a familiar model, but it is mismatched to the reality of how marketing is now starting to be delivered.
Much of what sits at the base of that pyramid is, at least in theory, automatable. Research from Anthropic suggests that around 65% of the tasks performed by marketing professionals could be handled by AI. So, content production, campaign execution, reporting, research and data analysis can all be handled – partially or, in some cases, almost entirely – through agentic workflows. As those capabilities mature further, the need for large executional teams starts to diminish. And that’s where the org shape begins to change.
Most commentators seem to agree that, as AI shifts from being a tool to something closer to an operator, the pyramid will compress and a different structure starts to emerge – one that resembles a diamond:
Narrower at the base, broader and more capable in the middle, and sharper at the top.
Execution doesn’t disappear, but it will become highly leveraged – one marketer, supported by a network of agents, will now deliver what previously required a team. AI stops being something you use, and becomes something you manage – effectively a peer on the org chart, within what some are calling “superhuman marketing organisations”.
The effect on the middle layer is just as pronounced. In this new model, Marketing Managers and Marketing Leads are no longer primarily managing people. Instead, they’re orchestrating systems. Their role shifts towards designing workflows, sequencing activity, integrating tools and coordinating a mix of internal capability, external partners and AI agents. It becomes less about supervision, and more about architecture.
At leadership level, the fundamentals remain intact – strategy, alignment and accountability. But the emphasis sharpens. And designing the system becomes as important as setting the direction.
Intelligence – across customer, market and competitor – also becomes a critical input, while commercial judgement matters more than ever: as execution becomes faster, cheaper and more accessible, advantage shifts away from the ability to do, and towards the ability to decide.
The skillsets and competencies we need to build
As the structure evolves, so too will the roles within it.
Some of these roles will be familiar, but with a different emphasis, others are only just starting to take shape. There is already a visible shift towards AI-enabled strategists, orchestration roles that sit across workflows and channels, architects who design journeys and systems rather than individual campaigns, AI operators who run the underlying data and AI infrastructure, and governance roles focused on risk, ethics and compliance.
Larger organisations are already hiring these as specialist roles, and a recent Microsoft study found that 79% of marketing leaders plan to hire for AI-specific roles within the next year.
For most marketing teams though, the exact titles and scope of roles will probably be less important than the underlying capabilities we’ll need to operate our new agentic-powered organisations:
- At the core is the idea of ‘AI fluency’ combined with a ‘systems thinking’ mindset. So, an understanding of AI capabilities and the ability to prompt a model, but also an understanding of how different components connect, how data moves through a system, and where decisions are being made. These are now vital skills for all marketers.
- Strategic judgement becomes more valuable, not less. When the cost and speed of execution collapse, the constraint is no longer delivery capacity, it’s choosing what to do (and what not to do), and why.
- Creative direction also steps up. AI can generate at scale, but it doesn’t replace taste, perspective, brand instinct. Someone still needs to define what “good” looks like, and appraise and manage output accordingly. Clear briefs become important here.
- Customer and market insight – marketing’s superpower – remain foundational. If anything, they become more important, as they shape the inputs that drive AI outputs. Poor inputs simply scale poor thinking.
- Data interpretation shifts from simply reporting to more proactive decision support. The challenge is no longer accessing data, but extracting signal and meaning from it, and then acting upon that with confidence. Further, genAI will allow for predictive modelling, at scale, and at speed, so marketers will be able to test campaigns before launch, and get data to reinforce recommendations.
- But running through all of this, though, is a need for ongoing curiosity. Change will be a constant, for years to come. The tools will continue to evolve, often faster than organisations can keep pace, so the advantage will sit with those who are willing to continually reassess how they work, rather than lock into a fixed model.
There is also an organisational question to resolve in the middle layer. Do you build broad generalists who can orchestrate across disciplines, or do you double down on depth through specialist centres of excellence?
The answer to that’ll likely depend on your marketing team size, and the number of business units being supported. But within larger organisations we’re already seeing hybrid approaches emerge: hub-and-spoke models, where centralised expertise supports decentralised teams, and ‘agile pods’ that combine different skillsets around a common objective with temporary teams forming as projects dictate.
New challenges we need to navigate
But for all the focus on efficiency and scale, the transition to a genAI-centric marketing world raises some new questions and challenges to consider.
The first is around talent. Historically, most marketers have learned through doing: writing copy, building emails, managing campaigns, erecting tradeshow stands, updating web pages, working through data in spreadsheets. These executional tasks were rarely glamorous, but were certainly formative, and provided foundations upon which judgement was built.
If much of that entry-level work disappears or is heavily automated, the pathways into our profession become fewer, and less obvious. Where will the next generation of middle managers come from? There’s a risk of creating a gap where fewer people have the opportunity to develop their craft through experience.
That places greater emphasis on structured development and learning, more intentional career pathways. Joel Harrison’s excellent blog explores this in detail – it’s a challenge that marketing leaders need to not only be aware of, but which they need to own.
That’s a medium-term problem to solve, but the second challenge is more immediate: increased cognitive load.
While AI will remove efforts from specific tasks, and will ultimately reshape the nature of work, recent studies have shown that it’s actually increasing workload in the short term.
A recent Harvard Business Review article introduced the concept of “AI brain fry” – the mental fatigue that comes from constantly interacting with multiple tools, reviewing generated outputs and keeping pace with accelerated workflows. A BCG study from early 2026 showed that marketing teams, sitting at the intersection of content, data and channels, are particularly exposed to this dynamic, and were found to be more likely to experience cognitive overload than other functions. Related, Marketing Week’s 2026 Career Survey showed that in 60% of marketers reported feeling overwhelmed, with remits expanding and marketers at all levels taking on more responsibility.
This isn’t a technology issue, it’s a management one, requiring clearer prioritisation, better workflow design and more disciplined use of tools.
The final challenge is differentiation. As access to AI becomes widespread, the baseline level of execution rises quickly. Activities that once provided competitive advantage become standard, and wat was once distinctive becomes expected.
That resets the bar. Advantage moves back towards areas that are harder to automate – strategic clarity, genuine customer insight, original thinking and commercial acumen. In many respects, the fundamentals of good marketing reassert themselves, but under greater scrutiny.
The AI-enabled marketing team of the future is not simply a more efficient version of today’s model: it’s a structurally different organisation. Smaller in some areas, broader in others, less focused on ‘marketing output’, and more focused on orchestration, decision-making and impact.
The organisations that succeed won’t be those that adopt AI tools the fastest, but those that are prepared to rethink how their teams actually work – how decisions are made, how workflows are designed, where real value is created, and the skills and competencies needed to support that.
This all needs thought, planning, and a purposeful management and change approach. (But if that’s something you need help with, I’d love to chat.)