Where AI meets Marketing: Building Better
06 Mar 2026
In a world where everyone has AI, a new operating model may be the ultimate differentiator
With AI use rapidly becoming standard in most industries – if not all – it is no longer a capability that sets your brand apart. According to McKinsey, 72% of organizations are now using AI in at least one function, and 65% report using generative AI regularly. This is particularly pronounced among marketers: Salesforce reports that 86% already use AI in their role. Yet the results vary widely: some companies are doing extremely well in this new milieu, seeing both significant gains and measurable cost savings from its adoption; others are seeing little benefit. The difference can’t be the technology itself – so could it be the way the tech is being used?
From tool adoption to workflow redesign
In many cases, AI is being bolted on to existing processes rather than fully integrated throughout its steps, bringing speed and productivity but not necessarily greater quality. Companies that are seeing genuine growth from AI adoption are the ones redesigning workflows around its potential applications. To take a simple example, under the old way of working, a marketing manager conceives of a campaign. The creative team builds campaign assets, the media team buys ads and the campaign runs. Analysts review the results and the team adjusts the campaign accordingly. The natural decision might be to employ AI to help generate variations on the campaign assets, but the process and decisions remain the same. AI is useful in this context for speeding up asset generation, but it’s not being used to its full potential.
The companies making gains are the ones embedding AI into every step of the process: first, set it to work analyzing past campaign data to identify high-performing messaging and target audience to inform the current campaign. Armed with this information, the strategist can shape the new campaign around a clear evidence base. AI can then generate multiple versions of the campaign which the media team then micro-tests across segments, keeping costs lower while establishing which content garners the most engagement from which users. Together, marketers and AI can then shift budget allocation towards the strongest performing campaigns and make strategic adjustments. At every key stage, a human must be involved in the process because AI cannot “own” decisions.
The rise of the in-house agency model
Rising up to meet this structural challenge is the in-house agency model. With ever-tighter budgets and fewer resources, marketing teams are expected to produce more campaigns, more content, more channels and more data-driven insights. CMOs can’t hire more permanent staff, but with AI and a core/flex approach, they can bring in the skills and capabilities they need to fill the gaps. Rather than outsourcing content wholesale to an external agency, the organization constructs one internally, built of people with core skills (permanent staff) rendered flexible by the temporary addition of auxiliary capabilities such as specialist knowledge, data expertise and creative talent in the form of contractors, freelancers, external agencies, data analysts and AI tools. This brings all the capabilities the team needs under one umbrella without driving up costs. When workload ramps up, freelancers and AI models can be brought in to ramp up capacity. When it scales down, they move on to another project or company.
Because AI increases both the speed with which tasks can be completed and the sheer volume, this flexibility enables swift adjustments to team capabilities to scale up or down as business priorities shift on a quarterly or even a monthly basis.
The productivity paradox
But productivity alone is not enough. The competitive edge still lies in human judgement. As AI enables creative teams to produce far more content, in more variations than ever before, this in turn produces more data to analyze and thus more audience targeting options to test. Output increases dramatically – but this doesn’t automatically translate into better results. So what’s keeping ROI down for so many players?
It’s highly likely that the sea of sameness is diluting the efficacy and individuality of messaging. As organizations move toward using the same tools in identical ways, their output risks becoming generic, boiling away any semblance of uniqueness and originality. Because AI models learn by identifying and effectively “mimicking” the patterns found in large datasets, it is easy for them to converge on familiar, similar tones and patterns which quickly lose their spark. And this is where human intervention is indispensable: using judgement, intellect and creativity to steer the AI away from the generic and towards the surprising.
From headcount to capability
The combination of AI and a flexible talent model is immensely powerful – if used correctly. Rather than asking “who do we need to hire for the long-term future?”, marketing leaders should be asking “what capabilities do we need right now?” Increasingly, the answer lies in fluid capability networks rather than fixed teams, providing crucial insight into whether those capabilities are best supplied by a full-time employee, freelancer, specialist contractor, agency or AI tool. Networked operating models are based on permanent access to core skills with additional help “parachuted in” as required, whether those are human or AI.
As AI expands what marketing teams can produce, the real strategic challenge is not hiring more people but designing organizations that can access the right capabilities at the right time.
For more on how marketing teams are evolving, you might find these articles useful: “How Leading CMOs Are Reshaping Their Teams to Deliver More With Less” and “Why the Old Marketing Team Model No Longer Works“.
