For much of the past two years, conversations around artificial intelligence have been dominated by one theme: productivity. Organisations have understandably focused on how AI can help employees write faster, analyse data more efficiently, automate repetitive workflows, and reduce operational costs. These applications are already delivering measurable value across industries, and they representan important milestone in enterprise AI adoption.
Yet I increasingly believe they are not the most significant development taking place.
The larger transformation is beginning to emerge at a much deeper level, one that has less to do with completing work more efficiently and far more to do with improving the quality of organisationaljudgement itself. Across industries, AI is gradually moving from being a system that assists execution to one that actively informs decision-making. Rather than simply helping people perform existing tasks faster, intelligent systems are beginning to influence which decisions deserve attention, where resources should be allocated, how risks should be evaluated, and what actions are most likely to create favourable outcomes.
Marketing offers perhaps the clearest illustration of this transition because it has always operated under a fundamental constraint: most intelligence arrives after the money has already been spent.
For decades, brands have followed a remarkably consistent sequence. Campaigns are conceptualised, creative assets are produced, budgets are approved, media is purchased, and only then do organisations begin understanding whether those decisions were correct. Attribution models explain performance after campaigns have concluded. Brand lift studies arrive weeks later. Consumer research attempts to understand behaviour once it has already occurred. Even the most sophisticated analytics platforms remain largely retrospective, helping organisations interpret outcomes rather than improving the quality of decisions before those outcomes exist.
This has never been a failure of marketing. It has simply been the reality of working with imperfect information.
Artificial intelligence introduces the possibility of changing that reality.
The real opportunity is not that AI can generate advertisements more quickly or write marketing copy in seconds. Those capabilities, while impressive, remain productivity improvements layered onto an existing operating model. The more profound opportunity lies in whether intelligent systems can help organisations understand how consumers are likely to respond before creative reaches the market and before media budgets are committed.
If that becomes possible at meaningful scale, marketing changes from a discipline built around measurement into one increasingly centred on prediction.
That distinction matters because advertising remains one of the largest commercial investments organisations make. According to WARC, global advertising expenditure is expected to surpass one trillion dollars annually within the next few years. At the same time, decades of research from organisations including Nielsen, Kantar, Ipsos, and System1 consistently demonstrate that creative quality remains among the strongest drivers of advertising effectiveness, often contributing more to long-term commercial outcomes than improvements in targeting, media efficiency, or channel selection.
Despite recognising the strategic importance of creative, however, most organisations continue evaluating it through processes that remain remarkably subjective.
Large campaigns often involve months of strategic planning, multiple specialist agencies, behavioural research, production teams, and significant financial investment. Yet the final creative review frequently becomes a conversation shaped by individual preference rather than objective evidence. Stakeholders debate whether the visual feels premium enough, whether the narrative should be more emotional, whether the logo appears prominently enough, or whether a particular headline better represents the brand. None of these perspectives are inherently wrong because creativity has always required intuition, experience, and human judgement.
The challenge is that consumers do not process advertising the way organisations review it.
Long before someone consciously decides whether they like an advertisement, their brain has already processed dozens of subconscious signals. Visual hierarchy influences attention before language is interpreted. Colour affects emotional response before rational evaluation begins. Familiarity shapes trust, movement captures attention, novelty influences memorability, and cognitive load determineswhether information is retained or ignored. By the time consumers articulate why they found an advertisement persuasive, much of the underlying neurological processing has already occurred.
This observation has formed the foundation of neuromarketing research for decades.
Rather than relying exclusively on surveys or focus groups, neuroscience attempts to understand how people actually experience advertising by measuring physiological responses that occur while the experience unfolds. Technologies such as electroencephalography (EEG), functional magnetic resonance imaging (fMRI), eye tracking, facial coding, and electrodermal activity allow researchers to observe attention, emotional engagement, cognitive effort, and memory formation in ways conventional research methods cannot.
Importantly, neuroscience is not attempting to replace traditional consumer research. Instead, it addresses one of its most persistent limitations. People are often remarkably poor at explaining the subconscious processes that shape their decisions. Behavioural economists and cognitive scientists have repeatedly demonstrated that preferences are influenced by mental processes operatingbeneath conscious awareness, making retrospective self-reporting an incomplete measure of actual behaviour.
Over the past decade, an expanding body of research published in journals such as Frontiers in Neuroscience, Nature Human Behaviour, and the Journal of Consumer Psychology, alongside work from institutions including Harvard, Stanford, and MIT, has shown that neurological and physiological signals can significantly improve predictions of advertising recall, emotional engagement, and memorability when combined with conventional market research.
For most of its history, however, neuromarketing remained an observational discipline. Researchers measured how consumers responded after exposure to creative, generating valuable insight but still operating within the same fundamental sequence that has defined advertising for decades.
Artificial intelligence is beginning to change that equation.
Instead of analysing individual experiments in isolation, modern machine learning models can identify relationships across enormous behavioural datasets, recognising patterns that consistently correlate with attention, emotional engagement, recall, and behavioural response. As these models become increasingly sophisticated, the conversation gradually shifts away from measuring what consumers did and towards estimating what they are likely to do.
That evolution represents far more than another analytical capability.
It represents the emergence of predictive intelligence as a strategic layer within modern marketing, where the objective is no longer simply to understand performance after campaigns have ended, but to improve the quality of commercial decisions before investment has even begun.
The implications of this shift extend well beyond creative testing. If artificial intelligence becomes capable of estimating behavioural response before campaigns reach the market, then marketing itself begins operating on an entirely different logic. Historically, marketing technology has been designed to improve execution. Customer relationship management systems organised customer information, analytics platforms interpreted historical performance, marketing automation software executed predefined workflows, and media platforms improved distribution. Human teams remained at the centre of the operating model, connecting information across systems and making every consequential decision themselves.
Artificial intelligence is gradually changing that relationship.
What makes this transition particularly significant is that AI is no longer confined to performing isolated tasks. Increasingly, it is beginning to connect decisions across functions. Audience research informs creative development. Creative evaluation influences media allocation. Market signals reshape positioning. Competitive intelligence updates messaging. Performance data continuously refines future recommendations. Rather than existing as individual software applications, these capabilities begin functioning as an interconnected intelligence layer that sits above the traditional marketing stack.
This distinction is often overlooked because much of the public conversation continues to revolve around generative AI. The ability to create text, images, videos, or presentations has understandably attracted enormous attention, yet content generation represents only one capability within a much broader transformation. The greater opportunity lies in systems that can reason across multiple inputs, identify relationships between seemingly unrelated variables, and improve commercial judgement over time.
In many ways, marketing is moving from an execution problem towards a decision problem.
This evolution is already becoming visible across enterprise AI. Research from McKinsey suggests that the greatest long-term value of generative AI will come not merely from reducing labour costs, but from improving decision quality across knowledge-intensive functions. Gartner has similarly argued that agentic AI will increasingly enable autonomous decision-making within enterprise environments, while Boston Consulting Group has observed that leading organisations are beginning to redesign workflows around intelligent systems rather than simply inserting AI into existing processes. Although these developments remain at different stages of maturity, they collectively point towards the same conclusion. Artificial intelligence is gradually becoming part of how organisations think rather than simply how they work.
Marketing leaders therefore face a different challenge than they did even five years ago.
For decades, the role of the Chief Marketing Officer revolved around managing campaigns, agencies, technology vendors, budgets, and internal teams. Success depended on selecting the right channels, building strong creative capabilities, and ensuring operational excellence across increasingly fragmented ecosystems. Those responsibilities will remain important, but they are unlikely to define the role for much longer.
Increasingly, the CMO will become the architect of an intelligent operating system.
That operating system will not simply consist of software licences or dashboards. It will determine how information moves across the organisation, how intelligent agents collaborate with human teams, how behavioural signals influence commercial decisions, and where human judgement should override algorithmic recommendations. Leadership will gradually become less about managing activity and more about designing systems that consistently improve the quality of organisational decision-making.
This represents a subtle but profound shift in enterprise management.
Historically, competitive advantage was often associated with access to talent, capital, distribution, or proprietary technology. While those advantages remain important, another layer of differentiation is beginning to emerge. Organisations will increasingly compete through the quality of their decision architecture. Companies capable of interpreting behavioural signals more effectively, learning faster from fragmented information, and coordinating intelligence across functions are likely to outperform competitors with comparable resources but weaker operating systems.
This is also where neuroscience becomes particularly relevant.
For years, neuromarketing has been viewed as an interesting research discipline rather than a core business capability. Artificial intelligence has the potential to change that perception because it dramatically increases the commercial value of neurological insight. Instead of analysing isolated experiments involving hundreds of participants, machine learning enables organisations to recognisebehavioural patterns across vastly larger datasets, allowing predictive models to become progressively more accurate over time.
Equally important is the emergence of multimodal AI. Unlike earlier machine learning systems that specialised in a single type of information, multimodal foundation models can interpret text, imagery, video, audio, behavioural signals, and structured data simultaneously. This enables a much richer understanding of how consumers experience advertising because human perception itself is inherently multimodal. We do not process headlines independently from visuals or emotional tone independently from context. Decisions emerge from the interaction of multiple signals, and increasingly, AI systems are beginning to model those interactions.
Some researchers have described this direction as synthetic audience modelling or behavioural simulation. Although the technology remains in its early stages, the underlying ambition is significant. Rather than waiting for campaigns to reach millions of consumers before understanding likely outcomes, organisations may increasingly evaluate creative concepts against intelligent behavioural models capable of estimating attention, memorability, emotional engagement, and purchase intent before production budgets have even been approved.
Naturally, this raises important questions.
Prediction should never be confused with certainty. Human behaviour continues to be influenced by culture, timing, personal experience, economic conditions, and countless contextual variables that no model can perfectly capture. Creative breakthroughs often succeed precisely because they challenge historical patterns rather than conform to them. Similarly, governance, transparency, privacy, and the responsible use of behavioural data will become increasingly important as intelligent systems play a larger role in commercial decision-making. The organisations that build lasting trust will be those that combine technological capability with clear ethical frameworks and meaningful human oversight.
For that reason, I do not believe artificial intelligence will replace creativity.
Instead, it will change where creativity creates value.
When intelligent systems become increasingly capable of evaluating probabilities, recognising behavioural patterns, and reducing uncertainty, human creativity becomes even more valuable because it shifts towards the work machines cannot easily replicate. Original thinking, cultural intuition, strategic judgement, ethical reasoning, and the ability to identify opportunities that lie outside historical data will remain fundamentally human capabilities. AI will strengthen these qualities by providing better intelligence rather than replacing the people responsible for exercising them.
Ultimately, this is why I believe the most important transformation taking place is not technological. It is organisational.
For decades, marketing has measured what happened after decisions were made. Artificial intelligence creates the possibility of improving decisions before consequences unfold. That seemingly simpleshift changes how organisations allocate capital, evaluate risk, develop creative work, and build competitive advantage. Over time, the companies that outperform are unlikely to be those producing the greatest volume of AI-generated content or deploying the highest number of AI applications. They will be the organisations that build intelligent systems capable of combining machine intelligence with human judgement to make consistently better decisions.
If that future unfolds as many current developments suggest, artificial intelligence will not be remembered primarily for automating marketing.
It will be remembered for transforming how organisations think before they act.
