Agentic AI & The End Of Persuasion

Advertising has never been an industry to concern itself much with economics, but in the interests of self-preservation, now might be a good time to start. Classical economics holds that in a marketplace with perfect information, advertising does not exist. Consumers make choices based on a fully informed understanding of their options. Advertising creates no economic value. It is, quite simply, irrelevant.

Until recently, this idea of perfect information felt academic — interesting, but safely abstract. Agentic AI has brought that reality much closer, much faster. We are now entering an era of near-perfect information, accelerated not just by digital platforms but by the emergence of AI systems that do not merely assist consumers, but act on their behalf. These agents increasingly search, compare and transact with minimal human involvement. As they optimise decisions using comprehensive, real-time data, consumers become less present in the shopping journey itself, delegating choice to machines designed to maximise value against clearly defined preferences.

In this environment, advertising built on information asymmetry, tactical persuasion and point-of-sale influence becomes structurally weaker. When claims can be verified instantly, alternatives compared exhaustively and outcomes optimised algorithmically, persuasion gives way to proof. Brand promises are no longer assumed; they are tested, ranked and validated. Shouting louder, targeting harder or repeating messages more frequently does not create advantage when machines coolly evaluate brands on evidence rather than rhetoric.

Yet for much of the past decade, advertisers have doubled down on tactics that assume presence and persuasion are enough. Data-driven targeting. Short-term activation. Relentless optimisation. Brands collect vast amounts of data, then deploy it to chase us across the internet with ever more precisely targeted messages. Programmatic advertising ensures brands are always on, always present. As Marc Pritchard, Chief Brand Officer of Procter & Gamble, once observed: “We tried to change the advertising ecosystem by doing more ads, and all that did was create more noise.”

In a world moving toward agentic decision-making, this approach is not just limited — it is misdirected. The problem is not the technology; it is what the technology is being asked to do. You do not influence people — or the machines acting for them — by being boring, repetitive or relentlessly annoying.

The strategic response for brands is therefore twofold.

First, brands must be genuinely clearer — and genuinely better — about the total value they create and deliver. Not just price or product features, but reliability, service, fairness, progress and long-term benefit. Crucially, this value must be explicit, real and tangible — structured and provable so intelligent agents can find it, interpret it and evaluate it accurately. In an agentic economy, being good is not enough. Value must stand up to scrutiny.

Second, brands must win with consumers before the shopping battle is even fought. As decision-making is increasingly delegated, the role of advertising shifts upstream — from influencing choice at the moment of purchase to shaping the emotional frames, beliefs and preferences that guide both consumers and the agents acting for them. Emotion does not disappear in a world of AI; it becomes the lens through which recommendations are interpreted. A brand that has already earned trust, affection or meaning benefits from confirmation bias — its recommendations feel safer, more aligned, more “right”.

This is something great brands have always understood. Nike, Apple, John Lewis and Coke know that influence has to be earned, and that the real USP of advertising is its ability to move us emotionally. The effectiveness of emotional communication has been consistently demonstrated in the work of Les Binet and Peter Field, reinforced by more recent studies from System1 and long championed by Mark Ritson. At its best, emotional advertising does more than communicate information; it alters the brand itself by attaching powerful associations to its meaning.

The job of advertising is therefore no longer to fight in the transaction itself, but to win hearts before the agentic AI’s dogs of war are unleashed — or at the very least, to tilt the emotional frame through which recommendations are sought and judged. In a world approaching perfect information, rational optimisation determines what is best. Emotion determines who we are willing to believe in.

Advertising will never win an information war against the internet — or against AI. But it can still win the battle for meaning. In an agentic future, that may prove to be its most valuable role of all.