There is a particular kind of institutional reluctance — polite, well-dressed, and ultimately self-defeating — that manifests when an industry confronts evidence that its foundational assumptions no longer hold. The publishing and marketing industries are, at this moment, engaged in precisely that reluctance with respect to the attention economy, and the cost of the delay is compounding by the quarter.
The foundational bargain of digital media, struck somewhere around 2004 and never quite made explicit, ran as follows: audiences would exchange their attention — rendered legible as page views, session durations, click-through rates, and scroll depth — for content that was free at the point of consumption. Publishers would monetise that attention through advertising. Brands would pay for access to the attention. The whole system was predicated on the audience having no alternative route to the information the content contained.
Large language models have destroyed that predicate. Not weakened it. Destroyed it.
When a prospective customer types a query into an AI assistant — and the evidence from multiple measurement studies is now unambiguous that this is happening at scale across every significant purchase category — they receive a synthesised answer that is assembled from the content of hundreds of publishers, none of whom are compensated for their contribution, and the prospective customer has no particular reason to visit any of the underlying sources. The attention that previously flowed through a publisher's front door now flows through an AI intermediary that has, from the publisher's perspective, no door at all.
The implications for brands are equally structural. The content marketing investments of the past decade — the blog posts, the whitepapers, the thought leadership series — were built on the assumption that they would generate organic search traffic, that traffic would generate awareness, and awareness would generate pipeline. Each of those links in the chain is being severed simultaneously. The content still exists. The AI is reading it. The traffic is not arriving.
I do not write this as a complaint. The organisations that will navigate this transition successfully are those that understand it as a design constraint, not a grievance. The brands that are investing now in ensuring their content is not merely crawlable but AI-legible — structured, attributed, authoritative, and continuously updated — are making a bet that compounds. The brands that are waiting for the search traffic to recover are making a different bet. I know which I would take.
The attention economy is not dead. It has been restructured, at considerable speed, around new intermediaries who play by different rules. The productive response is not to mourn the old rules but to learn the new ones faster than your competitors.