The Attention Economy Is Dying ## And What Comes After --- **Eleanor Whitfield** *Kelford Press* --- First published in 2026 by Kelford Press Copyright © 2026 Kelford Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means — electronic, mechanical, photocopying, recording, or otherwise — without the prior written permission of the publisher, except in the case of brief quotations embodied in critical reviews and certain other non-commercial uses permitted by copyright law. ISBN 978-1-7394521-0-8 (Digital) ISBN 978-1-7394521-1-5 (Print) ISBN 978-1-7394521-2-2 (Audio) Kelford Press Where Words Find Their Home Est. 2006 www.kelfordpress.com --- *For everyone who ever closed a tab and opened a book instead.* --- ## Contents 1. The Fifteen-Second Funeral 2. Peak Content 3. The Authenticity Paradox 4. The Collateral Damage 5. The Resistance 6. The Economics of What Comes Next 7. The Post-Attention World Acknowledgements About the Author Also by Kelford Press --- # Chapter 1: The Fifteen-Second Funeral --- Adunni Okafor pressed her thumb against the glass and held it there for a beat longer than necessary, as though the act of posting deserved some small ceremony. It was 11:47 p.m. in Lagos, and the ceiling fan above her desk stirred the humid air without conviction. She had spent fourteen hours on this video — a seven-minute investigation into how Chinese textile imports were reshaping the markets of Balogun and Oshodi, filmed on two phones and edited on a cracked laptop running DaVinci Resolve. She had interviewed three traders, a customs broker, and an economics lecturer at the University of Lagos. She had colour-graded the footage to feel warm but serious. She had rewritten her script four times. The upload bar filled. Instagram's servers, humming in data centres she would never see, ingested her work. Within ninety seconds, the platform's recommendation engine had made its assessment: the video lacked the velocity markers — the first-frame hooks, the text overlays, the emotional triggering — that its classifiers associated with viral potential. It was served to a thin slice of her 11,000 followers. A handful watched. Most scrolled past. By midnight, the verdict was legible in the metrics: 200 views. No meaningful shares. The algorithm had already moved on. Adunni is not a real person, though she is drawn from composites of several creators I interviewed in Lagos and Accra between 2024 and 2025. Her experience, however, is so common as to be banal. Every day, millions of people pour hours of skilled labour into content that is dead on arrival — not because it lacks quality, not because it fails to inform or move or provoke, but because a sorting mechanism designed to maximise advertising revenue has determined, in fractions of a second, that it will not generate sufficient engagement to warrant distribution. The funeral is over before most mourners know there was a death. This book is about the system that made that possible, why it is failing, and what is already beginning to replace it. But first, we need to understand how we arrived here — how human attention became the most traded commodity on Earth, and why the market for it now approaches something that looks very much like collapse. --- ## The Economist Who Noticed In 1971, Herbert Alexander Simon sat in his office at Carnegie Mellon University in Pittsburgh and wrote a sentence that would, decades later, be quoted so often and so carelessly that its original meaning would be almost entirely lost. "A wealth of information," he observed, "creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it." The sentence appeared in a paper titled "Designing Organizations for an Information-Rich World," published in a collection edited by Martin Greenberger for Johns Hopkins University Press. The context matters, because the context is precisely what the Silicon Valley mythmakers tend to omit. Simon was not writing about the internet. The internet, in 1971, was a handful of university mainframes connected by telephone lines; the first email would not be sent until later that year. He was not writing about social media, about advertising, about content creation, or about any of the phenomena later gathered under the banner of "the attention economy." He was writing about organisations — about how large bureaucracies and corporations could make better decisions when overwhelmed by data. His concern was managerial. How should a firm allocate its limited cognitive resources when available information exceeds the capacity of any individual or committee to process it? Simon was a polymath — Nobel laureate in Economics (1978), pioneer of artificial intelligence, contributor to cognitive psychology and political science. He moved between disciplines not out of dilettantism but because the problems he cared about refused to stay within departmental boundaries. His observation about attention was almost modest — an aside within a broader argument about information system design. The bottleneck in any information-rich environment, he argued, is not the information itself but the human capacity to attend to it. More reports, more memos, more data do not automatically produce better decisions. They may produce worse ones, by scattering focus and rewarding whoever shouts loudest. What Simon did not anticipate — what no one in 1971 could have anticipated — was that his observation would become the founding logic of an industry worth over seven hundred billion dollars. He described a problem of organisational cognition. The technology companies of the late twentieth and early twenty-first centuries built something different: a marketplace in which human attention itself was the product, harvested at scale and sold to the highest bidder. The distance between Simon's insight and what it became is the distance between noticing that a river flows downhill and building a hydroelectric dam. The observation is neutral. The infrastructure is not. --- ## The Long History of Selling Eyes Tim Wu, the Columbia Law professor and later Biden White House competition adviser, published *The Attention Merchants* in 2016. The subtitle — *The Epic Scramble to Get Inside Our Heads* — gestured at melodrama, but the history was careful and genuinely illuminating. Wu argued that the commodification of attention did not begin with Google or Facebook. It began in the 1830s, when Benjamin Day launched the *New York Sun* at one penny — far below production cost. Day's innovation was economic, not journalistic: a newspaper could be given away cheaply if advertisers paid to reach the large audience a cheap paper attracted. The reader was not the customer. The reader was the product. This model — attract attention with content, then sell that attention to advertisers — proved so durable that it would underpin nearly every mass medium for the next two centuries. The patent medicine hawkers of the late nineteenth century perfected the attention-grabbing advertisement, promising miraculous cures in prose so lurid it could make a carnival barker blush. Radio arrived in the 1920s and, after a brief period in which it was imagined as a public utility, was colonised by commercial broadcasters who sold airtime to sponsors. Soap manufacturers funded serialised dramas — hence "soap operas." The attention was the wife at home; the product was Palmolive. Television intensified the exchange. By the 1960s, the American networks had refined a system in which programmes existed primarily as vehicles for advertising. Newton Minow, FCC chairman, famously described television in 1961 as "a vast wasteland" — but a vastly profitable one. The economics were simple and brutal: attract the largest possible audience, then sell thirty-second intervals of its attention to corporations. The Nielsen rating became the currency — a crude but effective measure of how many eyeballs were pointed at a screen at any given moment. What changed with the internet was not the underlying logic but the precision and scale of its application. --- ## The Machine Learns to Count The year 2000 deserves a marker in this history. In October, Google launched AdWords, a system that allowed advertisers to bid on specific search terms and pay only when a user clicked on their advertisement. The implications were revolutionary, though they unfolded slowly. For the first time, attention could be bought and measured with something approaching precision. A newspaper advertiser in 1950 knew roughly how many people subscribed to the paper, but had no reliable way of knowing how many had looked at his advertisement, let alone acted on it. A television sponsor knew the Nielsen ratings but could not distinguish between a viewer who watched the advert attentively and one who had gone to the kitchen for a cup of tea. The old line, attributed to John Wanamaker — "Half the money I spend on advertising is wasted; the trouble is I don't know which half" — had been the industry's rueful motto for a century. Google abolished that ambiguity, or appeared to. With AdWords, an advertiser could track how many people searched for "running shoes," how many saw the ad, how many clicked, and how many bought a pair. The feedback loop was immediate, granular, and intoxicating. Advertisers believed, for the first time, that they could measure the return on every dollar. This was both the great promise and the great trap. Because once you could measure clicks, clicks became the thing that mattered. And once clicks became the thing that mattered, the entire apparatus of content production began to orient itself around generating them. Six years later, Facebook introduced the News Feed. Before September 2006, Facebook had been a static collection of profile pages. The News Feed transformed it into a river — an algorithmically curated stream of updates, photographs, links, and advertisements in an endless scroll. The design borrowed from television's logic: keep the viewer watching; never let the screen go dark; fill every silence with something new. The News Feed's algorithm — later refined, expanded, and replicated by every major platform — was the mechanism by which attention harvesting became industrialised. It learned, through billions of micro-interactions, what kept each individual user engaged: what they clicked, what they lingered on, what they shared, what provoked them to comment. It then served them more of whatever that was. The system was not designed to inform, to educate, or to connect, though it sometimes did all three. It was designed to maximise time-on-site, because time-on-site was the inventory that Facebook sold to advertisers. Mark Zuckerberg, in a 2007 presentation to advertisers, described Facebook's ambition with a candour that would later seem incriminating: "Once every hundred years, media changes. The last hundred years have been defined by the mass media... In the next hundred years, information won't just be pushed out to people, it will be shared among the millions of connections people have." The framing was democratic. The economics were extractive. --- ## The Metrics That Ate the Internet By the early 2010s, a vocabulary had emerged to describe the units of the attention economy, and each metric carried within it a set of incentives that would quietly reshape how information was produced and consumed across the planet. **Clicks** were the original currency — the direct descendants of Google's pay-per-click model. A click indicated, at minimum, that a user had been sufficiently intrigued (or misled) to follow a link. The optimisation of content for clicks produced "clickbait," a term that entered common usage around 2014: headlines engineered for curiosity gaps ("You Won't Believe What Happened Next"), emotional provocation ("This Video Will Make You Cry"), or deliberate vagueness ("She Walked Into the Room. What Happened Next Changed Everything"). The form was perfected by outlets like Upworthy and BuzzFeed, though it spread rapidly to legacy media organisations desperate for digital traffic. **Views** — or impressions — measured how many times content was displayed, regardless of engagement. On YouTube, a "view" counted after roughly thirty seconds of watch time; on Facebook, after just three seconds — so low that a user pausing while scrolling registered as a "viewer." In 2016, Facebook admitted it had overestimated average video watch time by 60 to 80 per cent — a discrepancy that drove a massive industry "pivot to video" in which publishers laid off writers and invested in video production on the basis of fraudulently inflated metrics. **Time-on-site** measured how long a user remained on a platform or page. This metric incentivised infinite scrolling, autoplay videos, and the fragmentation of articles into multi-page slideshows — anything that kept the clock running. It also incentivised outrage, because outrage is sticky: a user arguing in the comments section of a divisive post is a user who has not left the platform. **Engagement rate** — the ratio of interactions (likes, comments, shares) to impressions — became the metric by which social media algorithms ranked content. High engagement meant wider distribution; low engagement meant oblivion. This created brutal Darwinian pressure: content that generated strong emotional responses (anger, fear, amusement, moral indignation) was rewarded with reach, while content that was merely informative, carefully argued, or quietly beautiful was starved of oxygen. Each metric was, at best, a proxy for human attention. None measured comprehension, satisfaction, trust, or value. A user who clicked an article, realised it was misleading, and left in disgust registered identically to one who read it carefully and found it useful. The metrics measured activity, not meaning — and the platforms, by optimising relentlessly for activity, built a machine extraordinarily good at capturing attention and extraordinarily bad at doing anything worthwhile with it. The result, by the mid-2020s, was an information environment in which the most successful content was not necessarily the most truthful, the most beautiful, or the most useful. It was the most engaging — where "engaging" was defined circularly as "whatever the algorithm decides to distribute." --- ## The Cracks in the Cathedral The system began to falter, as such systems tend to, not with a dramatic collapse but with a slow accumulation of contradictions. The first and most measurable crack was the decline of organic reach. On Facebook, organic reach for brand pages averaged roughly 16 per cent in 2012, fell to 6 per cent by 2016, and by 2024 sat at approximately 2 per cent. A creator who had spent years building 100,000 followers could now expect 2,000 of them to see a given post — unless they paid Facebook to distribute it. The platform had offered creators free tools to build communities, then throttled access to force the purchase of advertising. A bait-and-switch executed at civilisational scale, entirely legal. The second crack was ad-blindness. By 2013, studies found that most internet users had learned to unconsciously filter out banner advertisements. Eye-tracking research showed gazes skipping over ad placements entirely. By the 2020s, the average user was exposed to an estimated 6,000 to 10,000 advertisements per day, and the cognitive response was not attention but perceptual anaesthesia. The noise had become so constant that the brain treated it as silence. The third crack was economic. Between 2020 and 2025, the average CPM on Facebook rose