The Attention Economy: What Advertisers Keep Getting Wrong

Most Ads Aren't Seen. They're Just Served.

Dentsu's 2024 attention report found that only one third of ads receive full attention. Not one third of bad ads. One third of all ads. The industry has spent decades optimising for impressions, viewability, and click-through rates while the majority of served ads pass in front of human eyes without registering a thing.

This is the central failure of attention economy advertising. Advertisers conflate delivery with attention. They measure what the platform confirms was served, then act as if the consumer saw it. These are not the same thing.

At its core, the thesis here is simple: most advertisers are optimising the wrong variable. They are buying exposure in environments where attention is scarce, poorly measured, and unevenly distributed. And as platforms get more expensive and AI reshapes how people consume information, the gap between served and seen is getting wider, not narrower.

The 2.5-Second Problem

Karen Nelson-Field and her team at Amplified Intelligence have done more to quantify this gap than almost anyone in the industry. Their research found that ads need approximately 2.5 seconds of active attention to form a memory trace. Not 2.5 seconds of the video playing in a tab. Not 2.5 seconds of technical viewability. Active, human attention.

A large share of online ads never get there.

Memory formation is the whole point. An ad that doesn't form a memory doesn't build mental availability. It doesn't shift brand preference. It doesn't warm the 95% of your audience who aren't in-market right now. It just burns budget.

Firms like Lumen Research, Adelaide, and Amplified Intelligence have built entire businesses around measuring whether this threshold is actually being reached, because the platforms themselves have no commercial incentive to tell you when your ad failed to land.

Attention Is Not Uniform. Advertisers Treat It Like It Is.

Here's where the conventional approach breaks down completely. Attention is not a binary. It exists on a spectrum, shaped by context, platform, format, creative quality, and the mental state of the person seeing the ad.

Chapter 12 of Never Always, Never Never covers exactly this: the attention spectrum from passive to active, and why designing for only one end of it is a strategic mistake. Patrick Gilbert makes the case, drawing on Daniel Kahneman's System 1 and System 2 framework, that most people spend most of their time in a low-attention, automatic mental state. They're not evaluating your ad. They're driving to work, half-watching TV, or scrolling Instagram before bed.

The creative strategy implication is significant. An ad that demands active engagement from a distracted audience will be ignored. An ad designed for the passive, automatic mind of a consumer who isn't actively shopping can still do real work: it builds familiarity, plants a brand memory, and keeps you in the consideration set for when that person does become a buyer.

That is not an argument for dumbing down your advertising. It's an argument for matching your creative approach to the actual attention environment you're buying into. A person might scroll past your ad at noon and watch it all the way through at 9 p.m. Context shifts attention. Creative that works across both states requires real variety, not just cosmetic variation.

Supporting this is the emotional advertising effectiveness literature. Research by Les Binet and Peter Field, drawn from the IPA Databank, shows that emotionally driven campaigns outperform rational, hard-sell approaches on long-term business outcomes. Emotional ads are better suited to the low-attention environments where most advertising actually runs, because they don't require the viewer to process a detailed argument. They just need to feel something.

The Attention Myth Advertisers Need to Drop

Widely repeated claims that consumers have shrinking attention spans drive a lot of bad creative decisions. Shorter. Faster. Louder. Interrupt before they scroll.

Addressing this directly, a chapter in Never Always, Never Never cites a study published in Nature Human Behaviour that tracked adult attention performance over three decades and found no meaningful decline. Attention spans have not shrunk. What has changed is selectivity. Consumers have become better at filtering out things that aren't worth their time.

That's a fundamentally different problem. If your ad is being ignored, the answer is not to make it shorter or louder. Making it worth the viewer's attention is the real fix.

People will watch a three-hour Joe Rogan podcast, spend an entire weekend on a LEGO build, or binge a ten-part documentary series. They will also skip your ad in half a second. The difference is not attention span. It's perceived value.

Reframing the problem matters for how you brief creative teams. You're not designing for people who can't focus. You're designing for people who have become ruthlessly efficient at discarding things that don't serve them. An ad that earns attention, whether through entertainment, relevance, or emotional resonance, will get more of it than most advertisers expect.

Closely connected to this is the attention spectrum framework explored in the book: the goal is a creative portfolio that can work across the full range of mental states your audience moves through, not a single format optimised for one end of the spectrum.

What AI Is Doing to Attention Inventory

Attention scarcity is getting structurally worse, and AI is accelerating it from two directions.

First, AI-generated content has massively increased the supply of content competing for attention. More content, same amount of human attention. Basic economics suggests this should drive down the value of any individual impression, but in practice it appears to be intensifying competition for verified human attention, which is making high-quality attention inventory more expensive, not less.

Second, AI is physically intercepting the path between advertisers and audiences. Seer Interactive's analysis, cited in 2025 industry reporting, found that organic click-through rates drop by 61% when a Google AI Overview appears on a query. Paid clicks fall by 68% on the same queries. Pew Research Center data shows that 26% of users end their browsing session entirely after reading an AI answer. They got what they needed. They left. No click, no ad, no impression.

For advertisers who built their model around search intent, this is a material shift. The zero-click search problem isn't theoretical anymore. Queries that used to generate high-intent traffic are increasingly resolved before the user ever reaches an advertiser-controlled environment.

Some analysts argue this shifts the priority from attention to relevance, specifically being cited or referenced within AI-generated answers rather than competing for clicks downstream. Whether you frame it as an attention economy or a relevance economy, the underlying problem is the same: the old model of buying impressions and expecting conversions is losing structural integrity.

Arbitrage in digital marketing, when cheap clicks were reliably available, is over. Google Ads CPC rose 31% between 2022 and 2024, according to Instreamly's 2026 industry report. Digital ad spend grew 14.9% year-over-year in the same period. More money chasing a shrinking pool of high-quality attention.

PwC Strategy& estimated the UK Digital Attention Economy alone at £21 billion in 2023, with global advertising spend projected to reach £690 billion by 2027. At that scale, the efficiency of how that money is deployed becomes a critical question.

The 95/5 Problem and Why Attention Measurement Makes It Worse

Most digital advertising is pointed at the wrong people anyway.

John Dawes and the Ehrenberg-Bass Institute have shown that at any given moment, only about 5% of your potential audience is actively in-market for your product. Everyone else is out-of-market, for now. Performance campaigns, constrained by ROAS targets and algorithm optimisation, effectively chase this 5% repeatedly while ignoring the 95% who will eventually become buyers.

The 95-5 rule carries a specific implication for attention strategy: the attention you most need to earn is from people who aren't yet paying attention to your category at all. That requires brand building vs direct response thinking, not just performance optimisation. Emotional, broad-reach campaigns that plant brand memories in the 95% are what build the pipeline that makes future performance campaigns work.

Agencies and vendors often focus attention metrics on whether individual ads are being seen, which is valuable, but rarely connect this to the question of who is seeing them and whether those people will matter to the brand in six months.

Brand advertising lifts performance across all channels through the halo effect in marketing. Build enough familiarity with the out-of-market 95%, and your performance campaigns get cheaper and more effective when those people enter the buying window. Neglect that 95%, and you end up bidding harder and harder against competitors for the same narrow slice of ready-to-buy traffic.

At AdVenture Media, this tension between short-term performance and longer-term attention investment comes up constantly when working with brands that have been purely performance-focused.

What Advertisers Should Actually Do

Measuring attention is a legitimate step forward from viewability. If Dentsu is right that only a third of ads get full attention, then buying on attention-adjusted metrics rather than raw impressions is a meaningful improvement. Firms like Lumen Research, Adelaide, and Amplified Intelligence offer ways to measure this that go well beyond what platforms report natively.

But measurement is not strategy. A better attention metric still doesn't tell you whether you're reaching the right people, building the right memories, or investing at the right ratio between brand and performance.

Evidence from Les Binet and Peter Field's IPA analysis points toward a 60/40 brand performance split as a reasonable starting point for most established brands. Most mid-market advertisers are nowhere near that. They've built their entire model on the assumption that the people who convert today are the ones worth spending on, which means they're consistently underfunding the work that creates future converters.

Meanwhile, emotional vs rational advertising research is clear: emotional campaigns are better suited to the low-attention environments where most advertising runs, and they produce business effects that accumulate over time rather than decaying quickly.

Practical conclusions follow directly from this:

• Stop treating attention as a single variable. It's a spectrum, shaped by context and creative.
• Measure attention properly. Served impressions and technical viewability are not proxies for memory formation.
• Design creative for the actual attention environment, not the idealised one where your audience is leaning in.
• Invest in the 95% who aren't buying today. They will be. Brand advertising works on a longer clock than most performance dashboards can see.
• Watch what AI is doing to search inventory. A 61% drop in organic CTR on AI Overview queries is not a temporary anomaly.

Attention economy dynamics are real. Where most advertisers go wrong is treating attention as a targeting problem when it's actually a creative and strategy problem. You can buy the best placements in the world, hit every viewability threshold, and still fail to earn a single memory if your ad isn't worth two and a half seconds of someone's actual focus.

That's the gap. And closing it starts with being honest about what you're actually buying.