MeasuringWhatMatters

The Attention Metrics

Attention measurement now exists as an industry, and it is worth saying what kind of evidence it produces before leaning on it. Lumen Research and its partners run commercial eye-tracking panels at scale, and the metrics they have built fit how the scrolling economy actually behaves better than anything the impression era produced. They are also vendor panels: the samples and methods are not published the way a journal would require, and this book has spent two chapters explaining why that matters. Use them as the best available instrument, not as a finding. Where a peer-reviewed number exists for the same question, prefer it.

The metrics that matter break into three categories:

1. Attentive seconds per impression (APM).

How long did a human eye actually rest on this creative? Not how long it was on screen, but how long someone looked at it. This is the fundamental unit of attention measurement, and there is peer-reviewed evidence that buying more of it pays: Simonov, Valletti and Veiga measured one extra second of gaze as worth 3.43 percentage points of ad recall and 0.73 of purchase intent (Journal of Marketing Research, 2025). What it does not do is scale in a straight line. Goldstein, McAfee and Suri found the returns on exposure time to be diminishing, with the first seconds worth disproportionately more than the ones after them (EC ’11, 2011). For a designer that is the more useful shape, because it puts the value at the front of the window instead of spreading it evenly across one. And the wider link between attention and effectiveness is a live argument rather than a settled one: Hartnett and colleagues report that eye-tracking measures have trouble telling high attention from low (International Journal of Advertising, 2024).

Lumen calculates this as attentive seconds per one thousand impressions (APM) and the cost of that attention (aCPM, attentive cost per thousand). That reframes media buying: you buy attention instead of impressions, and you optimize for the quality of engagement per dollar rather than raw reach.

2. Cumulative attention across touchpoints.

The scrolling economy’s third law states that cumulative signal beats transactional impact. The measurement system needs to reflect this. Instead of evaluating each piece of creative in isolation, the metric should capture total attentive seconds accumulated across multiple impressions for the same viewer.

The evidence for this is the Havas, Lumen and Brand Metrics study introduced in Chapter 3, and it is worth being clear that this is one study and not two: multiple impressions of “long enough” attention drive better results than single impressions of sustained attention. The strongest movements in brand lift occurred when people engaged with multiple ads, each for a moderate amount of time. The same analysis found the reverse for preference and purchase intent, where a single high-attention placement did more, so this is a rule for awareness and consideration rather than a general one. The compound effect from Chapter 3 has measurement behind it, with that limit attached.

3. Active vs. passive attention.

Kantar’s attention framework (Kantar, attention measurement framework press release, 14 November 2023) distinguishes between passive attention (the viewer’s eyes are on screen) and active attention (the viewer’s facial expressions indicate emotional engagement). Kantar reports a minimal correlation between the two within its own database (Kantar, “Decoding Attention: Five Lessons to Transform Your Creative,” 2024). A viewer can be looking at an ad without emotionally engaging with it, and vice versa.

The distinction is not academic. Passive attention is necessary but insufficient. It tells you the creative was in the visual field. Active attention tells you it produced a response. For creative teams working signal-first, the active attention metric is what confirms the signal landed: that the color and composition produced a reaction rather than merely occupying the screen. What it cannot tell you is that the reaction happened before attention did. Emotional processing is not free of attention (Pessoa et al., 2002), and the narrow exception that survives is fearful faces at low spatial frequency, not brand creative.

What to Stop Measuring

Introducing new metrics is only half the change. The other half is de-emphasizing the metrics that create misaligned incentives.

Impressions should be reported but not optimized for. An impression without attention is a cost, not a result. Reporting impressions tells you how much inventory was purchased. It tells you nothing about whether the creative worked.

Click-through rate should be reframed as a downstream indicator, not a primary measure of creative quality. A high CTR on weak creative means the media targeting did its job. A low CTR on strong creative might mean the creative is building brand recognition that converts through a different path: through search, word of mouth, or the cumulative effect of signal deposited across hundreds of micro-moments. CTR measures the transaction. Attention measures the investment.

Engagement rate (likes, comments and shares as a percentage of reach) measures social performance, not brand performance. Creative can generate high engagement by being provocative without depositing any brand signal at all. The benchmark reports usually cited to prove this are not solid enough to carry it: the most-quoted of them states on its own pages that it “contains 2025 values, presented as 2026,” and reports TikTok moving in two directions inside the same document. That report is out of this book. The claim rests on its logic, not on their numbers. Engagement is getting more passive. The feed is changing, and the metrics need to change with it.

None of these metrics are useless. They each measure something real. But once they become the primary evaluation criteria for creative work, they optimize for transactions instead of attention, and for platform performance instead of brand performance.

Selling Attention Measurement

The hardest part of adopting attention-based measurement is not technical. The tools exist, the data is available, and credible institutions have already shown the link between attention and business outcomes.

The hardest part is cultural. You have to convince marketing leaders, CMOs, procurement teams and agency partners that the metrics they’ve relied on for two decades are incomplete. That an impression is not a view, and a click is not a brand interaction. The number that is easiest to report is not the number that matters most.

The argument that works is economic. The Newsworks, Lumen and Peter Field analysis introduced in Chapter 3 found that high-attention media plans grew market share twelve percent more than low-attention plans, with the self-selection that comes with an award-entry databank. Display ads on trusted news-brand sites received forty percent more attention than those on non-news sites. High-attention campaigns achieved fifty-eight percent more attentive seconds per advertising pound spent.

Kantar’s data backs this from the creative side: seventy-one percent of marketers say attention has an important influence on creative effectiveness (Kantar, “Media Reactions,” 2024). Yet only thirty-one percent of consumers globally say ads on social platforms capture their attention, down from forty-three percent the year before. The gap between what marketers believe and what consumers experience is the gap attention measurement closes.

The creative team’s job in this conversation is to close the loop: show that signal-first design and recognition hierarchy produce creative that earns more attention per impression, then show with data that more attention produces more business value. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit. Be careful with that last step, because it is the weakest joint in the chain and the client’s finance team may already know it. Shapiro, Hitsch and Tuchman estimated television advertising elasticities and returns for 288 brands and found “negative ROIs at the margin for more than 80% of brands,” with the observed advertising schedule profitable for only a third of them (Econometrica 89(4), 2021). Lodish and colleagues, meta-analyzing 389 split-cable field experiments, found no strong relationship between the standard recall and persuasion copy-test measures and sales effectiveness (Journal of Marketing Research 32(2), 1995). Neither result says attention is worthless. Both say that a mid-funnel number moving is not the same as money moving, and that most advertisers are already spending past the point where more of it pays. The claim you can defend is narrower and still worth making: attention is closer to what the creative controls than an impression is, and it is the first link in the chain a creative team can move on purpose. Sell that. Do not sell a straight line to profit.

Measurement is not the opposite of creativity. It is the proof that creativity works, and it is the only argument that survives the room where the budget is decided.