The just-released IAB and MRC’s proposed Attention Measurement Guidelines (May 2025) aim to bring consistency to an increasingly fragmented and hotly contested space in advertising research - the measurement of attention. With marketers and agencies alike looking for clearer signals of effectiveness and accountability, the guidelines represent a practical and much-needed step.
But from the perspective of evolutionary psychology—a discipline focused on understanding the cognitive architecture shaped by natural selection—we are not there yet. The way attention is conceptualised, measured, and applied in the guide requires a bit more scrutiny and is a fair way off a standard.
Yes, it does offer a solid foundation for standardising metrics, but they overlook the complex, content-sensitive, and adaptive nature of human attention itself.
Attention is not a general-purpose thing. It is a set of evolved mechanisms that prioritise information in line with ancestral survival and reproductive goals. Understanding this deeper architecture is absolutely essential if attention metrics are ever going to produce meaningful insights into consumer behaviour.
The guide defines attention as a thing – a combination of exposure, engagement, focus, and cognitive impact. This categorisation is perhaps understandable in the context of trying to find industry standardisation, but from a cognitive science standpoint, it conflates distinct processes. Attention is not a single function, but a multi-faceted effort of specialised systems evolved to track different kinds of information - things like threats, social cues, faces, movement, food, or potential mates.
In simple terms, the mind does not allocate attention based on format or placement, it prioritises based on relevance. What captures attention in the real world (and also in an advertising context) is not arbitrary. It’s governed by evolved biases that favour certain types of stimuli. Things like emotional expressions, gaze direction, stories, status signals—because those were critical to decision-making and survival in our evolutionary history.
Attention proponents are quick to call out how all reach is not equal, but fail to acknowledge that not all attention is equal either. A second of gaze time prompted by a sudden animation is not equivalent to a second of attention drawn by an emotionally engaging narrative or a prompt that aligns with the viewer’s current goals. Attention must be interpreted through the lens of why it occurred, not just that it occurred.
The measurement call up all the usual suspects, a combination of visual tracking (eye movements), physiological signals (heart rate, pupil dilation), and neurological observations (EEG patterns) to assess attention. These methods offer valuable data for sure, and the inclusion of biometric approaches is a welcome sign of progress beyond simple exposure metrics. They measure something. But, from an evolutionary standpoint, interpretation is still key. Reflexive attention, captured by movement, sound, or brightness, has high utility in identifying threats or urgent stimuli. But these mechanisms evolved to prioritise evolutionarily relevant information. A gaze fixation triggered by design elements or prominent placement might reflect an orienting response, not any deeper processing or relevance.
Similarly, elevated physiological arousal might signal interest—but it might also signal anxiety, discomfort, or confusion. Emotional engagement is not a binary state, and the human body often reacts strongly to both positive and negative stimuli. This complexity demonstrates a need to distinguish between different types of attention. Some responses are automatic and unconscious, others are sustained and cognitively effortful. Effective interpretation of attention metrics should take account of the underlying systems activated, not merely the intensity or duration of the response.
That the authors also include survey and panel-based self-report methods as a measure is cause for a sigh. Self-report measures come with limitations well-documented in psychology and market research. It should be well known that people are poor introspects when it comes to their own mental processes. In any case, people tend to report what they think they should have attended to, especially when asked directly. Social desirability, confabulation, and memory biases can all distort self-report data.
Perhaps the most fundamental challenge for any future attention measurement system (and much of market research as a whole) lies in the mismatch between the environments in which human attentional systems evolved and the unnatural environments in which they are now being studied. Our attention evolved in response to physical, social, and ecological threats and opportunities - managing face-to-face interactions, monitoring status cues, and navigating uncertain environments.
(Quantum factoid - when users sense they are being tracked (especially via eye-tracking or biometrics), their behaviour changes, potentially reducing natural attention or biasing self-reports - we should know by now that methods themselves can distort the very attention they seek to capture.)
The media environment we inhabit is designed to extract as much eyeball time as possible through artificial stimuli, things like outrage, exaggerated emotional cues, repetition, stupidity and hyper-personalised content. These techniques often succeed by activating our evolved biases. But over time, the mind learns to discount repeated patterns that offer no useful outcome.
Banner blindness, ad-skipping, and fast scrolling are not merely behavioural patterns, they are adaptive responses to stimuli that the mind has learned are irrelevant. This makes ‘attention’ both more valuable and harder to capture. It also raises the bar for creativity. Content that taps into core motives—social bonding, moral emotion, curiosity, or narrative meaning—has a far better chance of sustaining interest than content that simply exploits attention reflexes.
‘Emotional’ response in the guide seems to be measured as a overly simple positive/negative scale - perhaps the authors are too much in thrall to the mysterious Feldman-Barrett school of thought - but the established evolutionary science shows clearly that emotions evolved as specific programs for specific challenges (fear for threats, disgust for contamination, etc). Ads that trigger or match the right emotion for the context are more likely to drive action than those that simply produce arousal. Nuance matters. Measurement should distinguish which emotions are triggered and in what context, not just how much.
Despite these concerns, the IAB/MRC guidelines do represent a meaningful step toward better measurement. They recognise the importance of triangulating different types of evidence. They also acknowledge that attention is not an end in itself, but a signal—one that should be interpreted alongside other outcomes such as memory, choice, and behavioural response.
Natural selection has designed attention as a means to an end, it exists to guide adaptive behaviour. Measuring it is valuable not because it is necessarily a good to capture more of it, but because doing so can tell us something about how stimuli interact with the cognitive systems that evolved to prioritise relevance and value.
But attention is not a commodity to be harvested, it’s a fleeting, adaptive filter. If we want to better understand how advertising works—and how to make it work better—we need to understand what human minds evolved to care about.
Metrics that help us quantify attention are useful. But metrics interpreted in the absence of theory risk becoming distractions in themselves. The evolutionary perspective reminds us that the best creative and media work doesn’t fight against the mind’s architecture or human nature. It works with it.
Eaon Pritchard | Strategy | Consumer psychology