The way we trade attention in the form of digital media in a murky market has led to a deep misunderstanding of this complex area, argues Faris Yakob - but some thinkers are advocating for alternatives.

Early this year I attended the inaugural meeting of The Attention Council in London where I saw presentations on the current thinking on attention in advertising, a topic in which I have a passing interest. A good deal of attention is being paid to attention of late, and the council’s mission is to “promote the use of attention metrics to create incentives that align all stakeholders in the media and advertising ecosystem”. This is a focused and intelligent approach, because underlying all media discussions are metrics, currencies and pricing. 

At the heart of the media business is the impression. In order to buy and sell access to human attention a currency was needed. In the pre-digital world, an impression was a set of eyeballs - one person exposed to one advertisement once in any medium. It allowed for cumulative and comparative analyses of different media, which was very useful for cross channel planning and optimisation.

It was also inherently probabilistic, since one impression was only ever an opportunity to see [or miss]. When digital media arrived, it allowed for a much more certain measure, but we opted for one ad being served, making the presence of a person secondary. This led to years of bizarre discussions in which the idea of ‘viewability’ was championed. Bizarre because anyone outside of a media trading environment could have easily explained that an ad no one sees has zero impact. Viewability went on to be defined meagrely, since only 50% of the ad must appear for one second to count [for the industry but arguably not for consumers]. 

Further, viewability only gets us back to the original conditions of media in the sense that we can have some confidence that our advertising actually could be seen by people, but it doesn’t tell us anything more about whether it is or the quality of attention that medium delivers. This matters because eye-tracking research from [founder member of the council] Lumen suggests that online “82% were completely missed, not recording a single moment when the user’s eyes were fixated on them.” 

That said, we now have data that helps to illustrate what happens in-between the binary of ‘no one saw any of it’ and ‘everyone saw all of it while fully engaged’. But, as an important aside, screen coverage is currently not factored into viewability, which is an obvious flaw, since an advertisement that has 100% of its pixels on screen means much less if the unit is vanishingly small on the page. Equally, if the ad unit takes over the whole page, it may be more likely to be skipped. The dynamics of advertising online are different because of the huge variation in formats, modes of consumption and possible consumer actions. 

Karen Nelson-Field launched her book The Attention Economy at the event, in which she explains her research that demonstrates high attention doubles the incremental rate of low attention exposures. This produces an attention response curve, which allows for new planning and pricing discussions. [This doesn’t measure the cumulative brand impact over time because that longitudinal data doesn’t exist yet but they are working on it.] 

In planning terms, considering attention from the outset is crucial because attention thresholds are very different on different platforms. Some naturally garner higher attention for longer durations and advertisers should play to the strengths of the media, rather than trying to force 30 second ideas into 3 second units. 

The impact of low attention has been well established by Nelson-Field and Robert Heath, but knowing how much you want and need is also key. Complex messages require more time and attention and the creative in high attention contexts need do less work on grabbing attention. How much attention does the communication require to be optimally effective, how much should that cost and how we most effectively use that opportunity are key strategic questions.

The metrics we use to trade create incentives in the market. The opacity of the programmatic media buying ecosystem makes it ripe for, and rife with, fraud, creating marketplace distortions. The cost of an impression therefore doesn’t properly price the increasingly scarce resource of chunks of aggregated human attention in the marketplace. This is the topic of an upcoming book by Tim Hwang, which suggests we are rapidly heading towards a “Subprime Attention Crisis.” 

New measures like quality cost per thousand [qCPM] are beginning to be traded which is important because they highlight that greater impact is worth paying more for but, as Nelson-Field points out, this will only help if q is measured in a way that represents real quality. Otherwise it will overprice opportunities. Additionally, any single intermediate metric becomes subject to Goodhart’s Law, because when a measure becomes a target it ceases to be a useful measure, which is why a broader set of evaluative measures are always necessary. 

Attention is a many-splendored thing, an analogue aspect of the most complex phenomenon known to man, our consciousness. Understanding it, planning for it and measuring it is challenging but crucial since it is the foundation upon which the entire advertising and media industry sits. 

For More See the WARC Guide Planning for Attention