It’s fair to say we have built a bit of experience in the few short years we’ve been considering attention as a measure of advertising cut-through. To date, we have run more than 4 million data points through our gaze models (6 models and counting), but there is still much more to learn about consumer attention and how it connects to a sale. And while we are advocates for the attention economy, human attention is hard to explain completely because quite simply, it is intangibly human. So, we continue to collect data that might help us understand human attention, just that little bit more each time.
Let’s consider why there is so much attention on attention in advertising measurement at all. The simple answer is this: there is a genuine and logical need for a comparable cross-platform impression measure. A measure that inherently reflects the vast modal differences across platforms. One that can be scaled and collected passively. One that can be traded upon to replace tired old demographics, which have limited relevance to actual buying behaviour. For these reasons, there is a genuine push from industry towards disrupting the measurement agenda. Add to this, the academic research (including our own) that confirms attention as a conductor for further cognitive and behavioural reactions, including building memory associations and, of course, buying stuff. Evidence helps the cause. We have said this before and will say it again—it is common sense, if you’re not paying any attention you obviously won’t process (or act upon) what is right in front of your face!
In the past few months we have dug deeper into the depths of our 5 frames per second attention data and combined it with some newer data points collected around the creative execution. It is important to note that we anchor our attention with a huge amount of reference data including product choice/sales, mental availability, previous buying behaviour, demographics, ad viewability elements, brand quality, emotions and (soon) search behaviour. All variables are included in our master model to not only answer quite specific research questions, but also to understand how the bigger picture looks. That is, how advertising attracts attention.
To date, most of our work has focused on the attributes of the media platform that foster/impede attention, if you have read any of our work you will know that ad visibility (specifically ad pixels) has consistently surfaced as the media attribute that matters the most. This time we thought we would switch sides. This new work focuses less on what the media buyer can do to improve attention and more on what the creative director can do.
We started with the usual suspects—emotions. No real surprises, everyone defaults to emotions on the premise that highly emotional campaigns are the holy grail. I confess, even in my work in on-sharing of content in 2013 (which resulted in a book) I started with emotions. Needless to say, after two years of work and several data sets later I found that while emotions play a role, they weren’t the greatest contributor of content diffusion (going ‘viral’). The outcome of this agenda showed similar patterns. In our new data, at a granular level, we can see that emotions do trigger attention spikes, but emotions are, in fact, less of a contributor to attention than platform choice. Meaning you could put a highly emotive ad on a platform that has poor visibility and the opportunity to gain attention is significantly reduced.
Secondly, we could see that while attention spikes were often triggered by emotion, not all high attention translates to a sale. To understand this, we transposed our overall aggregated sales measure (STAS) to a second-by-second sales differential. In essence, at high and low attention points we mapped uplift in propensity to buy. What was happening when this uplift was greatest? The answer is simple—branding. The mere presence of branding at attention peaks significantly increases the chance of buying.
So, we started to see an attention hierarchy here that we call Effective Attention. Choose highly visible media to place your highly emotive ads but ensure the high emotion (Ha) points are supported by bold branding. Voila!
((Ad Visibility + Ha Emotion) = Attention Spike) + Quality Branding -> Sales.
Most creative teams would balk at the thought of interrupting high emotive moments with branding, but alas we have that covered too. In my earlier content diffusion work we looked at how overt branding impacts emotional moments. Truth is, it doesn’t. It has no negative effect on how emotive the ad was rated nor how much the ad was on-shared via social.
There is, however, a dangerous dysfunction here. If you only focus on Ad Visibility + Ha Emotions, the ad will be misattributed to competitor. If you only focus on Ad Visibility + Quality Branding, the ad will have limited cut-through. If you only focus on Quality Branding + Ha Emotions, the ad will likely not get seen.
Make every attention opportunity effective, they are rare.