While our digital world has been changing rapidly, the metrics we rely on to buy and sell advertising are still based on the traditional opportunity-to-see. If a digital ad has been served and meets the MRC standard of 50% pixels for 2 seconds, then it counts as an impression. But the reality is that on average only 50% of impressions are looked at, even after viewability standards are applied. To make matters worse, the majority of that 50% of ads being looked at are less than 100% in view. Think of that math; this doesn’t leave much hope for being effective.
The moral of the story here is that even with viewability standards applied, buyers and sellers trade on traditional CPMs, that bring us no closer to understanding whether anyone was actually paying attention. And while time in view has been used as a ‘suitable’ proxy for human engagement, we can see in our data that time in view can equally represent distraction.
Most large advertisers know this and have been pushing for change in our current measurement system. It may sound too simple, but the only metric that will tell us if a human has viewed an ad, is human viewing (funny that). Simply knowing whether someone is looking at an ad is a massive improvement on the probability-based ‘opportunity to see’. Attention data, which is independent of platform owners, gives advertisers a way to identify, manage and even call out platform (and creative) differences.
This is what is referred to as an attention economy.
And the monetary value of the waste from inattention is immense. Distraction is worsening with an increased number of ads, ad formats and media types, all with different user experiences.
The value of attention
What does attention measurement offer that recall doesn’t? Does attention lead to sales? Does it mean that we will understand our consumers’ needs, desires and everything else in any text-book funnel? This always leads me to these ‘reality checking’ facts:
Key Fact 1: If there is no human paying attention to an advertisement, then there is no chance the message will get through. Nor can it have any effect on business outcomes; sales, awareness, recall, message take out, mental availability, upper/lower or whatever other funnel you use to determine success. I find it hard to believe that people find this hard to believe. The science part is this: when no attention is paid, ads have no impact.
Key Fact 2: We ‘want’ attention to give us a complete picture of a consumer’s brain, but it can’t. We ‘want’ attention to show us a true causal link with sales, but it can’t. Humans are too human for that. Particularly when we are talking about attention measurement at scale for real-time advertising effectiveness work. At its simplest, knowing whether someone is present and looking at an ad is a vast improvement on the current impression system, where many don’t see the ad but the advertiser still pays.
Key Fact 3: Human attention, while largely unexplainable due to the complex human element, does offer incremental value to advertising measurement.
I have mentioned the value of attention as an improved measure of opportunity-to-see, but what we also know from our own data that visual attention measures implicit and explicit processing. This means that attention captures beyond explicit recall, beyond explicit message take out, beyond explicit intention and it captures basic implicit memory (which is noted to be acquired and used unconsciously, and can affect thoughts and behaviours). Therefore even if you are unaware of or can’t remember the ad exposure, passive attention does play a role in behavioural outcomes. It is not as strong as looking actively at the ad, but it still plays a role.
Key Fact 4: Human attention ≠ dwell-time or time-in-view.
Attentive seconds as a measure represents, on average, how many seconds of human attention were actually paid to advertising. Time on screen, does not. We see in our data that time on screen, a metric that is used as a proxy for attention, in reality tells us little about whether a human has actually seen the ad or not. We find that on average around 50% of time on screen has no attention paid at all – so it can equally represent distraction as it can engagement. So beware, proxy measures of attention/engagement, opaque metrics of time on screen or ad in view are not what they seem and can represent distraction just as easily as they can engagement.
What other value does attention research bring?
It brings lots of juicy facts about how advertisers can optimise their creative and media buying to squeeze every bit of value from its investment. At Amplified Intelligence, knowing that human attention was key in advertising effectiveness, we developed gaze-tracking and machine learning technology that could take a consenting participant through a natural viewing process on their own device right through to virtual shopping. What this technology meant for us is that we could look at attention and a host of other variables to begin to understand (to some degree) what tweaks attention. So far we have generalised results across 5 regions, 6 modes including: Linear TV, BVOD, Instagram, Facebook, Twitter, YouTube.
Key Fact 5: Highly emotional moments in creative gain more attention
Humans tend to wander around in a state of pre-consciousness and prediction, and it takes certain triggers to jumpstart our brains and prime us to take on new information. This is important for creative devices. According to Professor Jared Horvath Cooney, “If you ever want to truly and completely grab an individual’s attention, you must break their prediction.” He says that unexpectedness can break the prediction and trigger attention. We see this in our data. Emotion is an unexpected trigger, but it has to be high emotional arousal content which is long noted to drive behavioural outcomes. Many people think that creative devices, such as babies and animals, are automatic triggers of unexpectedness. But some animals are boring. Creative devices only drive attention when they are connected to high emotional arousal.
Key Fact 6: Branding should be placed at attention peaks. Given our constant state of autopilot-like pre-consciousness, it is more important than ever to make it easy for the viewer to attribute the advertising to the correct brand. Viewers quickly switch between active, passive and non-attention. If the brand is not placed at key attention trigger points the viewer will miss it. We find that the mere presence of branding at attention peaks increases the chance of buying and that sales are amplified when attention peaks and branding are aligned. Our brains are trained to satisfice, and we make many mental shortcuts in our busy day-to-day lives. For advertising this means if the brand is missing, we fill in the blanks. A word of caution: without correct branding, the attention you work so hard for could benefit your larger competitor. Quality branding is one of the most underrated variables of success. It is actually not much more complicated than this.
Key Fact 7: Attention increases with a higher percentage of pixels, and pixels matter to attention and brand choice. Anything below 100% pixels means diminished attention (and sales). We see this consistently across multiple sets of data. The nature of how different platforms are designed relative to the level of ad visibility can greatly affect the amount of attention people pay to them. For example, and as you would expect, the larger the ad in the platform feed the less competing stimuli there is around the ad, the more likely the ad will be noticed.
Key Fact 8: Attention and ad decay is related; the more attention paid, the longer the brand stays in memory. Advertising decay is the rate at which advertising erodes over time. We can see consistent patterns across multiple sets of data that show that ads that gain higher attention at the initial exposure, stay in memory for longer, so the impact of the ad is more enduring. We can see that each attentive second on average leads to 3 days in memory.
Key Fact 9: 2 seconds is not enough. If someone is telling you that 2 seconds is enough, don’t believe them. Our data shows conclusively that there is a difference in advertising impact above and below 2 seconds, and this is consistent across all platforms, even TV. This is not to say that less than 2 seconds of attention renders no effect, because it does, but the difference between brand choice impact above and below 2 seconds is sizable. This shouldn’t be overly surprising though, and there are many studies other than ours that show that time viewed is an important factor in advertising effectiveness.
How can attention be applied to the advertising ecosystem?
The real value of human attention to the advertising ecosystem lies in the fact that it is a human-centred, scalable, cross-platform metric. It can dispense with so many assumptions and record the reality of human viewing. It measures what people do rather than what platforms do (as per Key Fact 5), and this is important. It means that, as a metric, it can retain its usefulness and accuracy in the face of platform and device disruption. It avoids the distraction of device measurements and by focusing on the actual target of advertising, it becomes a resilient, portable and impartial measure. After all, advertising and sales are about people.
Up until now, applying true attention metrics has been challenging. Even if we all agree that human attention is valuable, many are undecided around how as an industry we can apply it. Most of the use cases on attention application use it to improve creative output. But the benefits of attention metrics run far deeper. As an independent metric, attention has the capacity to inform media planning, buying and, ultimately, how the entire advertising ecosystem values the space upon which it trades.
In talking to our customers about media planning, advertisers asked us to help them understand whether the cost they are paying per 1000 views is worth it relative to the level of human attention gained. It’s a simple value for money, relative to impression quality, question. Over the past 6 months we have been building our own tool to test the thesis that attention data can be applied to planning and can be used as a second quality parameter to reach.
At its core, our tool (called attentionTRACE™ Planner) can calculate the monetary value of an ad impression based on two important ad effectiveness factors – human attention and time viewed. In other words, our tool describes whether performance differences relative to human attention gained between platforms and ad formats is accounted for by cost. The output is an Attentive Seconds Cost per Thousand (asCPM), a metric that can be directly ingested into planning systems and used for mix planning and CPM negotiations. Importantly, our tool uses real human attention – no assumptions or proxies.
Why is this so important to the industry? Because our measurement system is broken and human attention data, collected independently, can bring transparency back. Because we know that around 50% of media spend receives no attention, yet advertisers are still paying for it. Because on average, every attentive second leads to the brand being remembered for 3 days. Because at the very least we know this data takes media planners away from the race to the cheapest ad spot and brings effectiveness to the forefront.
Key Fact 10: Not all reach is equal. Attention adds a quality layer that other non-human data simply can’t. When quality parameters are considered, mindsets change, investment is re-calibrated and effectiveness emerges.
About Amplified Intelligence
Based in Adelaide, Australia and with global reach, Amplified Intelligence believes that media trading should be fair and accountable. We are driven by the knowledge that human attention is highly valuable, and we build superior and rigorous human attention research products to measure real human attention on mobile and TV. Amplified uses machine learning together with innovative human minds to amplify what is possible; digging deeper and pushing further in their understanding of the value of human attention in the advertising ecosystem. Using (quality) attention as an easily accessible supplementary layer of measurement, is how impression reform begins.