SAAS-tool Brainsight predicts viewing behavior and visual performance of (digital) ads and videos
New developments in artificial intelligence show that A.I. models are excellent predictors for viewing behavior. In today’s attention economy, consumers are time-poor and easily distracted, which is why the marketing industry often refers to the so-called ‘3-second rule’, indicating that advertisers need to get brand recall and their message across instantly; quite a challenge for creatives and digital marketers. With new A.I. tools entering the market it is possible to pretest ad-performance and tackle this issue.
Most digital ads are tested post-live, for instance with A/B testing. Researching performance of (digital) out-of-home is even more difficult and sophisticated. By combining neuroscience and A.I. research, visual attention and visual clarity can be predicted in advance.
Predicting viewers’ attention
In the past years, studies* have been conducted by renowned universities, like MIT, to research and map viewing behavior. It appears that instant visual attention is influenced by elements like color, size, contrast, shape, location, etc. With the help of A.I. and machine learning, predictive models have been developed, the latest with accuracy rates of 94% compared to live eye tracking research.
Now that visual attention can be made predictive, new commercial tools enter the (digital) marketing industry to help them with the 3-second rule.
Eye Tracking Heatmaps
Researching visual- and viewing behavior is generally conducted with eye tracking technology, where dozens of people (‘s eyes) are being tracked while they watch images and videos. With heatmaps, the consolidated data of their gaze-points are visualized, showing where people (do not) look, how their attention is distributed and whether they focus on only a few or many parts in the image. These are all indicators of visual attention. Did viewers see your logo? Was your ‘call-to-action’ button overlooked?
“Apart from qualitative research for large budget productions, data-analytics has become the norm for testing ad-performance. But data only comes in after a campaign has been launched. Predictive A.I. models are a cost-efficient way to pretest and optimize designs before starting a campaign,” says Roger van der Spek, Managing Director of Braingineers, an Amsterdam-based neuromarketing research/tech agency.
Predictive A.I. tools
Predictive models have found their way to the market with tools that can instantly create eye tracking heatmaps, without the need for extensive research. Pretesting viewing behavior will therefore become more common and with a fraction of the cost of live-eye tracking research, also smaller marketing agencies and independent professionals can use this method. Roger: “With online tools, pretesting or benchmarking ad content on clarity or attention can easily be done instantly, by one person and from any location.”
Max van Kaathoven, founder and CTO, explains: “In the past years we’ve conducted hundreds of neuromarketing-studies for digital customer journeys, always including eye tracking. By combining the latest A.I. models with that expertise and data, we have been able to develop a front-running tool.”