How can you avoid attention mistakes during your brand building? A recent post at Neurons showed that almost 50% of ads were placing the brand in the "corner of death." This area is well known to be a place where nobody pays attention. Here, we show how you can use Predict to ensure that your own brand building does not fall prey to this problem.
When you're building a brand, a critical component of any communication is to link to the brand. If people love your ad but miss the brand, your effort (and budget) is wasted! For years, it's been known that certain areas of pictures, images, videos and other visual elements are "dead" on obtaining attention. For example, the bottom-right corner is the least favorable place to put any information: it's simply so rarely attended that in eye-tracking studies it is labeled the "corner of death." So it's astounding that when analyzing several ads, we found that almost 50% of all ads - half of the ads - had the logo in the bottom right corner.
Clearly, something needs to be done: the easiest is to get the word out there: don't show your brand in the bottom-right corner. This type of so-called "translational science" allows us to be inspired by basic science in doing marketing right. Another solution would, of course, be to run eye-tracking studies to ensure that people see where we want them to. Through both exercises, we would learn that the bottom right corner just doesn't cut it!
Predict to the rescue
But there's another solution that is even more handy: running a Predict analysis. To demonstrate the utility of Predict for avoiding the corner of death, we ran a comparison on data from Predict and eye-tracking on the same images. Here, we analyzed the relationship between areas of interest (AOIs) for the bottom right corner, aka the "corner of death." Would the Predict algorithm also show that this area was a no-go for us? The results were staggering:
The relationship between Predict and eye-tracking data for the bottom right corner in ads show that the relationship is close to perfect! Each dot represents the AOI value for Predict (x-axis) and eye-tracking (y-axis) for the same ad. The line represents the regression line where the Predict AOI values predict over 92% of the variation in the eye-tracking AOI data. These results clearly show the power of Predict -- the Predict analysis is closely resembling the actual eye-tracking data, to the extent that you would make the very same conclusions whether you looked at the Predict or eye-tracking data.
The early vetting tool for brand building
By using Predict and eye-tracking alike, the advertisers of these ads can learn that their ads perform sub-optimally because of the position of their brand. But the worth of Predict goes way beyond just to let you ruminate over past advertising mistakes. Instead, think of Predict as your early vetting tool! Run your ad through the platform, run 100 iterations of the same ad through the model, and see which performs best on grabbing brand attention!
Examples of just how closely Predict output are to eye-tracking can be seen in the following images. Amazingly, these are ads that the Predict AI model had never seen before, yet the heat maps are virtually indistinguishable. So what's the recommended action for using Predict?
- Make different versions of your ad, or other visual materials
- Include some "oddball" and crazy ideas -- if nothing else, you learn what to rule out
- Identify the purpose of your ad and use this to define the most important areas of interest (e.g., the brand, product, and specific words)
- Run all images through your Predict platform
- Draw AOIs on each image for your predefined areas
- Compare the scores (if you-re sophisticated, download the data and run a statistical analysis)
- Choose the image(s) that perform the best
- If you still prefer another version than the winning ad, take inspiration from the winning versions to ensure optimal attention to the key AOIs
Test your ads for optimal brand attention with Predict. Book a demo today to see how it works.
.png)
