During the kick-off week for our second term we got the chance to get our hands on a Tobii eye-tracking system. As a part of our assignments we got the task to experiment with the Tobii system, so we decided to test some of the popular assumptions about eye-tracking which you can find all over the web.
One of these assumptions is that you can guide your users’ attention by using images of faces gazing at areas of interest. The theory basically goes as follows: Humans are social beings and naturally drawn to look at faces. They’ll also search for visual clues in these faces and act accordingly, so if they see a face which is not looking at them but a specific area of the website they’ll follow the ganze of the face.
Examining existing studies
However, there has been some back and forth on the validity of this theory and there are a few contradictory studies on the subject. One of the supporting studies was conducted by UX Designer James Breeze, he published his findings on his Blog Usable World.
He claims “Faces can be used to guide a person’s attention to key content and make sure they actually read it.” and provides two pictures showing the results of a 106 subjects strong study. The top picture shows a baby looking at the user, the second a baby crawling and looking directly on the text to its´ right. As you can see, the baby on the bottom picture seems to draw a lot more attention to the text as the one in the first picture, apparently because the test subjects were guided by the baby’s gazes.
A study with contradictory findings was just recently conducted by the team behind EyeQuant, an artificial intelligence software which predicts eye-tracking results. While their main intent was to find statistical patterns for their software, they also shared some insights they found along the way. According to their findings the users’ attention isn’t as often focused on faces and also doesn’t necessarily follow their gazes.
Unfortunately we don’t know how many websites with adequate images they tested, but judging from the screenshot above, there might be some reasons why they couldn’t reproduce Breeze’s results. Fist off, the headline’s font is not the heaviest and the woman seems to have her eyes almost closed, not really looking towards the headline. Secondly, the website is visually heavy towards the middle, drawing attention away from the image. And thirdly, users which are just browsing might be more susceptible than users with a clear goal/task in mind (e.g. buy stuff from eCommerce websites).
Our own A/B-Test
To find out for ourselves whether you can or can not guide your users attention through faces, we decided to do our own small eye-tracking A/B-test. We quickly did two mock-ups of a simple charity site with the help of Macaw, one with a banner in which a child is looking at the viewer and another where a child is gazing at the short text and donate button.
After we set up the eye-tracker and defined areas of interest (AOI) around the text/button area we asked our participants to read the short story on the page. Unfortunately we didn’t have enough time to gather more than eight participants, so you obviously should take the following with a grain of salt.
Below you can see a typical gaze plot for variant A (left) and B (right).
And here are the accumulated gaze plots:
And the heatmap of all users:
As you can see, the picture in version B seems to draw more gazes to the text area and donate button. The mean fixation count in the defined area of interest for variant B was 12, in variant A it was only 5. Overall, we got 20 fixations inside the AOI in version A and 48 in version B. Unfortunately the differences are not significant, but this was to be expected.
What´s next?
Because of our limited number of participants we can’t definitely tell if you can grab the attention of your users with this technique, but the results seem to indicate that you can. That being said, we are planning to do a real study in the near future with more participants and proper methodology to get more definite results, so stay tuned for updates.
Sources
[1], [2] You look where they look by James Breeze
[3] The 3 Most Surprising Insights From a 200 Website Eye-Tracking Study by EyeQuant
The Images in the mock-up were taken from GettyImages.
