Do you think you're more likely to look at an online ad if it contains 1) a picture, 2) an animation or 3) just text? The answer: just text. Surprised? Well, consider the man who was checking his e-mail when he came across a dating-service ad featuring a picture of a bikini-clad woman. He looked at the woman's face and chest once and then at the surrounding text five times.
The Internet has cracked open a brave new world for folks whose job it is to spend ad dollars. The ability to track where a Web user clicks provides a sort of precision intelligence advertisers could have only dreamed of in decades past. But before a click comes a look, and according to new research, advertisers are often wrong about what attracts our attention.
The findings are presented in a chapter of a new book, Eyetracking Web Usability, by Jakob Nielsen and Kara Pernice of the consultancy Nielsen Norman Group. Don't let the bland title fool you: what Nielsen and Pernice have done is track the eye movements of hundreds of people as they navigate websites, looking up advice on how to deal with heartburn, shopping for baby presents, picking cell-phone features, learning about Mikhail Baryshnikov. By bouncing infrared beams off a person's retinas and recording head movements with a camera, the researchers were able to deduce what sort of ads garner attention in real time a methodology that runs laps around later asking people to recall what they saw.
The headline result: simpler is better (not to mention probably cheaper to produce). Participants in the study looked at 52% of ads that contained only text, 52% of ads that had images and text separately and 51% of sponsored links on search-engine pages. Ads that got a lot less attention included those that imposed text on top of images (people looked at just 35% of those) and ones that included animation (it might seem movement is attention-grabbing, but only 29% of these ads garnered a look).
Now, looking at an ad and being vaguely aware of it are two different things. Plenty passes through our peripheral vision, but because of the way the eye works, we only thoroughly see things that we stop at and observe deliberately. By that measure, people in the study saw 36% of the ads on the pages they visited not a bad hit rate. The average time a person spent looking at an ad, though, was brief one-third of a second.
Interestingly, people who were just browsing the Web looked at only 5% more ads than those trying to accomplish a specific task. Even when we're on a mission, we're still fairly willing to stop and look at an ad. However, there was one sort of website where ads rarely registered: pages built around search boxes. Think Mapquest or Expedia. Google's tribute to white space on its home page might be sleek design or it might have something to do with knowing that no one would look at an ad there anyway.
Then there was the result that most surprised the researchers: text-only ads received the most looks. Part of that might be our accidentally thinking text-only ads are part of the information we're looking for. But as Nielsen explains it, the nature of the Web itself might be coming into play as well. Unlike television, which is a passive medium, the Web is all about taking action searching, clicking, registering, buying, downloading. It might be the case that as we're out there on the Internet, what we're attracted to is content that gets us to where we want to go.
That's one possible reason the man presented with the dating-service ad quickly moved past the woman's body and fixated on the text surrounding it. "Even in a case like that, the real information is still the strongest point," says Nielsen. Odd as it may sound, the way to grab people's attention online might be to simply level with them.