EyeQuant, a German artificial intelligence software company that offers predictive eye-tracking for websites, is in the business of figuring out exactly what people will look at on a site.

Results from the company's most recent study suggests that some of the conventional wisdom in web design may not always be correct.

The 46-person study was designed to help EyeQuant improve its internal algorithms that predict where people look on their clients’ sites. The study subjects looked at a total of 200 product and landing pages.

The resulting heatmaps show where people fixed their gaze the most.

The average “fixation” lasted for around 15 seconds.

This study found that people were more likely to look at website’s headline text or a search bar than the faces prominently displayed on the page.

Participants didn’t look at huge typography, even when it was advertising a sale. Instead, they looked at relatively smaller fonts on the page.

Users who already knew what they wanted to buy tended not to look at the branding elements of a page, like the company logo.

On sites that advertised a free product or free shipping, people were more likely to look at the product descriptions.

“Data shows us the common sense, rules of thumb about how things work… should be taken with a grain of salt," EyeQuant CEO and co-founder Fabian Stelzer told Co.Design.

Co.Design

Eyetracking Study Reveals What People Actually Look At When Shopping Online

Not just boobs! (But yes, definitely boobs.)

When you pull up a website, what's the first thing you look at? The stock photo model splashed across the page? The large company logo at the top? The search bar?

EyeQuant, a German artificial intelligence software company that offers predictive eye-tracking for websites, is in the business of figuring out exactly what people will look at when the page loads. Results from the company's most recent study suggests that some of the conventional wisdom in web design, like that our attention is drawn to faces, might not always be correct.

The 46-person study was designed to help EyeQuant improve its internal algorithms that predict where people look on their clients’ sites. The study subjects, who came into a neuroscience lab at the University of Osnabrueck in Germany, looked at a total of 200 product and landing pages with the intent to buy things like new Levi's jeans, a TV from Target or a weekend trip to Paris. The resulting heatmaps show where people fixed their gaze the most. The average "fixation" lasted around 15 seconds. What they found:

Faces Aren’t That Powerful

Humans love to look at faces, even in inanimate objects that don’t have them. Designers tend to direct users' attention to important text by placing the text in the line of sight of the person in the photograph, with the thought that you’ll want to look at whatever it is that person is looking at. This study found that people were more likely to look at a website’s headline text or a search bar than the faces prominently displayed on the page.

Big Text Is Easy To Read

Participants didn’t look at huge typography, even when it was advertising a sale. Instead, they looked at relatively smaller fonts on the page. When looking at English Proofread, a copyediting service, for example, users ignored the company name taking up most of the left side of the page in favor of slightly smaller subheadlines. (They were still larger than the main body text.)

Free Is The Magic Word

Everyone loves free stuff, right? They may not look at the word free as much as we think, though. On sites that advertised a free product or free shipping, people were more likely to look at the product descriptions, or in one case, definitely the Victoria’s Secret model’s boobs.

Your Awesome Branding May Not Be Appreciated

Users who already knew what they wanted to buy tended not to look at the branding elements of a page, like the company logo. "Users ignore a lot of the branding on these landing pages," EyeQuant co-founder and CEO Fabian Stelzer tells Co.Design. "They really want to find the elements that get their job done."

Now, 46 people isn’t a huge sample, statistically speaking, but many previous studies of eye-tracking have explored how people look at websites in the context of finding information—like figuring out what the site is all about, or reading an article. This hints that when people shop for something specific, they may see a little differently.

"Human creativity is considered to be this holy space where data can’t help us, but I think that’s not true," Stelzer says. "Data shows us the common sense, rules of thumb about how things work… and should be taken with a grain of salt."

Add New Comment

15 Comments

  • I think this article is misleading and EyeQuant's artificial intelligence software is very 'artificial.'

    1 - Some of the pages are home pages, some are so the customer journey may be 3 to 4 clicks in and they already know what to ignore. Taking these images as a whole to deduce anything about font, faces... is not valid.

    2 - If the tester already knows the brand, they will not invest a lot of time in the branding

    3 - Faces are not important. There is no data I have seen that validates that comment. Why would anyone spend a lot of time looking at a face. They are moving to what they need to see.

  • Martin West

    I feel this dumbs down the role of the brain. Our brains 'see' more than our eyes, it's aware of things its comfortable with, like faces and logos because it expects to see them. The eyes are hovering over information that is unfamiliar and needs processing. So to give the impression that people, faces, branding are somehow less impactful than we assume, then I think it's a rather narrow view that excludes a much richer world.

  • Agreed!

    The part about site branding: the design and colors of a company's website are something I see in a flash. I don't have to study them, but I know they're there, because that's how I know where I am. When a familiar website is redesigned, I always have this period of disorientation.

    With those blocks of huge text, it's like, "Why is this thing in my way? This isn't what I want to know. Where's the menu hiding?" I have this urge to swat it away like a bug.

  • Christine Mark

    My thoughts exactly!

    This study shows where people "fixed their gaze" ... where they needed to study and process information. It is interesting in its own right, for sure. But one cannot conclude that branding elements, human faces, etc. are not important or didn't matter.

  • Where are the facts about devices used to conduct the study? I think we can all assume it's desktop but they don't explicitly mention that.

    Would love to see a mix of devices thrown in, and see if there were differences in how they consumed information based on the screen size and the way their bodies changed to look at the content.

  • That was vague - what i meant by "the way their bodies changed to look at content" is ergonomically users might be approaching devices differently, and potentially in different areas. They might be at a desk for a desktop, and they are probably sitting looking straight ahead - I'm assuming they are using a mouse but i wouldn't know for sure. Laptops they might be somewhere less rigid. On cell phones they might be standing up and more alert, at the very least they will be doing something different with how they hold the content. Tablets they might be reclined on a couch and more relaxed - regardless they'll be interacting with the content directly via touch unless they are on a Win8.

  • Heat maps for eye tracking is very useful but it can be very easy to ignore years of experience from a human being who understands their craft (designers, marketers), in favour of a graph that seems like its revealing something new.

    I would argue that the faces help make the sites feel better/safer for the user, just because they are not looking at the faces first or for a long period of time doesnt mean they are not aware of them or that they are not having an influence.

    This is also very quantitate information. We still need graphic designers as well as our own brains in interpret this information.

    Users look at these pages BECAUSE of the branding. Once they are at the site in question, they are already primed, they want to buy so I'm not surprised they are not looking at the graphics, that happened somewhere else.

  • After thinking "Ugh! More big boob marketing on the horizon!" - I had the same thought as Brad about the software and peripheral vision. It would be a mistake to discount the influence of non/low target areas because peripheral vision affects the final effect. But this study is good news for copywriters and content writers. :)

  • I'm concerned by the veiled criticism on the importance of branding. This reads as ‘users don't tend to care, they're more interested in content’ which is specifically what good branding does well. It instills a relationship with its audience which means they overlook branding because trust is already established.

    I am also referring to branding as every experience between vendor and consumer, not purely a logo.

  • I agree. Theres a bit of an undertow of 'well advertising doesnt work on me' in the tone of this article, although it brings up some interesting points. The branding is playing a huge part in influencing people long before they go to a website. (Seems a bit first-year-of-university level to be having to mention on a site like this).

  • What about the sequence in which the eye fixates on different elements on the page? That would seem more useful than how long a fixation lasts, since most of us would naturally need more time reading text, than looking at a face; or trying to decipher smaller fonts than huge headlines.

  • True but users are giving attention to things that they feel require it.

    It does take some context. They might not understand what a visual element/ button/ field does and look at it longer. They might not care to read text they deem unnecessary to getting the job done.

  • Brad Zavakos

    What I'd be interested in exploring is the size of type conversation. Couldn't you argue with larger type, coupled with short character lengths, lowers the time and attention required from someone viewing it? I myself could see the Spring Fever offer in the above image through periphery alone, which I doubt the software could accurately detect.

  • This was my first impression on that point also — users won't linger on content they've already consumed unless they fail to understand its meaning, which is unlikely with headlines.

  • What this study doesn't capture is whether the visuals are an important impression on the user experience. I would contend that even if people don't look at them, they are extremely impactful to the emotional reaction someone takes away from the business.

    The last quote about data is spot on.