Most Web users don't read; they skim
A woman reads her messages online in this 2011 file photo. (AP Photo/Kathleen Malone-Van Dyke)
Michael Oliveira, The Canadian Press
Published Wednesday, March 27, 2013 1:35PM EDT
TORONTO -- Ian Everdell knows exactly where people's eyes generally go when browsing a web page, and how women act a little differently than men when it comes to reading online.
If you can even call it reading.
Everdell has it on good authority that Internet users typically aren't really reading at all on most web pages. Their eyes impatiently dart across the screen in search of something specific, or something interesting to pique their attention, but there's not much actual reading going on.
He knows because of the eye-tracking technology at the Toronto office of the marketing company Mediative, where web users sit down to do a little surfing and have their flits of vision recorded and studied. Web designers and advertisers can then see what viewers are looking at and what they're not paying attention to.
As it turns out, not too much has changed in the years that Mediative has been watching web surfers.
"The biggest and most interesting thing we know about eye tracking from how we read is that we don't actually read. We spend probably about 80 per cent of our time just scanning through web pages," said Everdell, the company's manager of user experience and research.
"At the beginning of a headline, or a heading, or paragraph, people are seeing at most 12 to 15 characters -- you're not getting a whole sentence, you're not getting six words, you're getting two or three.
"That means that the content of the writing has to be front-ended, because something like 'Welcome to the ...' is all they're going to see, so doing that is not going to catch their attention and get them to read further."
Men and women generally surf similarly, but there are some differences in how the sexes scan for information, Everdell said.
"We ultimately end up looking for the same pieces of information, but if you look at the order that we look at things in, men are very sort of all over the place. We'll pull things out from here and there. There's no real structure or method to the way we scan a page," he said.
"Women are more methodical. They're very sort of item by item. We end up looking at the same things, but the way we look at them is different."
At Mediative's office, test subjects are parked in front of a computer monitor that has two built-in infrared cameras which track their eye movements and record audio and video of the web-surfing session.
"Our eyes jump around three to five times a second and when they're still we call that a fixation: that's when we're actually gathering visual information," explained Everdell.
"We're very much hunting and pecking for that information that's relevant to our intent."
At the end of the session, graphs illustrate where the most fixations took place and for how long. A heat map highlights the most-seen areas in red, followed by hues of yellow and green.
Studying users' behaviours on search engines revealed a pattern that became known as the Golden Triangle, a hotspot of red and yellow in the upper left corner of a search results page. The consistent pattern suggests users mostly focus on the first few search results down the page and only about a third of the way horizontally across the screen.
Sometimes the red and yellow hotspots appear more like a capital F, but one thing's for sure: the right side of the page, where ads are typically laid out, does not get much attention.
"The general pattern hasn't changed (over the years). You see people might glance quickly at the paid ads (but) then jump over them very quickly," Everdell said.
"The scanning through the page really goes down that left side."
Another type of graph, called a gaze plot, draws a chronological line between the different user fixations across a page, typically creating a messy web of zigzagging lines that jump across the screen.
Everdell said the eye-tracking technology is important for researchers because surveys don't always tell the full story of what users were looking at and why.
"We're so bad at self-reporting why we do something. This sort of gives us a window into the brain," he said.
"We can understand what's attracting visual attention and what isn't attracting visual attention, which arguably is often more important."
These days, many eye-tracking experiments are being done on mobile devices, which are exposing a new set of user habits.
"One of the interesting behaviours we see on the mobile device, which is somewhat different than what we see on a desktop, is that people will actually focus their eyes in one spot and move the results with their finger, instead of moving their eyes through the results," Everdell said.
"There's a lot more keeping the eyes in one spot and reading as stuff flies by."