You may think visitors to your company’s Web site are carefully reading every
word on your home page. But they’re not. People glance at a fragment here, a
fragment there, and decide within a matter of seconds whether to stay
Now a remarkable study has measured this hummingbird-like effect that your
site’s visitors have. Web pages that have been changed to exploit the findings
of this study have
sometimes generated two to three times the response rate from visitors, at
little cost to the site owner.
Web experts sometime say, "No one clicks a button they didn’t see." How could a visitor get to
your site and not see a button that’s right there, waiting to be clicked? Read on to
find the answer.
A Landing Page’s Work Is Never Done
The study, entitled "The Landing Page Handbook," is by Marketing Sherpa, a publishing
group with a hard-earned reputation for digging up what works in online
marketing. The 185-page PDF report, complete with numerous color illustrations
(one of which is shown below), sells for $247 USD but is available for a
discounted price of $197 until March 15 (I provide details later).
Using sophisticated "eyetracking" tests and the staff’s own experience
interviewing online marketers, the study provides seven "overarching
lessons" for companies that have a Web presence:
• Most Web pages are scanned, not read.
Unlike printed pages, Web pages tend to be viewed by vistors for only a handful
of seconds. During those seconds, people glance at a few words in the headline
(but not all of the words), look at any images and captions on the page, and
then make a snap decision about whether or not the page is worth looking at any
• Images trump left-to-right reading.
In the English language, there’s a strong tendency to read from left to right
and from top to bottom. Marketing Sherpa found that this tendency goes out the
window, so to speak, when you put images on a Web page. People experience such a
strong pull to look at on-screen images, especially human faces, that pictures overwhelm everything else. A picture placed on the right-hand side of a
Web page can keep your readers’ eyes from ever getting back to words you may have on
the left side.
• Keep important images on the left.
One big lesson of the study, therefore, is this:
If your Web page bears a dominant image, such as a product "beauty shot" or a
photo of a seminar speaker, put it on the left side of the page. This allows
readers, once they’ve examined the photo, to move their eyes either down a bit or to
your right-hand column, as they wish. This is a more natural reading style than
trying to get people to move their eyes to an image on the right, and then "backwards" to text on the
• The upper-left corner is always seen. The study’s eyetracking
findings indicate that the upper-left corner of a Web page is almost always
viewed by visitors. This is why most Web sites intuitively place their corporate logo there. But Marketing Sherpa found that a
large number of
visitors also let their eyes drift to a spot just below the logo. This makes the
tag line under your logo very valuable real estate to explain the benefits of your
• Captions are high-readership. For the same reason as the
upper-left rule just mentioned, the space immediately below any image on a Web
page is extremely valuable. The eyetracking study found that material underneath
images was viewed quite often. Instead of using this space for unrelated phrases,
or nothing, make sure to state the benefits of your product there.
• Hyperlinks capture attention, for good or ill. Any text you
underline or color in blue (like a hyperlink, whether it’s clickable or not)
will get high readership. This effect is so strong that many visitors will read
only the emphasized text before deciding, "This site is/isn’t for me."
That’s not a reason to eliminate hyperlinks from your landing pages, but it’s
definitely a reason to make them relevant to what you’re selling.
• Lose the navigation on landing pages. Unlike the hyperlinks
mentioned above, navigational links or buttons on a landing page usually
distract visitors from the main purpose of the page. You should eliminate
navigation bars, including "Site Map" or "About Us" buttons, on pages where you want a
visitor to take a single action, such as placing a product in a shopping cart or signing up for a
free trial. Sites that don’t expect anything from visitors, but are bringing
traffic to a single page solely to invite them to click around within a site,
might be exceptions to this rule, but probably not.
What does all of the above boil down to? "Don’t let your Web designer dress up
your pages with clip art," said Ann Holland, president of Marketing Sherpa,
in a telephone interview.
Visitors will lock onto your images and buttons, she says, possibly to the exclusion of
everything else on your precious entry page.
Eyetracking As The Future Of Web Design
The company that Marketing Sherpa hired to perform the eyetracking work for its study
is Eyetools Inc., a San Francisco firm that specializes in human interaction
with Web sites.
The “heat map” shown below is an example of an Eyetools analysis that’s
included in the landing-page study. Areas of the tested Web page that are in red
were viewed by 100% of the visitors in the study. Areas in green were
viewed by 50% of visitors, and areas in dark blue were viewed by almost
The tested Web page is an advertisement for a “Webinar,” a seminar on the Web. Notice
that most visitors glanced at only a few words of the boldface headline (the
words shown in the red spots). The seminar speaker’s photo, on the right-hand
side of the page (also in red), was viewed by 100% of the visitors. The majority
of the visitors skipped most of the words on the left-hand side of the page. The
design of this page could be tweaked to attract a higher percentage of visitors
to register for the seminar.
Eyetools has improved the eyetracking process by eliminating the bulky
eyeglasses and other headgear that was associated with TV-ad eyetracking years
ago. Now, sensors atop a computer monitor record each visitor’s eye movements,
which are surprisingly consistent from subject to subject.
Greg Edwards, Eyetools’ CTO, said in a telephone interview that the firm had
previously charged $20,000 to $30,000 for customized eyetracking studies
performed at a customer’s location. Now, Eyetools offers tests of a single page,
using 10 test subjects, for as little as $1,000. The only catch is that the test
must be performed in Eyetools’ lab in San Francisco, so all travel and setup
time is eliminated.
The Marketing Sherpa study cites numerous cases in which improvements of 50% to
200% in the response rate of a landing page have resulted from changes made
after eyetracking. "The biggest improvement," Edwards says, "was a 10-fold
increase in clickthroughs from a home page to the interior pages of content."
Eyetracking seems to be such an advance over previous methods of measuring the
effectiveness of a Web design that I’ll devote next week’s column to ways your
company can take advantage of it.