Forum Moderators: Robert Charlton & goodroi
New EyeTracking Study verifies the importance of page position and rank in both Organic and PPC search results for visibility and click through in Google.
Nothing really shocking here, but the images which identify the important "triangle" are kind of neat.
[prweb.com...]
I just called Tobii Technologies (Sales office in San Francisco) and was quoted $27,000 for 1 of their eye tracking systems (17" monitor with built in tracking, and software).
I just got an idea for a new site ;)
Interesting study though :)
The only thing it reinforces which I just learned is my skyscraper ads on the left hand side works better than they did on the far right.
And also (shock) people read top to bottom.
Thanks for the great link to the "First Impressions of a Website" page.
If you look at the web sites they review, it doesn't really surprise me what people are looking at and in what order. Basically proving with technology what photographers, designers and artists have known for years about that the eye travels to the most dominant object on the page, then the second most, then flows around. If you construct a good organic flow you can drive customers to look at exactly what you want with very little effort.
Photographers also use the "squint" test - you squint when you look at an image or scene (or a web page) and the most dominant object that will catch someone's eye jumps out at you.
Try it, much cheaper than what Eyetools charges :)
I think that is what most of these type surveys miss. The fancy triangle don't mean squat if none of the items in that zone are useful.
From what we have seen, and from my own personal habits - if I am seriously looking for something I will often go several pages deep. If I am just casually browsing, I tend to follow the same pattern they found (what a surprise).
As you pointed out, window shoppers are not buyers, even thought they may be clickers. Prospective buyers usually tend to dig a lot deeper, especially on expensive purchases.
Another thing I have noticed lately is that we are getting more traffic lately from the places like PriceTool.com that allow customer ratings of merchant sites. I think a lot of people - especially on the internet - tend to shop more for reliable service etc than for just price or placement on the search engine page. I know that when I personally go online shopping, I sort computer stores by rating, not price.
Would be great to get some valid research on a sample of different sites too. Then i could match the demographics of my target market to the research.
I would pay to access this type of research.... anyone know any?
Part of this may be that we have conditioned people to scan this way because of the broad use of left hand menu systems
I think you (and a few other people) are missing the point here. This survey is only relevant to Google's site. It does not apply to other web sites.
The reason that the triangle is there is because Google has put the main content down the left hand side. If they were to flip the design and put the content down the right, the 'triangle' would be inverted.
This 'triangle' will not appear on all sites. On sites where the navigation is on the right hand side of the page, the eye focus will be placed over it.
This reasearch does not suggest that people should start modifying their own pages and putting their important page elements where Google's 'triangle' is.
Also, it shows that a title like "domain.com - cool website this is! Page on Widgets" is by far worse then "Widgets! domain.com - cool website this is!". Since further down the SERPs, the right end of hte title might never be scanned.
SN
Photographers also use the "squint" test - you squint when you look at an image or scene (or a web page) and the most dominant object that will catch someone's eye jumps out at you.
It's always worked for me! :)
agree, not only worthless, but harmful.
they can survey 20 people all they want, it is
a) non-scientific (i.e. done to create a press release, or to deceive people)
b) for Google - makes no account for the price/position. Lower Ads already discount for this "phenomenon".
c) possibly only applies to Google search page - and that is unapplicable to other websites (with a different navigation structure, more images, different content, etc., etc.)
d) also does not take into account "banner blindness" effect.
Wouldn't a survey of 50 webmasters be more revealing (but of course we wouldn't want to tell what works really would we) ;-)
Wouldn't a survey of 50 webmasters be more revealing
Webmasters aren't the visitors.
That would be like asking the warden if all the prisoners enjoyed their stay.
This is akin to a higher-tech version of what Jacob Neilson already has been doing for years, I just think it's been done before, the usability results are in, this just confirms it.
I think the main take away here is that the general searching public rest a lot of their trust in the top results that Google spits out.
“It’s #1 so it must be the best.” they say.
We’re a biased audience because we know how SE’s operate. Moreover, we’re programmed to scan and “hyper read” whereas many people read slowly, and have trouble digesting all of the content that spills out into a single results page.
I did a test and if you take the word count on an average Google results page and dropped it into 12 pt Times single space in MS Word, you’d read the equivalent of 3-4 pages!
Needless to say people don’t have or refuse to invest the time into digesting so much data.
Wow, I’m coming up with a great idea for a new SE to rival Google.. Anyone want to invest? :)
The problem is that they end up measuring not only user behavior (which is what we want to know), but also how good Google's search results are. And they don't make any efforts to separate these two things.
In the study they told the subjects to search for something and then tracked their eye pattern. Well, if you're searching for information and looking at a search result page, you'll look only until you find what you want, right? So if Google does a great job, you'll find what you're looking for right at the top, and won't bother checking out the rest of the page. But if the results aren't very good, you'll keep scanning down the page until you find what you want. So the outcome of this study is influenced more by how good the search results are, than by intrinsic user behavior.
Imagine if somebody did a study on "when people look for lost car keys in their house, how many places do they look in?". Such a study would probably indicate that 60% of subjects look for their car keys only in 3 places, 20% look in up to 6 places, and only 5% look in 10 or more places. Well, that's great, but that only tells us about how messy people's houses are. It doesn't give us any useful information about how people look for car keys.
[poynterextra.org...]
........................................................Don't look at this
LOOK HERE
.............................No here.
Not here.
wait
...............................................This is what I don't care about.
here is where I . . . . . ;)REALLY;). . . . . want you to look.
.
.
.
.
.
.
.
based on this study you should not be scanning this smaller text but actually reading it
[edited by: Reid at 6:14 pm (utc) on April 15, 2005]
