« MapQuest Embeddable Sharing Map; Launches Voice-Guided iPhone App | Main | Bipartisan Group Calls for FCC Probe into Google Voice Restricted Calls »
October 8, 2009
Eye Tracking Without The Eyes - AttentionWizard.com Beta

Eye tracking studies can be a very valuable tool that can help identify significant problems with your website or landing page. Unfortunately eye tracking requires expensive and specialized hardware and software to be used, and live test subjects to observe and measure.
In-page Web analytics can also provide detailed heatmaps of people's clicking and scrolling behavior. But these also require the landing page to be properly tagged and measure the behavior of real site visitors. This data takes time to collect and can only be gathered from "live" pages.
Recent advances in the study of computational attention and human visual processing now offer a new and exciting alternative. Computer algorithms can be used to simulate where people will look during the first few seconds of interacting with your site and create a detailed attention heatmap of your landing page.
This approach has several advantages:
- Instant results - upload an image of the landing page and get your answer
- High degree of accuracy - 75%+ correlation with eye tracking and mouse tracking
- Works with page mock-ups - can be used with in-progress mock-ups before the page goes live
SiteTuners.com has recently announced the launch of a FREE private beta program (limited to the first 5000 sign-ups) for it new AttentionWizard.com visual attention simulation tool. Free daily heatmaps will be available for all participants for the length of the beta program. A paid version with per-image based pricing is expected by Q1 2010 for individuals, companies, and interactive agencies.
Posted by Tim Ash on October 8, 2009 2:01 PM
Comments
Ric P. October 8, 2009 3:12 PM
Just curious... with the new guidelines laid out by the FTC, you are required to disclose whether you were paid or given anything for free for promoting this company and their services.
So... were you?
Roe October 8, 2009 3:36 PM
Hi,
Just guessing...but the links are not affiliate links...they go directly to the site(s).
Seems like just... helpful news to SEW readers and visitors. I would say that no they are not.
IMHO
Peace,
Dan
___
Sokule
Sokule October 8, 2009 3:51 PM
I have tried to sign up. but I don't have a priority access code. Does anyone know what it is?
carey October 8, 2009 4:04 PM
Hi Roe,
I basically think the new FTC disclosure guidelines are an ill-though-out load of crap that are an unconstitutional restriction of free speech rights.
Regardless, the guidelines apply to blogger "endorsements" and "testimonials". Since I am not endorsing or saying "AttentionWizard is great", they do not apply. I am simply reporting about a news announcement that should be of interest to people in online marketing.
Be that as it may, SiteTuners and AttentionWizard are my companies (I am the CEO of SiteTuners as it clearly states in my profile: http://searchenginewatch.com/3629087 )
So there you have it - I hope that is enough disclosure for one day :-)
Tim Ash
October 8, 2009 6:08 PM
Ric P.,
Both are based on theories of computational attention.
Tim Ash
October 8, 2009 6:11 PM
Carey,
You do not need a priority access code. The review times are currently short - with most people getting an answer in 48 hours.
Tim Ash
October 8, 2009 6:18 PM
Pretty interesting i must say. In my opinion, this will be pretty much for the big companies as small SMB wont have that kind of traffic to justify this investment
SEO Singapore October 8, 2009 10:49 PM
SEO Singapore - I am not sure I understand your comment. There is no "investment" since anyone can access it for free. Also, it has nothing to with "traffic" - you just upload a screenshot of your landing page and get the heatmap back instantly.
Tim Ash
October 8, 2009 11:36 PM
Looks like a nice tool Tim, thanks for sharing. Also seems that you can't please all the people all the time, even with a nice free tool.
Adi @ The Management Blog October 9, 2009 4:17 AM
Adi,
Looking forward to your feedback on the tool.
Tim Ash
October 9, 2009 12:58 PM
Tim,
Thanks for this tool. I was fairly happy with the first screen capture I sent through.
Do you have plans to upgrade the algorithm over time or will this be static for awhile?
Jeremy Brown October 11, 2009 8:44 PM
Hi Jeremy,
Enhancements to the algorithm are being worked on and are rolled out periodically to the public.
Tim Ash
October 13, 2009 1:00 AM
The claim of 75% accuracy - can we see the evidence?
Hoover October 14, 2009 4:25 AM
Hi Hoover,
We will be publishing much more detiled supporting info after the private Beta is completed. We also welcome your experiences and examples. Please send us your original screenshot, AttentionWizard heatmap, and any eye or mouse tracking images. We plan to create a "user gallery" of these.
Tim Ash
October 14, 2009 5:02 PM
I uploaded the screen shot, but did not get any results.
On a side note, I watched your video on 7 deadly sins of landing pages and was really blown away. I have only been using Google for about 18 months, and the information you gave was just awesome.
James October 26, 2009 5:06 PM
Tim,
Love your podcasts and currently salivating over the AttentionWizrd tool. Having some problems uploading a file at the moment (FireFox and IE). Worked for me a couple weeks ago but now it fails to load the file, resulting in an error message. I checked naming, size (KB and dimension) restrictions and they are fine. Is your server have loading issues?
Thanks,
Greg Bogdan
www.twitter.com/GregBogdan
Greg Bogdan October 30, 2009 12:22 PM
Tim,
Does the algorithm take into account white space? What I mean is if I have the same call to action in exactly the same space but in one I have more white space around it, will there be more of a hot spot? Also, if I have less information on a page can I expect a larger hot spot around then if I had the same call to action in the same place but with more text on the page (more distractions for the eye)?
Both these make sense to me but not sure if your algorithm takes this into account.
Finally, I am wondering if you have tested your algorithm with a number of live subjects and how the results compare.
This is a fantastic tool if it is truly accurate. Repeated tests and comparisons between multiple live subjects and this tool would be a nice proof point.
Thanks,
Greg Bogdan
www.twitter.com/GregBogdan
Greg Bogdan November 4, 2009 4:20 PM
Hi Greg,
The algorithm does take white space into account because it looks for changes in color and contrast as key features. The placement of elements with more visual buffers around them will change the results on the AttentionWizard heatmap.
The algorithm is about 75% correlated with eye-tracking and mouse-tracking approaches. It is not meant to be a replace those technologies. But is is terrific for evaluating the visual balance of your page instantly. In addition, it can be used on mock-ups of in-progress designs and does not require the page to be live.
Tim Ash
November 5, 2009 5:12 PM












Tim,
What is your sense on how this works compared to Feng-Gui?