Thxs bryan

This is definitely relatively unknown technique in CRO world. Actually, sometime in the past I had researched a lot on the science behind this predicting gazes on an image. This paper summarizes how it is actually done http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?ar… (though it is technical in nature).

Basically, the gist of the technique is saliency. Any part of the image that is more salient from the background (difference in contrast, color, texture, etc.) tends to stand out and the algorithm detects all such areas.

Though when applied to web it has couple of deficiencies because on web people read in F-pattern and have some other intricacies such as ignoring pretty much everything in sidebar.

A much better alternative to predictive heatmap is a service such as GazeHawk which plots you a real heatmap.

No but I have seen a strong correlation between minimized frustration levels if people can find what to click on right away and not have to go searching for it on a page.

LandsEnd.com is very good at tracking the ROI of their changes. I am sure they would not have kept the revised version if it didn't increase revenue.

Right, forgot about that…where's my tobii eyetracker 😉

No the question is, is there a strong correlation bewteen what people first see and where they click (or not) on ? I can imagen a picture or flashy color object can track the first impression but is not necessary an area that require action (thinking about a product picture for example)..

Dave,

These tools simulate the first 5 seconds of attention to a page your visitors see. Crazyegg aggregates all the mouse movements of all your visitors and represents it as a heatmap. They are two very different things and you should expect different results.