Before I start, I was hoping to bring you some figures on thenumber of users who own a scroll-mouse, or at least some recentsales statistics. While I haven't yet found that data, I can saythat I did find that the scroll-mouse (especially optical) areamong the most popular because of the ease it offers inscrolling documents.
With that in mind, the "Adsense Scroll Mouse Theory" goessomething like this:
People scroll a webpage in mainly two different ways: A) Theyare browsing quickly and looking for something to catch theireye, or B) They are reading a document where their scrolling ismuch slower and paced. When a user scrolls the page, there arenatural breaks which occur due to limited finger movement. Basedon the two kinds of scrolling, finger movements and the factthat eye-level content gets the most visibility, an estimate canbe made (in pixels) which determines where the majority of thesebreaks occur at eye-level.
People scroll a webpage in mainly two different ways: A) Theyare browsing quickly and looking for something to catch theireye, or B) They are reading a document where their scrolling ismuch slower and paced. When a user scrolls the page, there arenatural breaks which occur due to limited finger movement. Basedon the two kinds of scrolling, finger movements and the factthat eye-level content gets the most visibility, an estimate canbe made (in pixels) which determines where the majority of thesebreaks occur at eye-level.
In this experiment, I recorded results from 10 differentscroll-mouse users to measure their scrolls in pixels, based ontwo different scrolling habits.
A) Fast browsing - the average mouse scroll is 600-700 pixelsvertically, meaning when the user pauses, their eye-level fallswithin the 600-700 pixels range on the page (from the top of thedocument, not the browser). They scroll and stop, scroll andstop. They have no choice.
B) Slow browsing - the average slow mouse scroll is 400-460pixels vertically (scrolling in smaller chunks while reading),meaning that they pause at eye level every 400-460 pixelsvertically. As they read a section, they usually start to scrollbringing more content up into their immediate eye-level.
Considering these two key mouse scrolling behaviours, it wouldmake sense to place ads exactly at these points, or use anoverall average to concentrate ads within every 400-700 verticalpixels on the page. The best
fit would be a 336x280 ad block asit catches both ends of the slow & fast scroll.
fit would be a 336x280 ad block asit catches both ends of the slow & fast scroll.
You might be wondering about people using a browser's scrollbarinstead of a scroll-mouse. With a browser's scrollbar, fastscrolling habits either stopped at around the 2000 pixel mark(for very long pages equal to just slightly more than half-waydown a page) or the bottom of the page. For slow scrollinghabits, the pauses occurred at eye-level about every 400-460pixels, just as with a scroll-mouse.
The consistency in slow mouse & slow browser scrolling habitswould leave you to assume that Adsense ads placed around the400-460 mark might be something worth testing on your ownwebsite.
But consider this too: Initial eye-level on a webpage issomewhere around the 200-231 pixel mark. You could try placingads at this point, the 400-460 mark, and then at the 600-700mark. This way you are getting immediate visibility with the topads first presented when the page loads, caching slow readers asthey scroll, and catching fast scrollers too. The only groupyou'd have a problem slowing down are the ones using browserscrollbars. Images used next to Adsense ads can help catch theirattention, especially if the page has a lot of text.
Now it's very likely someone else has already thought of thisand it's already been discussed to death. I haven't foundanything related to it. Everything here is really just theory,but something I plan to test myself using live test subjects (itsounds cooler than 'search users') and will post results as theycome.
***Literature :
The Adsense Scroll Mouse Theory
By Carole Nickerson
No comments:
Post a Comment