Thursday, August 10, 2006

Using the AOL Search data

Okay, enough of the voyeuristic freak-show stuff. The AOL data was released for research, right? So what does it tell us?

Several people have analyzed the data to see what position gets most of the clicks. Not surprisingly -- at least it shouldn't be surprising, although some SEO types had been overthinking this -- the number one position gets more than 42% of the total click-throughs. After that it's a pretty sharp drop down to the last position on the first page (10). Position ten actually does a little better than nine, apparently because searchers suddenly see the second page coming and in a panic will click on anything rather than continue to a second page. Anyway the tenth position gets almost 3% of clicks and then it falls to two-thirds of a percent for position eleven.

I guess that tells you that if you want traffic, beyond those who are really determined to find exactly you, you'd better be willing to scrap and claw your way to page one of the SERPs or else forget about it. And that's gotten harder thanks to the value search engines are now placing on domain age (although one might buy existing domains). Probably better in most cases would be to find some niche that isn't well filled. How to do that? One way would be to analyze the data for searches that get no click-throughs, on the assumption that failure to click through means the search results weren't satisfactory. The ambitious webmaster could then construct sites designed to respond to those searches.

The problem with that is that most searches are very ill-formed, and most searchers appear to have short attention spans. So there is a large random factor in all this. From what I can see there isn't a glaring difference between searches that get click-throughs and those that don't. I assume that this is an area that will be explored in coming weeks.

No comments: