For a long time now, I have wondered whether Google considers how often your site is visited via a direct click on a search engine results page.  For example, if a Google user enters search term X and your site returns on the first page.  The user then elects to click on your link (instead of the other nine on the page).  Does Google consider this in evaluating the future rank of you page?  It seems like a fairly objective measure of the appeal of your site’s descr i ption when presented on a page against nine of its competitors.

I have yet to fully explore this, but I do have some semi-empirical evidence to suggest some correlation. 

About three months ago I started two blogs.  They were both on the same broad topic.  I was the sole writer for each and kept the post rate constant.  SEO work was identical for each.  Each now has about 15 000 words on the same subject matter. 

Until a month ago, they were both returned in the SERPs at about the same position for the same search terms.  However, an article on one blog contained a slightly different keyword variation for the post topic.  This subtle variation in the keyword meant there were more collisions with a popular search phrase and therefore a few more Google clickthroughs (this was reflected in looking at traffic and referral statistics). 

This small perturbation seems to a contributed in a very disproportionate way and the blog in question now ranks far higher on many search terms, sometimes on page one or two of the results for quite general terms, and also a good deal higher in Google Blog Search than its sister blog about the same topics.

I realise it is difficult to definitively establish whether this factor alone was responsible for the sudden upturn in the SERPs, but given all else was kept fairly constant and the very marked nature of the change, it might suggest that clickthroughs are a ranking factor.