I'm not so sure this is an urban legend any more.

I maintain a website for a business, and we seem have done quite well with all of the major search engines. It's part of my job to keep an eye on the competition as well, and I've noticed that some sites, with poor (by any standard) meta information and incoming links from non-relevant sites rank quite well for popular keywords.

I've heard it whispered that the popularity of a website (as measured by the number of clickthrough visits it receives via the search engine results) is a factor in how Google ranks a site for that particular keyword. More traffic, higher ranking.

I'd appreciate hearing anyone's thoughts on this.

RKeefe

with the recent release of Google trends im leaning more towards this being true.

Heres why:
One of my clients which has multiple (diff content, diff design, diff c ip address) websites that offers similar products recently saw its .ca version of its website jump from the 100s to #6 with little to no optimization.

About a week ago after noticing this, he informed me that his biggest client base is in Canada. A look on G trends for his top keyword shows that 6 out of the top 10 cities that search for his keyword are in Canada. A look at alexa traffic for his .ca site and an analysis of his logs shows a spike in traffic to the .ca site over the past 6 weeks.

I dont know if traffic is a ranking factor or not, but if it's not that's one crazy coincidence.

Google et al have had this data for quite sometime now and do make use of it. It's only minor, but it would make sense in relation to sandboxing of new sites.