Great in-depth post about the results of research into bounce rate as a quality metric for SEO and why it probably can’t be used a factor (too much signal to noise and potential explanations that might not reflect quality etc). Agreed in principle but I think it would be a foolhardy marketer who can’t see a position for usage data overall to be a factor in the rankings. I left a comment to that effect and don’t have time to expand on it fully right now, but it’s a theme I want to come back to as my gut feeling is that it will be massively important as Google try to co-opt the internet through a combination of search dominance, widespread Analytics take-up, Webmaster Tools and search tracking history. In fact, a system based on that kind of data should, in theory, knock the whole Algorithm/pagerank/paid links/ shit out cold.
Interesting enough… but for another day.