A few months ago I wrote about how page views were not going to be around much longer as a used metric to determine the effectiveness of a website. Well, apparently Nielsen/Net Ratings is replacing the page view metric with "time spent", which they feel is much more valuable.
Frankly, this metric (IMO) can be tough to get accurate data on. For example, if I go to an article on NYTimes.com and spend 5 minutes reading it, my time spent should be 5 minutes. But how is the website supposed to know I spent 5 minutes if my next action is to go to Gmail in my browser? I never followed up with the NYTimes.com server, so it has no way of knowing if I was there for 3 seconds or 3 hours. The only way it can know is if I return a few minutes later and click on another article. But even then it's not terribly accurate since maybe I was on a page, read it for 20 seconds, went to the bathroom, and then a few minutes later came and clicked on a new page. Was my time spent 20 seconds? 5 minutes? As far as the web server is concerned, it was 5 minutes, but that's not really true, is it?
The nice thing about page views is that it's easy to measure and objective. However, with AJAX, widgets, etc., it has become suspect as an effectiveness measurement. Time spent is a lot more subjective, but if we can find a way to feel confident in the data, is more telling about actual site usage.