Suspicious StumbleUpon Bounce Rates?
3 Jan 2008 edit
I've been promoting some of my blog posts by submitting them to StumbleUpon, and it's been generating a fair bit of traffic -- my post "How I Learned CSS" has done especially well. I've been watching my site using Google Analytics, and I've noticed that visitors from StumbleUpon have a substantially lower bounce rate than, say, visitors from dzone. So I think, "yay, StumbleUpon" and start submitting anything interesting I write to there.
But in poking around in my Google Analytics a bit more, I've also noticed that, on pages that I've promoted through SU, I'm seeing strange navigation patterns: for "How I Learned CSS," 40% of visits to the page came from the page itself (and, as it logically follows, 40% of next pages were the page itself). The page was viewed a total of 4,700 times, but a little more than 2,800 of those pageviews were unique. Since bounce rate basically measures whether a new arrival went to another page before leaving the site, I'm not sure it shouldn't count as a bounce if the second page the user visits is the same as the one they were already on.
I'm not seeing this pattern with pages I don't promote on SU, which makes me curious. SU is clearly driving a ton of traffic to my site; it's just not clear that the traffic is actually resulting in the remarkably low bounce rate that Google Analytics shows. If I took out the 40% of people whose next page was the page they were already on, I'm left with not even 1% of people who went to another, different page on the site before leaving -- and suddenly dzone is looking good again.
I'm curious whether anyone else is seeing this, and whether it's actually related to StumbleUpon, Google Analytics, or something else. If I had to guess, I'd imagine that this is a result of people using the StumbleUpon toolbar (since it's not happening in all cases), and maybe the toolbar is pre-visiting the page on the user's behalf to make sure it still exists?
Update: See the comments and a more in-depth discussion.