3 ms·
A time-based bounce rate measurement would largely minimize all the downsides listed in the article. (That plus some sort of event tracking for outside links.)
by isleyaardvark 17y ago
A time-based bounce rate measurement would largely minimize all the downsides listed in the article. (That plus some sort of event tracking for outside links.) It's unfortunate that's not the standard, default measurement for bounce rates.
- briancray 17y agoBy "time-based" do you mean how long they spent on the page?
- isleyaardvark 17y agoExactly. Some analytics programs go by whether they viewed a single page, some go by visits that viewed page(s) and only stuck around for less than 3 to 5 seconds. Most, if not all, analytics programs measure "time on site", so it's theoretically possible to segment out the groups that only viewed a single page and spent only a few seconds on the site, it's just inconvenient to set up if that's not the default. In Kaushik's book, Web Analytics an Hour a Day, he defines bounce rate as "The percent of traffic that stayed on your website for fewer than 10 seconds." So that's probably why he's so keen on bounce rate, using his definition you avoid some of the issues listed in that article.
- briancray 17y agoSure - I've read that - but even if you record >10 second visits, you're still not finding out if users did what you wanted them to do based on bounce rate, unless your metric is time on site. I don't think the 10 second rule minimizes all other points in the article.
- niels_olson 17y agowhen we made tmedweb (http://tmedweb.tulane.edu http://tmedweb.tulane.edu), one of our initial goals was a high bounce rate and minimum time on site. We were simply tying a calendar held elsewhere to course material held elsewhere. The basic design was our interest. You know you're doing it right when your biggest detractors set you as their homepage. :-)