Comments on: Decreased Bounce Rate – A Metric For Success? http://blog.purevisibility.com/2009/01/decreased-bounce-rate-a-metric-for-success/ Fri, 03 Sep 2010 07:36:08 +0000 hourly 1 http://wordpress.org/?v=3.0.1 By: mwilliams http://blog.purevisibility.com/2009/01/decreased-bounce-rate-a-metric-for-success/comment-page-1/#comment-1463 mwilliams Fri, 30 Jan 2009 15:30:55 +0000 http://blog.purevisibility.com/?p=390#comment-1463 I agree, looking at bounce rates are more meaningful when you look at individual keywords. If you lump them together and look at the bounce rate by page you won't be able to determine which words caused your good or bad bounce rate. I agree, looking at bounce rates are more meaningful when you look at individual keywords. If you lump them together and look at the bounce rate by page you won’t be able to determine which words caused your good or bad bounce rate.

]]> By: paisley http://blog.purevisibility.com/2009/01/decreased-bounce-rate-a-metric-for-success/comment-page-1/#comment-1462 paisley Fri, 30 Jan 2009 14:38:09 +0000 http://blog.purevisibility.com/?p=390#comment-1462 Bounce rate works better when you use it on a per-visit basis correlated with keyword. Bounce rate works better when you use it on a per-visit basis correlated with keyword.

]]> By: jhullman http://blog.purevisibility.com/2009/01/decreased-bounce-rate-a-metric-for-success/comment-page-1/#comment-1461 jhullman Fri, 30 Jan 2009 14:19:23 +0000 http://blog.purevisibility.com/?p=390#comment-1461 You are right that this is how Time on Site would show up given the scenario you describe (30 sec on page 1, 0 on page 2) I don't know of any package that will by default get around this, but something we've done to track engagement with Flash includes setting a timer variable that periodically logs time if the page is still open. This way you get some information on how long the user was actually on the page watching the video, even if they leave without further navigation. So you could probably do something similar with javascript variables - initialize a timer when the page loads, and have it periodically (every 30 seconds or so) record a reading, and pass that as a user defined variable to your web analytics package. You are right that this is how Time on Site would show up given the scenario you describe (30 sec on page 1, 0 on page 2)

I don’t know of any package that will by default get around this, but something we’ve done to track engagement with Flash includes setting a timer variable that periodically logs time if the page is still open. This way you get some information on how long the user was actually on the page watching the video, even if they leave without further navigation.

So you could probably do something similar with javascript variables – initialize a timer when the page loads, and have it periodically (every 30 seconds or so) record a reading, and pass that as a user defined variable to your web analytics package.

]]> By: awhinston http://blog.purevisibility.com/2009/01/decreased-bounce-rate-a-metric-for-success/comment-page-1/#comment-1459 awhinston Thu, 29 Jan 2009 21:06:00 +0000 http://blog.purevisibility.com/?p=390#comment-1459 Clouding the picture even more is the way analytics packages record time on page/site. The common approach is to take the timestamp of two subsequent pages and subtract them. Seems straightforward enough, but think about a user who visits one page for 30 seconds, visits a second page for 10 minutes, then closes their browser. The application will report the 'time on site' for that visit as 30 seconds and report 'time on page' for the second page at zero (not 10 minutes). Anyone understand this differently? Anyone know of an analytics package that can report true 'time on page' for all pages? Clouding the picture even more is the way analytics packages record time on page/site. The common approach is to take the timestamp of two subsequent pages and subtract them. Seems straightforward enough, but think about a user who visits one page for 30 seconds, visits a second page for 10 minutes, then closes their browser. The application will report the ‘time on site’ for that visit as 30 seconds and report ‘time on page’ for the second page at zero (not 10 minutes). Anyone understand this differently? Anyone know of an analytics package that can report true ‘time on page’ for all pages?

]]>