In addition to the discussion above on the subject of Time on Page (or even Time on Site) I would like to add that there are useful articles to read who question these data because of several reasons (one of them is Tabbed Browsing), quite interesting stuff !

Dear B,

Thanks a lot for the Sherlock Holmes stuff. Now I can finally comment on Time on Page mechanism being used by WA tools. Tim stamp is a cost effective way and more precisely a bandwidth conservative way because with this, WA tools don’t only have to sent data to their servers once i.e. on page load.

I think they should maintain a stream of time stamps for every visit to a particular page in order to calculate Time on Page accurately for bounces and exits.

Dear Karl,

I agree to your point that if exit rate is used to monitor performance of key sections then it proves more useful.

That is why I said, Exit Rate should be able to help you when you’re assessing one page or a group of related pages (section or a directory).

Thanks for the insight

Not sure specifically about GA but I would imagine they work in the same way as other industry leading WA tools.

Omniture:
The Time Spent on Page report shows the amount of time that a visitor spends on each page accessed during a visit. When a page loads in the visitor’s browser, the image request populates the time variable in the SiteCatalyst code with the exact time up to the second. When the visitor leaves the page and accesses another page, the new image request populates the time variable in the SiteCatalyst code on the new page. SiteCatalyst calculates the time between the two pages and populates an instance to the proper time interval in the Time Spent on Page Report.

Definition of Visit Duration from the Wab Analytics Association:

Visit Duration
Type: Count
Calculation:
The length of time in a session. Calculation is typically the timestamp of the last activity in the
session minus the timestamp of the first activity of the session.

While G/A discloses the “sampled data” right at the top of the reports, it does not provide any clue whatsoever what the actually sample rate is.

To be able to even begin to properly evaluate the time on site and or bounce rate effectively, can you confirm what the standard data “sample rate” G/A utilizes for its reporting?

Please confirm.

Thanks.

Using directory filter to set up independent profiles of key sections so can compare average exit rate of each directory and also trend over time.

Average exit rate can in this way reveal a low performing section of a site, or changes that warrant further research into cause.

Though likely more relevant in lead generation sites, page specific exit rate can be useful to identify trouble pages that are beyond the first step of a conversion path. Bounce will not show you this.

You are right – exit rate is only useful when in the context of a single page (or collection of related pages).

Whether the averaged metric is too high or too low depends on the type of site you have – if you expect your site to have a low exit rate on all pages, then your average will be low. However, if you have few pages and one or two of the pages has a really high exit (perhaps supposedly so – maybe these pages have highly insightful answers) then the average will be high – which is ok…

There is a difference between Time on Page and Time on Site. What you have indicated in your comment looks like more of a way of calculating Time on Site. I am yet to confirm the Time Stamp method from GA team.

I am sure Google Analytics has better ways of calculating Time on Page and if not then you can record Interactions and Events with your AJAX page (or any page) by passing a few extra parameters to Google Analytics script on your page. That way you can also be sure of the efficiency and time spent on that page.

In regards to your previous comment, I think it’s important to remember that by definition, the Time Spent metric requires a second page in order to be calculated (difference between two time stamps). So, Time Spent on a landing page with a high bounce rate is not necessarily going to provide insight in to the value/efficiency of the page.

“In case of dynamic / interactive pages written in AJAX or Flash one would have to also consider the Time Spent on Page in order to finally conclude efficiency of that page.”

If visitor stays for more than 10 seconds then its should not be considred a bounce page as he/she might have consumed the content of the page.

Thanks for the comment, I’ll try to come up with more of such informative posts in the time to come.

-A

I personally think exit rate is a pretty meaningless metric anyway. As you mentioned in your post, everybody has to exit at some point however, taking the example of an e-commerce site, some who make a purchase will return to the home page before exiting the site. That increases the exit rate of the home page but doesn’t make it any more of a problem page.

Bounce rate is a far greater measure of a site or a page’s ability to meet expectation.

I agree that a site wide Exit Rate is pointless and also to your point that a high Bounce Rate does not necessarily mean irrelevant landing pages. In case of dynamic / interactive pages written in AJAX or Flash one would have to also consider the Time Spent on Page in order to finally conclude efficiency of that page.

Thanks for such an insightful comment.

-
Abhishek

e.g.

My Site has 4 pages

Page1 60% exit rate
Page2 20% exit rate
Page3 10% exit rate
Page4 90% exit rate

Average exit rate = 45%, although all visits still exited

There’s nothing wrong with it as such, but it is a little pointless because, as you say, you want to monitor exits by individual page, not whole site in order to view potentially problematic pages.

Bounce is best described as the percentage of single access (or page) visits, but it does not neccessarily mean the landing page is irrelevent, especially in a time of AJAX & interactive pages as the user may be doing something on the page of value, even if it is only reading an article & then leaving.

Unfortunatly GA isn’t advanced enough to construct deeper engagement type measures which would show a more colourful picture than just Bounce or Exit rate

Thanks for the effort, but what’s the point you wanted to make?

-
Abhishek

Bounce Rate: the rate of people who left your website from the landing page, that’s mean when someone click on you’re banner or pub and land on your landing page, he immediately leave you site (so no surf) it’s tell you that you’re land page is irrelevant, people expect something else when they click on your ads. so you have to modify your ads or change your landpages then people can find what they looking for.

send me a msg if you want more information about google adwords.

I am glad you think so

Though personaly i never got an opportunity to try Google’s web analytics but found the terminology description and the difference between the two very insightful.

Rgds