Average Time on Page – how is it calculated?

This totally reminds me of the Abbott and Costello “Who’s on First?” routine.

Google Analytics (GA) doesn’t count the exit page in average session duration.  It can’t because it doesn’t know by default how long users are on the last page they view.  GA measures time on page 1 by looking at the time a user starts viewing page 1, and comparing it to the time the user starts viewing page 2.  If there is no page 2 viewed (i.e., if page 1 is the exit page), it can’t determine time on page 1.

On the other hand, if a page is not the last page viewed in a session (i.e., it is not the exit page), the time viewing that page can be calculated and is accurate.  For a session where a page is the exit page – there is no calculated time on that page, and Google does not include instances where a page is an exit page in the average time on page calculation for that page.

So – if a page is often an exit page, the average time on page metric for that page has less meaning.  Here’s an example.

A Practical Example

You have a weekly blog article you publish on your website.   You send out email to subscribers with a link to that blog article each week.  Chances are, even if users spend several minutes reading your blog article, that same blog article page will often be your most frequent exit page (that’s not a bad thing – it’s just natural user behavior in this scenario).  And, that blog article page will have an average time on page calculation that only considers instances where it was not an exit page.

As such, where a relatively few users that spend a bunch of time reading that page then move on to view another page (whereas most users exit from that page) – the average time on page for that blog page will be inordinately high (e.g., 10 times the average session duration).  But, as noted, that metric will not have much meaning for this page.

And now you know.  I hope.  But, do you know “Who’s on First?”

Why is my Google Analytics Average Time on Page so high?