re [webanalytics] PPC Cost (Clicks) vs. Web Site Visits
- Hi Rob,
We looked at a cross section of our client index; which included many
different websites and industry sectors;
It turns out quite a few people who click on a Google PPC campaigns
(between 10% - 30%) immediately press "back button" and don't get the
Of course this is not 100% accurate for all and sundry however it is a
pretty good generic indicator.
The page only partially loads (no images or style-sheets have time to
"visit". Typically we recommend putting the code at the bottom of the
page, since the whole page must load before the tag measures the page
view. Since these clicks don't result in a real page view, I'm happy that
WebtraffIQ accurately recorded this as a no-visit by not recording it.
This is odd user behaviour as it seems that the user would not even have a
chance to see much at all before making a decision to leave so very
The best way to catch more of these no-visits (if you want to record them)
is to put the WebtraffIQ code way up the page HTML inside the "head"
I believe this would stand the same for Webtrends, Omniture et al
- --- In firstname.lastname@example.org, "Marcos Richardson"
> It turns out quite a few people who click on a Google PPC campaignsthat
> (between 10% - 30%) immediately press "back button" and don't get the
> whole page
> view. Since these clicks don't result in a real page view, I'm happy
> WebtraffIQ accurately recorded this as a no-visit by not recording it.Google supposedly has a quality threshold for what they consider a
legitimate billable click. For instance, in this case, Google says that
they discard click charges for these type of immediate back button
As another person mentioned, the original poster will want to get the
click and cost data from Google for the campaigns of interest and any
3rd party tools like their web analytics package or tracking URLs.
Checking your server logs for the Adwords traffic is a good
triangulation point too if you have distinct tracking parameters in
your destination URLs.