We have actually already discussed that case. Right now, it is exactly as
you described. But, those cases cannot explain how 50,000 people can hit
the home page in one day and 48,850 are dropping off.
People might be interested in the solution we came up with for the home
page drop-off problem in a huge place like this:
At first, we had talked about filtering by IP range. But, that would be a
maintenance nightmare, and it wouldn't allow us to track computers that
did not have the website as the homepage, like staff, mobile, etc.
Then, we realized that we do have control over the machine images in the
public areas. And, we could set up aliases for the library home page,
like library.yale.edu/image1, library.yale.edu/area2. So, we could set
the browser home pages to the aliases, and we could tell exactly which
request were coming in from our public computers, and wether they really
were dropping of or not.
Clayton "Andrew" Predmore
Manager, Web Operations
Yale University Library
[log in to unmask]
On 2/7/12 10:52 AM, "Andy Kohler" <[log in to unmask]> wrote:
>If your library's machines all have www.library.yale,edu as their home
>page.... are you assuming that users actually click links to leave
>that home page? User sits at library machine, user sees library home
>page, user types url for gmail or youtube or facebook,,, does GA track
>On Mon, Feb 6, 2012 at 12:55 PM, Predmore, Andrew
><[log in to unmask]> wrote:
>> Thank you for your help. But, there appears to be another problem. The
>> main landing page is on www.library.yale.edu, but almost every link of
>> that page goes to resources.library.yale.edu. Right now, I am seeing a
>> 98% drop-off from the home page. It looks like Google is not tracking
>> visit across the sub-domains.
>> Is there a way to fix this?
>> Clayton "Andrew" Predmore
>> Manager, Web Operations
>> Yale University Library
>> [log in to unmask]
>> On 2/6/12 3:10 PM, "BRIAN TINGLE" <[log in to unmask]>
>>>This can be really tricky to get right when you have a more complicated
>>>site with lots of domains. Since you are all on .yale.edu it should be
>>>easier than crossing .cdlib.org to .universityofcalifornia.edu. If I
>>>understand correctly, you should be able to
>>>_gaq.push(['_setDomainName', '.yale.edu']); on every page and it should
>>>This debugging plugin for chrome is pretty useful
>>>It will help you confirm what is getting sent to google.
>>>On Feb 6, 2012, at 11:53 AM, Predmore, Andrew wrote:
>>>> I have been tasked with updating the Analytics for the Yale University
>>>>Library, and I am having quite a bit of trouble.
>>>> Specifically, I was hoping to only track domain names that included
>>>>library.yale.edu, like www.library.yale.edu,
>>>>resources.library.yale.edu, but the instructions don't seem to cover
>>>>sub-sub-domains like this.
>>>> Also, I was hoping to set up a profile/filter that would show me the
>>>>sub-domains in the reports. Again, I followed the directions but I am
>>>>not getting any results. Well, that's not entirely true the reports
>>>>showing about 30 visitors a day (and no page hits, how is that
>>>>possible?). The main profile is showing 5,000 ¡© 10,000 visitors day.
>>>> Does anyone have experience with this that could help me out? Maybe
>>>>there is even someone from Google at the conference?
>>>> Clayton "Andrew" Predmore
>>>> Manager, Web Operations
>>>> Yale University Library
>>>> [log in to unmask]<mailto:[log in to unmask]>