Comparing google analytics with weblog expert

Avatar
  • updated
  • Answered
The weblog files and the google analytics files differ enormously.

For example in July alone my weblog files said I have 798,000 visitors to the site yet my google analytics say 172,000. What can be causing the anomaly.

I feel the google analytics is too low. But the weblog is too high. The only explanation I can think of with the google analytics is that the code is not inserted into every page. Yet I inserted into the header which is on every page.

With the weblog expert it is recording my page views in the millions. Making the page views per user look ridiculous. So I know the weblogs are wrong but don't know why?

Could the server be giving wrong data? If so how?
Avatar
Michael
WebLog Expert is a log analyzer, so it shows all requests for all files on your site. Unlike it, Google Analytics works as a JavaScript counter so it only counts visitors and page views when JavaScript code runs on the pages.

There are few main reasons why WebLog Expert may show more visitors and page views than Google Analytics:

1. There are some visitors that request only non-page files like images and downloadable files. Such visitors aren't counted by Google Analytics but are counted by WebLog Expert.

If you wish to get information on visitors who accessed pages only we recommend you to create an include "File" filter with value %PageFiles%

2. There are spiders that request pages without running the JavaScript code. Most spiders should be detected by their useragent by WebLog Expert, in such case they aren't counted as visitors and their requests aren't counted as page views. However, ones that aren't detected will be counted such way.

3. If you have AJAX requests for files that are detected by pages, they may also cause increased number of page views. If you know that such files indeed exist on your site, you should also try to exclude them using filters.
Avatar
m wild
Thank you that is most helpful
Avatar
Celand
This reply was created from a merged topic originally titled GA vs Weblogs unique visitors is 10 to 20 times different. Why??.

I am not an expert on software and how it works. I use Weblogs for years but recently my programmers also installed GA. There is a huge difference in traffic and in referrer numbers. For example today ( 24h period) my Weblogs shows 25,000 unique visitors while GA shows 3,000, and a forum I contribute to frequently, shows 350 referrers and GA shows 6. Why such a huge difference?? It cannot be just robots?? Any advice would be appreciated. Is there a collection of data difference??
Avatar
Michael
Quote from Celand
This reply was created from a merged topic originally titled GA vs Weblogs unique visitors is 10 to 20 times different. Why??.

I am not an expert on software and how it works. I use Weblogs for years but recently my programmers also installed GA. There is a huge difference in traffic and in referrer numbers. For example today ( 24h period) my Weblogs shows 25,000 unique visitors while GA shows 3,000, and a forum I contribute to frequently, shows 350 referrers and GA shows 6. Why such a huge difference?? It cannot be just robots?? Any advice would be appreciated. Is there a collection of data difference??
It is probably caused by files (e.g. images) accessed directly from other sites. You should try to include data on pages only as described above.
Avatar
m wild
Can you tell me where I enter the ajax files. Do I enter them as urls?
Avatar
Michael
If you wish to exclude AJAX requests for specific profile, you can do it by creating exclude "File" filters in Profile Properties > Filters. You need to enter full file path without domain, e.g. /controller/action/ - you can view the URL in the Access Statistics > Pages report, you just need to remove the domain here.

You can also use wildcards in the filters, e.g. /controller/*
Avatar
Celand
Being a business person only, without any knowledge of programming... I have concluded from the above as follows....( please forgive me the layman's logic)

1) When I post, lets say 10 jpgs on a page of some unrelated forum and some visitor clicks on that page ( maybe to see something else) and even that this specific visitor has never clicked on any of my 10 jpgs, by this, the visitor opening this unrelated forum page Weblogs "sees" this and reports it as 10 visitors.

2) Using the same example as above, GA would ignore them completely unless the visitor clicked on one of those jpgs. Then, GA would report this single click as a single visitor.

If this is at least "generally" correct assumption ( no need to be "geeky" about it ;) ) then that is all what I need to know. I would not know Java from AJAX anyway.

thanks in advance for the confirmation
Avatar
Michael
Actually the situation is different. If you post some JPEGs on a forum, all visitors of the page with the images (not important how JPEGs they viewed) are counted as visitors of your site in WebLog Expert unless you use filters like I suggested. If visitors click on the images and go directly to images (not to a page on your site), the situation doesn't change. And such visitors aren't counted by GA as they don't visit pages on your site.

If a visitor clicks on a link and visits your site, he/she will be counted as one visitor by both GA and WebLog Expert. The numbers may still differ (e.g. because of unidentified spiders), but there shouldn't be such drastic difference like you mentioned.
Avatar
Celand
OK Michael, thanks for your answer... it is clearer now, thus:

1) I will ask my geeks to set up Weblog filter to "reject" AJAX requests

2) If I ONLY want those who clicked on my page then I need to set up a "File" filter with value %PageFiles% ( I hope the geeks figure this out )

3) If I do both ( 1and 2, not only ONE of these two) the other the spiders the numbers between Weblogs and GA will be a lot closer.

Please confirm
Avatar
Michael
Yes, it sounds good, you should try these settings.