I stand confused. I’ve got Google Analytics running on this blog, and a system provided by Quantcast, but according to these systems, I don’t really get many visitors per month. These hardy types view around 4,500 pages. Nothing special at all.
I recently had a look at the Urchin 5 statistics system that comes with my Media Temple hosting service plan, and I saw over 88,000 page views for May alone this year! OK, so this includes my own visits and my other two domains, but even so, that is one heck of a difference. The average session length is around 6 minutes according to my hosting server stats too, which seems to be quite a long time for search engine bots and the like to be spending sniffing around this blog.
The difference is weird, and I don’t know which numbers to really believe. I’d love to think that 88,000 was right, and it is an ego massaging figure, but I suspect that 4,500 is the more accurate representation of the actual number of pages that are read by real humans. Even so, I’m not sure how to explain the enormous difference.
If anyone out there can tell me why there is such a difference, then I’d love to know.
Article continues after the advertisement
About Alex Roe
Italy Chronicles Feed
Alex,
From what you’ve written, you have both a server side measurement tool (Urchin) and a browser based measurement tool (Google Analytics).
Server side measurement tools generally capture ALL website activity by default – page views, file downloads etc, by both robots (such as, but not limited to, search engine crawlers) and people. Activity can include both success and failure (page not found). Browser based tools on the other hand generally ignore robots, file downloads and non-successful page requests.
I say generally as there is a lot of latitude in how a tool can be configured. In both cases, successful data gathering and subsequent analysis is dependent on the proper configuration of a tool to fully exploit its functionality for a given website. Browser based tools, such as Google Analytics, will blissfully ignore any pages you haven’t tagged, with no possibility to retroactively fix the problem. You can track some file downloads with Google Analytics and other browser based solutions, as long you add specific tracking code to your setup. Yet none of these browser based tools will track downloads by someone who directly links to a file from a third site or otherwise bypasses the tracking code in a web page by using wget or similar.
Despite what many web analytics sales people will try to tell you, neither the server side nor browser based methodology is intrinsically better than the other. The key is to understand each methodology’s strengths and weaknesses and to insure an optimized tool configuration for your site. I do recommend that a web dependent company consider use of multiple tools, resources permitting.
For more on the topic of web analytics: http://www.antezeta.com/web-analytics/#WA5
- Sean
Hi Sean,
Thanks for all the informative info.
However, the difference between 88,000 and 4,500 is still huge! Even so, differentiating between what are human visitors and bots/spyders seems to be just about impossible, unless some serious manual analysis is carried out. There must be some accurate system around though. I would have thought views via browsers would have been a good guide – always assuming that the browsers do not have javascript deactivated!
I guess that one day, someone will come up with a reliable way of measuring how many times a site is visited, and which pages etc are viewed. So far though, it appears that nobody has done this.
Have a good weekend,
Alex