Heather Green and Sarah Lacy wrote an article regarding the often noted ‘fuzzy math’ of web metrics reporting. They have a point, because third part tracking software that is ‘guessing the visitors’ without actually being installed on the server can in no way be called reliable.
For example, though we are not YouTube or MySpace, we get roughly 65,000 to 100,000 visitors per month (depending on what news is out there) thoughout Yooter owned sites.
We understand that this traffic level might be only 1 day or even 1 hour of traffic for larger sites, however we feel that 100,000 people per month is a large enough sample to try to hammer and nail some standard reporting.
We use statcounter, though it’s entry level in terms of deep reporting, it gives you a vaild picture of who and when and how they came to your site. We tried Google Analytics , Webtrends and virtually every reporting program out there, at the end we settled on a simple solution that fits our needs excellent.
We feel that the problem with Web tracking isn’t with the tracking software, it’s how to read the reports.
Unique Visitors per 24 hour visit is the proper way to read how many people have visited your site, period.
Tracking ‘hits’ or ‘pageviews’ might have some useful metrics, but they don’t measure people, they measure server load, and pageviews does measure how many times that unique visitor moved around on your site.
But it still centers around unique visitors, any other way of tracking visitors is fooling yourself, and your investors.