However, they need some work on the way they present the data. Check Alexaholic.com for an example of how they ought to be giving the info.

http://blog.compete.com/files/compete_flash/

By launching Compete.com we are opening our data and hoping to include more contributors in the exchange – the Compete Toolbar is the community’s vehicle.

We are excited to open our data to the world, but are also realistic and understand it will be scrutizined and challenged. From today forward Compete’s user base will grow, the time and rigor we invest in normalizing the data will increase and we’ll work further and further down the tail…

We would love to hear from you if you have any questions or suggestions.
membersupport(at)compete(dot)com

Thanks.

In the FAQ, Complete explains why their stats would be different than what shows on a site, using internal log files:

“Compete’s definition of ‘people’ is also different compared to traffic reported through common local analytic solutions and traffic log analyzers. Compete measures U.S. consumers, which means we will only count a person once no matter how often he or she visits your site throughout the course of a month. In comparison, local analytic solutions will include domestic and international traffic and often include spiders and bots that appear as traffic, but do not represent actual human activity. Compete’s services do not rely on log files or cookies and therefore do not count this activity as traffic.”

So how does Complete explain being as far off as they are (and for the sites I checked they were waaaayyyy off) as compared to perfectly capable stat applications that already filter out international traffic, spiders/bots and repeat visitors?

I found the answer in the FAQ to be quite arrogant (gee, you mean all those “hits” from some guy called “googlebot” weren’t real visitors? Wow, I had no idea…gee, thanks, Compete). Maybe it’s good for larger sites (top 50,000) but for those of us with sites ranked in the 250,000-500,000 ranges I wouldn’t consider Compete a valuable or fair measure at all. Do people really just use these things to compare and analyze monster sites? Seems to be another Alexa-like service that relies on its own toolbar to give information about the surfing habits of people that use the toolbar. Not much more.

Snip:
Compete gets its data in the following ways:

* The Compete toolbar – works very similar to Alexa toolbar, in that “community members” download the toolbar and opt in to share information on the web pages they visit, as well as other marketing data.
* ISPs
* Opt-In Panels
* Application Providers

Comparing THEIR estimation for battellemedia and Federated Media – with the private data only you have available…

How close are they?

snapshot.compete.com/battellemedia.com/?domain2=http://federatedmedia.net/

All of the readers with popular sites could offer their input about the accuracy as replies to this Post – allowing for a collection assessment – that would be valuable for everyone