Let’s Have Some Statistical Consistency

If the blogosphere is ever going to be taken seriously by advertisers and investors, industry standards need to be established when it comes to measuring traffic. This issue, which has been bubbling under the surface for too long, is highlighted by VentureBeat, which put Quantcast, Alexa, Compete and Google Analytics through their paces by using the popular Streetfire.com blog as the guinea pig.
Quantcast and Compete’s data was far below Streetfire’s server logs and its Google Analytics data, while Alexa’s data was not taken into consideration. (Is Alexa still relevant anymore? Any can one consider its tracking tool at when it relies on people downloading its toolbar, and then using these people as a representation of the entire Internet population). In the end, Quantcast re-loaded with Streetfire by putting a tracking pixel on the site, and discovered traffic was higher than earlier estimate. Compete, however, wouldn’t put on a tracking pixel due to patent concerns.
The VentureBeat experiment makes for an interesting story but it is illustrative of how a hodge-podge of different methodologies, algorithms and techniques is making a mockery of the Web stats business. Sure, it’s a competitive market and everyone wants to use a different secret sauce to come up with the “most accurate” results but how does it benefit the blogosphere’s evolution if you are not sure who’s right and who’s off the mark?
Whether it’s the Internet Advertising Bureau or Nielsen/NetRatings or Hitwise or ComScore, someone has got to take the lead on blog statistics and propose there be industry standards established in how information is collected and presented. This must happen before anyone take’s the blogosphere’s claims of rapid growth seriously.
Note: I recently tried Quantcast but removed the tracking pixel after its data reported significantly lower traffic than AWStats, Google Analytics and The Good Blogs. I’ve also tried Get Clicky but dumped that pretty quickly.