Web Site Analytics Revisited: Numbers Numbers Everywhere

There's nothing more important than measuring how many people are coming to your web site, and what they're doing. And there's nothing more frustrating than trying to believe these numbers are 'accurate'.

The web wasn't built for precise measurement. So analytics vendors do their best to reliably capture data and overcome the many strange and unfriendly technical realities. And using a good web analytics package (meaning reading it daily not just installing it and checking overall site traffic once in a while) can improve nearly every aspect of your online business efforts.

But when you try to match what your analytics say as compared to your accounting system, or your ad server, or your CRM package, or outside services such as Comscore and Alexa, or your paid search engine reports, any other number that by all rights should line up darn near exactly, you'll find that website analytics is latin for 'a hopefully close approximation'.

Or the analytics are perfect an each of these others are screwy. Of course, since NONE of these ever seem to line up to any other, there's a pretty strong case to be made that there's plenty of blame to go around. (Excluding your accounting system of course, that one I'd believe...)

This all came to mind this morning because the 'A-List' is an uproar about a Comscore study on blog traffic and advertising, which publicly reported some numbers that challenge the assumptions of who's "A+" and who's just a lowly "A". A summary and links galore can be found on Buzzmachine.

This study was a panel study - where they asked real people what they do - so of course both the methodology and 'wetware errors' guarentee these numbers are way wrong. Jarvis of course, nails this:

My first reaction is that all this shows how messed up panel research is. This is the method used by Nielsen et al to measure TV and radio and print readership — affecting billions of ad dollars — and it is and always has been relative bullshit.

The bigger picture is that unless the basic technical mechanisms of the web change, we're all going to have to get used to the fact that web site numbers are approximate. This doesn't make them any less valuable or important. Think of them as right but not accurate. Look at all the numbers you can get from as many reasonable sources as you have, check your gut, then take some action based on the average of it all.