Doing some research on the ComScore�s methology and tech behind it i stummbled upon a relatively hot story, where the ComScore prediction actualy had a rather nasty impact on Google`s stock in 2008

The background story :http://www.wired.com/epicenter/2008/04/the-comscore-go/

Google had however exceeded by far the prognosed revenue. The discrepancy is explained by ComScore in the fact they measured only US market and other factors such as Googles clickprice policy, wich they can not measure. Indeed!

ComScore presumably validated their data:

ends here: http://blog.comscore.com/2008/04/the_proof_comscores_google_pai.html

i tried to explain in a comment exactly why they can never hope to valid any data that corelates Google PPC click volume to Google`s revenue. But the comscore blog had errors and i couldnt publish a comment there.

So here it is:

"HI,

In all probabillity a situation when Google determines a decrease of PPC click volume yealds more revenue is very plausible. And they would make sure a decrease happens in such a situation.

This means you can not measure anything that  you can correlate with google`s revenue.

To add a methodology issue: a corellation is never a causational relationship of factors - but then again it can be a explanatory factor of previous factors ;) "

Does that make sense?