Monday, March 03, 2008

Enough of the Google / comScore fiasco

There's been plenty written about the hit Google took, partly as a result of the numbers released by comScore. Enough in fact that I don't care to link to any of the reports or numbers. I believe in part it has been an over-reaction to comScore's numbers so for everyone who's been crying the demise of Google - shame on you. And for those who jumped to conclusions on what comScore's numbers meant - you should know better.

Any metrics without consideration of context may lead to misinterpretations - as it obviously did in this fiasco. Few take the time to understand methodologies used by companies like comScore, Compete, and others. Even if you have no idea about their methodologies and take their reports as gospel, when reading these reports, forget about the stats and ask yourself "does this make sense and why".  Panels, and most metrics for that matter, make a number of statistical assumptions which in most part we can live with. Panels in particular by their nature have a number of inherit bias' to them and by using statistical methods, estimate and project numbers. 

I believe there is a great deal of responsibility on individuals and organizations reporting numbers, particularly when they may have an economic impact. Did comScore's latest blog posting come across a bit of an 'oops'? Possibly, but I believe this was a lesson to them that they have a responsibility to keep their metrics in context. While it is not common for them to comment on particular market impacts beyond what the numbers say, I was happy to see they took the opportunity to explain what fewer clicks may mean. Though I am sure it being Google had something to do with it.

I am tired of seeing headlines of clicks, impressions and page views declining. These are poor metrics at most.

I just wonder considering the punch Google stock took, what will the impact be on SCOR? 

No comments: