Posted by Melanie Phung on Monday, November 24, 2008 at 11:54 am
We all know that panel data can produce some odd results when sample size is really small — exaggerating trends that in reality might not signify anything at all or missing some data altogether. In the case of reporting search traffic to my site, however, the panel data from Compete.com seems to be pointing at something that really doesn’t exist.
Check it out:
Compete.com says that the term “Jesus Christ” is responsible for one quarter of my search traffic.
Obviously I don’t expect Compete’s free data to match my (also free) analytics program perfectly, but I can say with a lot of confidence that this data appears to be sampling something that can’t possibly exist.
Now before you go around decrying me as a heathen and a heretic, my point here isn’t that Jesus Christ doesn’t exist … simply that he certainly does not exist on this blog.
I’ve never, ever used the phrase “Jesus Christ” on this site. Until now, of course. No, in general I tend to favor exclamations like Jeebus! or Good Gawd! or Sweet Lawd Almighty!
Nor have I, to the best of my knowledge, ever been Googlebombed with that term.
In short: I do not rank, and there’s no reason for me to rank, for the search term [jesus christ] — And showing up in search results would seem to be a prerequisite for driving search traffic.
Here are the terms that drive search engine visits according to my analytics program (although no single term drives anywhere close to 25% of my search traffic):
See? No Jesus.
While I do use Compete.com for research and competitive intelligence, I’m going to be taking their data with an ever larger grain of sand. Their data isn’t just skewed, in some cases it’s just patently wrong.
The idea that someone looking for Jesus Christ would find Him on my blog struck me, frankly, as insane. But I hear He works in mysterious ways, so I’m just going to go ahead and go with it.
Sorry, the comment form is closed at this time.