google-predictive-queriesMany of us are familiar with are familiar with Google Webmaster Tools.  Regardless of how experienced you are with this particular Google feature, we all know that GWT can give us some nice information on how our sites are doing.  This alone is what it takes to give marketers what they need to help create even better websites for themselves or for their clients.

But there seems to be a problem.  This problem could potentially be pretty big depending on what you need.  It seems that GWT isn’t always totally accurate with some of their data.

“What’s this” you may ask?  “Sacrilege!”

Well, I can tell you I’m not the first to say this.  Rand of Moz has said this during his AMA on the BigSEO subreddit.  One thing to keep in mind about the current topic is the focus on the accuracy of the Avg. position metrics, not the accuracy of other metrics like impressions, clicks, or CTR.  The problem with using averages, and some of you may be a head of me on this, is that you’re not getting very specific with you numbers.

As an example, if my site, on one day, got 200 visitors, and on the next day, got 1000 visitors, and on the third day I only got 150,  and GWT presented me with the average of those three days, which is 450, I would have a number that clearly doesn’t really help me much on showing me the numbers I want to see.

I know the numbers I used in my example might be a little extreme, but I’m simply using them as an example to prove the  point that averages aren’t necessarily the best way to go when trying to measure certain metrics.

 has written a post for Moz entitled “Testing The Accuracy of Average Position for Search Queries in Google Webmaster Tools“, and talks about the test he ran to see how accurate the averages for position for search queries really was.

Check out what he did, how the test was run, and see the results that he came up with in the following link.

YouMoz: Testing The Accuracy of Average Position for Search Queries in Google Webmaster Tools