If you're like me, you use the SEOmoz Tools on a regular basis.  An old favorite of mine is the Keyword Difficulty tool. 

If you've used this tool, you know that it requires input from the Wordtracker Keyword Suggestion Tool.  WordTracker has been around since the early days of search marketing.  Their data is mined from two search engines - Metacrawler and Dogpile.  Remember them?  If not, you're not alone - they don't hold much of the search market.  As an interesting aside, though, search for "search engine" in Google - Dogpile.com ranks #1 (at least on my end).  Beyond the small portion of search usage that WordTracker is privy to based on their license for Metacrawler and Dogpile data, the numbers they report are odd.

Let’s use an example:  “search marketing.”

Using WordTracker’s Free Keyword Tool, I ran a query for search usage for keywords related to “search marketing.”  Here’s a snippet of the results:

The local-specific keywords are intriguing here.  According to these results, “las vegas search engine marketing” was searched for 14 times during the last recorded day, “search engine marketing maryland” 9 times, and “vegas search engine marketing” 9 times.  Are these numbers accurate?  Of course we should take them with a grain of salt as with any keyword research tool, but I don’t see any reason to think that they’re way off base.  The issue here, though, lies in what is missing.

To put it simply, I find it hard to believe that people are only searching for search engine marketing in Las Vegas or Maryland, even on this given day.  What about New York?  Los Angeles?  Chicago?

Let’s try querying for “search engine marketing new york” with this tool, just to be sure.

So what gives?  Why are we seeing no results for “search engine marketing new york,” but 14 searches in the last day for “las vegas search engine marketing?”  Did I just pick a keyword that has no search usage?

I ran these two keywords through Google’s External AdWords Keyword Tool, which started reporting actual search usage a few months ago, and here are the results:

• “Las Vegas Search Engine Marketing” - average 390 searches per month
• “Search Engine Marketing New York” - average 880 searches per month

This is just one example - play with the WordTracker Free Keyword tool, and you’ll find the data to be unreliable in this regard across the board.

So I pose the question: is WordTracker data actionable or even informative?  Should SEOmoz be relying on this data for the calculation of the Keyword Difficulty tool (a tool I love in concept but have stopped using for this reason)?

I don't want to pick on WordTracker - it's great to have another source of keyword data other than Google, and they provide a lot of other tools that are very useful.  But it seems clear that we need to take this data with one ginormous grain of salt, if we're going to take it at all.

What do you guys think?