Major Stats Changes on Google Adwords Keyword Tool

Sep 08, 2010 8 Comments by

A friend passed along a nice observation made on David Naylor’s site: the stats on the Google Adwords Keyword Tool (GAKT) have changed dramatically.

David Naylor and his associates are known for being continually ahead of most SEOs. Their point here is not to put too many of your eggs in one basket. If you are buying domains based on multiples of GAKT stats without any further analysis or if you are choosing keywords to optimize solely based on GAKT statistics– you are setting yourself up to fail.

I wholeheartedly support using GAKT, especially in an IDN context when it is helpful to have a rough tool to help non-natives separate the wheat from the chaff. But beware of the inaccuracies and drastic stat changes as seen below:

As first noted here, look at this change:

Before & After:


About the author

Aaron Krawitz is an active domain investor and a co-owner of

8 Responses to “Major Stats Changes on Google Adwords Keyword Tool”

  1. Twitted by idnblog says:

    [...] This post was Twitted by idnblog [...]

  2. Leonard BrittNo Gravatar says:

    Wow – so if GAKT is that unreliable what else should one use?

  3. NickNo Gravatar says:

    I’ve noticed this as well, if perhaps not as dramatically, in the context of some merchandise keywords. Had one drop from 360,000 per month to something about one tenth of that. I had to go back and check my old notes to see if I hadn’t lost my marbles. Your point about putting eggs in one basket is well taken. If the GAKT number was where it is today I might have passed on the name.

  4. em johnNo Gravatar says:

    So either the numbers have aways been inflated or they are maybe leaving out numbers that they got from search partners. In any case, that’s a pretty drastic drop for “marketing consultant” . I doubt it only has 1000 searches per month, just instinctively speaking. I look at my keywords today and it all seemed kind of screwy. I was getting 0 results for a couple of major keywords. I’m sure Google will have to come up with some kind of explanation for this.

  5. IDNBlogNo Gravatar says:

    @Leonard – In addition to using GAKT one must think about the name’s appeal to end users in a non-statistical way.

    Put yourself in a company’s shoes and think to yourself — is this the perfect name for my target audience? How many end user companies like this are there out there? Are these end users usually wealthy enough to buy premium domains? Rick Schwartz has a number of good posts on this topic. Many call this using your “gut” but it is really just critical thinking.

  6. MarkNo Gravatar says:

    When I search “marketing consultant” I get 110,000 global monthly searches when I do a “broad” search – and 8800 global monthly searches when I do an “exact” search.

    I have always done an exact search.

    Haven’t these 2 search types always had different figures?

    Was it an “exact” search that showed the 110,000 before?

  7. AllenNo Gravatar says:

    From our database, virtually every keyword has changed down 10x-100x or more. As for which is more accurate, who knows, but for the keywords we track on our sites, it looks like the old numbers were closer, but still estimates.

    The real issue is that the change is way, way, way outside the realm of mere statistical error. Google does owe an explanation on this change, and what it means, since they had held out the previous tool as reasonably accurate enough for making decisions, whatever they were.

  8. Steroids UKNo Gravatar says:

    not much change in any of the steroid and bodybuilding eywords i monitor on a daily basis.

Leave a Reply