Is Google Keyword Tool Search Volume Accurate?

I don’t know if anyone else has notice a huge anomaly in the search volume being reported with the recently updated Google Keyword Tool, but something just isn’t right here.

In doing research for PPC campaigns and content writing I’ll use a number of different free keyword tools just to get a feel of the amount of searches being performed on terms I’m thinking of using. My favorite tools are KeywordDiscovery.com, the Free Keyword Suggestion Tool at WordTracker and of course the Google Keyword Tool. WordTracker reports daily search volume and KeywordDiscovery and Google reports monthly search volume, or at least that’s what their supposed to be reporting. The only problem is that Google seems to be reporting 50 to well over 100 times the volume the other search tools are as you can see below:

Goole Search Volume vs Other Tools

I didn’t notice this right at first, but as I continued to do more research the numbers began to look a little to good to be true. Then I decide to compare the same terms through these three tools and came up with similar results for lots of different keywords. Although the ratio of search volume in Google is probably comparable to other research tools, and still make Google’s tool useful,  it’s a complete mystery as to why Google would report such huge search volume, or am I just completely missing something here? Maybe over-reporting search volume would encourage more people to advertise on keywords in the hopes there would be lots of room in a certain niche when there really isn’t, just a guess.

I have an email in to the helpers at Adwords customer support and we’ll see what kind of answer they come up with. It should be interesting.

Popularity: 6% [?]

Tags: , , ,

If you have enjoyed this post, please consider subscribing to my full RSS feed or you can have posts send directly to your inbox if you subscribe by email.

Bookmarks: | del.icio.us | Digg it | Furl | reddit | StumbleUpon

15 Responses to “Is Google Keyword Tool Search Volume Accurate?”

  1. Yeah I noticed this trend too but i’m using Wordze. I’ll be interested in finding out what google says.

    FYI…. I ran ‘apartment rentals’ through worze and it came out at 4447.

  2. I think Google’s just searching their own services while other free and paid keyword tools look in different crawlers and searchengines. Maybe that will give a huge different between those results…

  3. I’d trust Google, who has this data & knowledge first-hand, before I’d trust any other service at estimating the number of searches.

  4. Eric,

    I would tend to think that too, but having a couple of sites in the apartment rental niche makes me think that it’s not accurate.

    Doing a quick search on apartment rental terms using the Google Keyword Tool I’m guessing Google is saying there are 60 to 70 million apartment related seaches per month. If that were so, I’d be getting 10 times the traffic I’m getting.

  5. They’re pretty rounded, but I run accounts for a bunch of clients in addition to my own affiliate stuff and the search impressions are pretty close to what they’re reporting.

    Of course wordtracker is off because of where they get their data. They can make estimates based on market share of searches, but they’re ignoring who typically uses their search engines.

  6. 1.The reason may be what Manuel said. Keyworddiscovery and wordtracker emphasise on other search engines

    2.Google will be the best source to know how many times people search for a keyword.

    3.Adwords tool is for advertisers and PPC marketers and google wont try to cheat them. DO NO EVIL.

  7. I think that tool is still in “beta” because i see empty bars for competition on some keywords, and when i type that in search, i get tons of ads.

  8. I get what appear to be inaccurate results too, most well overestimated but some underestimated. I have a number one ranking in Google for one of my sites’ main keywords, and I got 3 times the number of visitors the Adwords tool suggested there were searches for. And not everyone’s going to click on the number one result, meaning it must be even further out.

  9. I appreciate everyone’s input. I’m kind of anxious to see how Google responds to my inquiry.


  10. I have been using Google and can say that not all keywords are correct. Most of the time they are, but not always and that explains the huge numbers.

  11. My own take on the new tool:

    Most keyword tools use an exact search method for the results, in other words, surrounded with brackets or quotes. Ex; “apartment rentals”

    When you search G’s new tool, it is returning the most broad match to your string.

    Searching for apartment rentals on the G tool, returns not only the exact phrase, but every variation of the phrase as well. (Broad Search)

    ex: apartment rentals = 1
    dallas apartment rentals = 1 search, but incrementally increases the first phrase by 1 as well.

    Thus, while apartment rentals shows a volume of 673,000 June searches, that figure also includes every long tail search that included the phrase as well.

    Another anomoly in the Google search data is that they (G) have a HUGE network of sites using their search widget onsite. AOL, Netscape, and hundreds of thousands of webmasters using the Google search widget.

    My assumption is that ALL of these are included in that figure…

    WordTracker and every other keyword tool can only estimate the volume in Google based on Nielsen data, ie: Google has 65% of search volume and Yahoo has 20%, therefore, if Yahoo reports 100 searches, Google must have 300% more in its own engine.

    Anyhow… thats my .02 cents!

    Good post!

  12. Interesting. yeah it would be nice to know what is really going on.

  13. If you choose the exact “Match Type” instead the “Broad” one you will get 33,100 results – only. This is also not even close to the other free tools. Maybe we should get an average of all these tools to be more sure what we should expect for each keyword.

  14. I definitely agree with Mark – you have to pay attention to the match type. I tested it out with a bunch of our high-volume, well-ranked keywords, and so far everything I’ve seen appears to be pretty accurate.

  15. The only real way to know is to run a PPC campaign at a high CPC for your keywords and make your ad unclickable (meaning a title of “kjf fsaj flsj flj” with garbage for a description and destination url.

    You can then check your impressions and get an accurate number from google … multiply by marketshare.

Leave a Reply