Google Keyword tool - Showing very inaccurate data

[corepartner]

I have been doing keyword research for hours using google previous keyword tool as I liked it more.

After finding few keywords which shows high search volume in google keyword tool (previous version), I just thought to double check these keywords using there new interface.

And I found that the new google keyword tool only showing "400" search volume (exact match) for the keyword which has "40,000" (exact match )search volume according to the previous google keyword tool.

I then checked the keyword which gives me good amount of traffic and I picked that keyword using the google previous keyword tool and again the new google keyword tool showing only "50" search volumes/month for that keyword, which is totally incorrect.


I checked more keywords and found that google new keyword tool is showing far less search volumes then the previous one and I found it very much inaccurate.

My criteria of search:

I used exact match
I only focus on local search volumes

I guess google is trying to make it more difficult for the webmasters to pick the right keywords, I hope it would be just a bug....please share your views about it.

[mrez74]

I've noticed the same for any of my keywords. The new keyword tool shows exact match search volumes that are 50%-80% less than what the old keyword tool shows.

I'm thinkg that the true number somewhere in between those numbers. But the weird thing is that some of my sites that I have been number #1 for a few months and get most of my traffic for one main keyword the new numbers kinda match with what I'm getting.

[tedster]

that's my experience so far - as well as some other people i've talked to about it. The new numbers are a lot closer to the real world.

[corepartner]

Since you guys are saying that the data new keyword tool is producing is more accurate but it showing so too much difference in search volume.

I found a keyword which has almost "50000" local searches using exact match.

Now I have checked that keyword using their new tool and it is showing only "46" exact match search volume for that keyword.

This is a huge difference now I am very confused, I don't know if I should run after that keyword or not. ANY ADVICE?

[corepartner]

I have also found another interesting data about as I told you I am seeing huge difference in search volume of a keyword.

that's why I went to google insight and search for those keywords and google insight is showing the data for those keywords and we all know that google insight wont show the data for any low volume keyword.

I used the google insight using the quotes "keywords"....please share your experience here as we all know that keyword research is the most important thing and I don't want to target the wrong keywords.

[tedster]

Any one data point can always be botched. However, a 50,000 search LOCAL terms sounds pretty dubious to me unless it's [pizza] and the location is New York.

More seriously, for any decision that has a lot riding on it, I always augment Google's data with at least one other source, usually a paid service.

You can also see if anyone is actually bidding for ads on the term and get a feel for it that way. Heck, you could even buy an ad with a low budget cap jand learn a bit in real-world application that way.

[tedster]

Here's a relevant comment from a Google employee on the Adwords Support forum:

If you use both the previous and updated versions of the Keyword Tool to search for keywords, you may notice differences between the tools for statistics on Global Monthly Searches and Local Monthly Searches.

This is because the previous version of the Keyword Tool provides search statistics based on Google.com search traffic AND traffic from search partners, while the updated version of the Keyword Tool provides search statistics based on Google.com traffic only. We've updated these statistics based on user feedback, and hope you find them helpful for keyword selection.

[google.com...] [emphasis added]

[edited by: tedster at 6:03 pm (utc) on Sep 16, 2010]

[McMohan]

Can I make a related question? I use the old interface, mainly because I like the way it shows search volumes of phrases that include the keyword I typed in. Whereas I see that missing in new interface, which lays more thrust on returning numbers of related phrases.

The new interface does have an operator "Show results for Ideas containing my search terms", but if I selected that, almost always it returns no result.

[corepartner]

Thanks Ted for the link but I am not liking asking questions again and again but I don't understand "you may notice differences between the tools for statistics on Global Monthly Searches and Local Monthly Searches. This is because the previous version of the Keyword Tool provides search statistics based on Google.com search traffic and traffic from search partners, while the updated version of the Keyword Tool provides search statistics based on Google.com traffic only" what does this means?

Kindly reply :)

[tedster]

Google provides search results not only through their own website but also through partner websites such as AOL.

So there are two different data-sets for search frequency numbers. One (the original tool) includes all the sites where Google provides search, The other (the new tool) only includes search frequency on Google.com

[jonathanleger]

While that explanation certainly makes sense, it's not playing out for all of the keywords (unless I'm doing something wrong).

The new interface for some keywords I follow is showing more searches than the old interface (both global monthly and local monthly searches). For the particular keyword I checked, the new interface shows 40,500 global searches, whereas the old interface shows 27,100 global searches.

If the old interface always includes the search partners, shouldn't it always show higher volumes?

[jonathanleger]

Woah! Found another major discrepancy with the reply from AdWordsPro.

I checked another set of keywords for my sector, and the new interface shows 9,140,000 global searches, whereas the old interface shows 246,000 global searches (exact match).

When I switch the old interface to broad match it still only returns 1.2 million global searches -- far, far less than the new interface.

I just don't see how these discrepancies can be attributed to the new interface not including search partner data.

[added]FYI, I have both interfaces set to US/English only data.[/added]

[seabreeze]

My interpretation is a) "google local exact word match" is the exact count for the prior month; b) "google global exact word match" is a rolling 12-month average for the prior 12-months. So "bikini" or "ice cream" would have higher "local exact" matches for June/July/August than December. The "global exact" can be higher or lower (and should be) than the local because of the seasonality issue - if the searched item has seasonality aspects.

The new interface is flawed as the numbers don't represent reality. "Insurance" has 5,000,000 global and exact. "Car Insurance" has 90,500 local exact. There are to many inaccuracies to use this data, the new google interface, with any conviction or certainty.

I appreciate the fact 3rd party search parties are separated from the new interface, but even factoring in this variance, the data simply isn't logical.

[tedster]

I just don't see how these discrepancies can be attributed to the new interface not including search partner data.

even factoring in this variance, the data simply isn't logical.

Agree with both comments. I just thought the "official" input was a useful factor to consider.

[Sgt_Kickaxe]

My best keywords don't show up on any Google suggestion tool, go figure.

[nickreynolds]

Last month I had a keyword (low number of searches but enough to show up on the tool) that I had way more visits from Google than is shown on the new keyword tool per month. I'm about position 3 on google for that search term.

[vetofunk]

Can I make a related question? I use the old interface, mainly because I like the way it shows search volumes of phrases that include the keyword I typed in. Whereas I see that missing in new interface, which lays more thrust on returning numbers of related phrases.

The new interface does have an operator "Show results for Ideas containing my search terms", but if I selected that, almost always it returns no result.

I see the same thing. Anyone else having this same issue when they just want to see keyword results having the keyword you are searching for in the phrase itself only?

[mkelly7777]

Agree that the new keyword tool is much less user-friendly and provides a lot less valuable data versus the new keyword tool. Google - if you are listening, please don't get rid of the old keyword tool. Or, if you do, please get feedback from this forum before you do so we can help you fix the problems with the new keyword tool.

[jskrewson]

I continually go back to the old keyword tool, because I need to know the phrases that contain the keyword I specified. How else can we reliably generate negative keyword lists and ascertain the most probable "intent" of a search?

The new keyword tool feels more like a tool to increase spending, not a tool to increase ad quality... Can't we have both?

[bluelook]

With the old tool gone it's time for everyone to tell them we want it back!

[JasonP]

Typically I use it more as a comparison between phrases more than caring about exact search volumes.

[bluelook]

Of course we use these tools to get rough ideas, but even if some advertisers asked for the option to only have Google.com results the truth is that many of use also like to see the total numbers, including search partners, so why not giving us that, as an option, even with a big disclaimer?

Without those search partners data you don't loose only numbers, you loose a lot of quality related phrases because the new tool says that some of them have 0 or xx searches, when I know that they have much more than that... (checking my own referral stats).

Google, you had the best keyword tool in my opinion, because most others fail for non-english keywords. Please don't kill one of the bests tools you have. I ask this as an advertiser and as a publisher that wants to continue providing quality content for search phrases that are really being used. I don't have any complaint about the old GAKT. Some numbers seem overestimated? I learned how to read them. But now I'm loosing precious related phrases...

[corepartner]

I think Google has shut down the old keyword tool, as I cant find a way to access it.

let me know if anyone can access it?

[Robert Charlton]

I think Google has shut down the old keyword tool, as I cant find a way to access it.

I no longer see the "link to the previous interface" on the new tool.

Additionally, the Google Inside AdWords blog yesterday carried this announcement...

Updated Keyword Tool: Out of Beta
Friday, September 24, 2010
http://adwords.blogspot.com/2010/09/updated-keyword-tool-out-of-beta.html [adwords.blogspot.com]

This is the latest official wording on the stats change, essentially the same as noted earlier in the thread...

...we’ve also changed how we calculate Global Monthly Searches and Local Monthly Searches. Statistics in these columns are now based on Google.com search traffic only. Previously, they also included traffic from search partners. We've updated these statistics based on advertiser feedback, and hope you find them more helpful for keyword selection.

So it's likely the old tool has been taken down.

[bluelook]

I sent an email to the Adwords team asking them to give us the option to use the old data.

If some advertisers only want Google.com's data, I understand, but what about the others that want the information from search partners as well and don't like the new interface?

I thought that the user experience was very important to Google? If they search they will find many people saying that they prefer the legacy tool.

We just ask for the possibility to use it, hidden or not.