New Google Keyword Tool Shows Different Results Than Old Tool

[ChrisWV]

I'm comparing the new Google Keyword Tool to the old one while I still can. I notice with the keywords "widgets" and "widget" the results are staggeringly different.

Widget: "Local Monthly Searches" USA 30,400,000 on the old tool and 3,350,000 on the new tool. I double checked to see if I missed a zero and I did not.

Widgets: "Local Monthly Searches" USA 24,900,000
on the old tool and 37,200,000 on the new tool.

I used Google and Search Partners in the Targeting. Without that the difference was even greater.

Does anyone know af anything that might be affecting this? Thanks.

[edited by: goodroi at 7:29 pm (utc) on Jul 22, 2013]
[edit reason] Welcome! Now please go read the forum charter & dont post specific keywords :) [/edit]

[Robert Charlton]

Were the match type settings the same in both tools?

Your question brings to mind what happened when a large change to the keyword tool was made in 2010. Google started basing its data on different Google properties.

Possibly the same thing is happening again, but the differences were much smaller then than what you're reporting here. I'm guessing it's either a glitch or a different match type.

See these discussions from 2010 for reference...

Google Keyword tool - Showing very inaccurate data
http://www.webmasterworld.com/google/4200253.htm [webmasterworld.com]

Search Suggestions Overhaul - Just Before the Google Instant Launch
http://www.webmasterworld.com/google/4200592.htm [webmasterworld.com]

Quoting a Google employee, back in 2010...

...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.

[ChrisWV]

So it looks like Google used to base statistics on "Google.com search traffic" AND "traffic from search partners", then they changed to "Google.com search traffic" only with the 2010 update and now they are giving us a choice. When I used the default "Google" and not the second choice "Google and Search Partners" the variance was even wider. Restaurant was 201,000 Monthly USA and restaurants was 7,480,000. Which numbers should we trust. I'm inclined to just go with the new ones but I don't like highly questionable statistics. Any input is appreciated.

[aakk9999]

I'm guessing it's either a glitch or a different match type.

I would agree with Robert that it is probably because of a different match type. We had this question recently, see [webmasterworld.com...]

[JS_Harris]

Which numbers should we trust?

None of them. You're getting a list of similar keywords but Google intentionally is not making it a comprehensive list and is intentionally not making it current. Unfortunately the tool could be used to game serps if it was overly accurate and current.

The same applies to the figures. Some of them simply make no sense while others seem ballpark. What I do know is that I rank #1 for some fairly hard to get keywords that the tool says 30,000,000+ people look up each month locally but GWT tells me that I only got 12,000 impressions in the #1 position.

I'm fairly certain that the figures include images and partners as well as web searches too making the figures even less useful.

[easylearning]

I've been experiencing the same, and actually stumbled upon this forum looking for others who are experiencing what I am.

The differences between the keywords with the old and new versions are extremely big ... so big that my only conclusion is that Google shows wrong numbers in their results. The differences are also still huge when catering for Google & Search partners, which I've done here:

* example phrase1: old 1,500,000 ... new 14,800
* example2: old 300,000 ... new 90,500
* examples3: old 368,000 ... new 27,100
* example phrase4: old 823,000 ... new 301,000
* example phrase5: old 16,600,000 ... new 1,100,100
* example phrase6: old 2,240,000 ... new 201,000


Well, I definitely do not trust the new numbers ... Noone can tell me that "example phrase1" only has 14,800 monthly global searches ... Not only is it logically way too low from a global perspective, but it is also less than 1% of its previous 1,5 mill avg global searches. How is that possible?

For "example phrase6" it's a reduction of 90% from 2 mill to 200k.

Would love Google's take on this, as this is frustrating to say the least - if you can't trust these numbers, what do you do for decision support?

[edited by: goodroi at 1:53 pm (utc) on Aug 6, 2013]
[edit reason] Welcome to WebmasterWorld, please go read the TOS and dont post specifics [/edit]

[smithaa02]

I'm seeing this too...so weird that the new tool has completely different figure than the old one.

eg I search for xyz widgets.

In the new system it shows 75k per month for keyword ideas. Then for ad group ideas I get a keywords like at 174k and the phrase(25) for 60k.

On the old system I get for broad/exact/phrase:

500k
110k
450k

It gives the same for global and monthly.