This post forked from a big one that I started here, because new and very strange data popped up...

During the discussion me and another WPW member shared our screenshots about search demand that we see for a specific keyword.

I was shocked to see that under the same search parameters (English, All Countries and Territories, Exact Match) Google keyword research tool showed to carpediem (thanks for the help, Denielle) 480 searches and to me (freetraff) 2,900

THE SAME SEARCH PARAMETERS, check out these screenshots, people.

This one is from carpediem:



And this is mine:







How is that possible that Google shows us different search demand?

The search parameters are the same (I am repeating this 3rd time, because some people did not pay attention to it when this issue got clear in the previous post).

We are just sitting on different IPs, but this should have nothing to do with the search demand, because search demand for a specific language and territory does not depend from where I am checking it.

480 vs 2,900 - this is 6 times difference!

Why? What the heck? What is Google doing to us?

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

My 30 year of IT experience is that in 99% of the cases it is an user error.

__________________
Mini Network:: Financial information at your fingertips
Learn object oriented programming where it started

If you look at the two screen shots again, the first one has the "use synonyms" box checked, the second one does not, maybe that is what produced the different results?

__________________
2009 Hairstyles - Pictures of 2009 hairstyles and a virtual hairstyler demo.
Price Comparison Site - Compare prices of well known brands and products.

But you saw the screenshots. What is the mistake? Parameters are the same, words are the same - screenshots show it all.

I swear I did not do any changes in PhotoShop or other stuff. I am sure that carpediem did not change anything.

I do not see any possibility for a mistake, or these screenshots are a MATRIX and we do not see what we see, but this is a cranky explanation.

I also thought about it BUT if carpediem would use Google help in synonyms that she would have even MORE than me, because Google would also use the synonyms.

Plus, I checked this on my IP - ticking "Use synonyms" does not change the number for me, I have 2,900.

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

Let's do a test - what is YOUR number for:

0 car finance
English
All Countries and Territories
Exact Match

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

I tested it here from UK and got the exact same results as yourself. It is odd, all I can think is that maybe ip location does have an effect, not sure why it would, but it does look like it does.

I do not use that tool, but I see three possible explanations for the different results:

1. You do not use the same tool.
2. You use it differently.
3. The input / data is different.

So to those who test it, agree on 1, 2 and 3.

__________________
Mini Network:: Financial information at your fingertips
Learn object oriented programming where it started

Maybe if you are based in UK the keyword tool calls for results from a data center that has different results. Although you would expect that query all countries and territories to be an amalgamation of all data centers results, maybe it is not, because maybe how often your ads are shown even if showed worldwide, is somehow still linked to which data center your adwords account is associated with...

Its just a thought

Ah. This reminds me of something.

I used to work for a large telephone directory enquires firm in the UK. We had three servers. Each server had the same data set on it. But often they would throw up different returns on a search.

I.T. had various answers as to why this happened including that the servers were rebooted at different times and took a while to synchronise, so would give different returns while this was happening.

Might be a similar problem with Google?

__________________
http://thatsnews.blogspot.com http://thats-business.blogspot.com/ http://thatschristmas.blogspot.com

1. The same tool - please check screenshots.

2. We use it the same way - please compare the parameters of search (it is impossible to use this tool another way).

3. Input data is the same - please check the screenshots.

I know this 'check the screenshots' sounds annoying, but they show everything you need - we are doing the same.

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

I understand we are just guessing - we are not Google.

But if your assumption about datacenter is correct, then AdWords tool is mostly of no use.

Let's say Mike in Canada wants to buy some AdWord traffic of the French people who are searching for the 'visit Canada' (he wants to milk the French tourists for money in Canada - legal wish).

So, Mike wants to see how many people in France are searching 'visit Canada' via Google.

Forget it, Mike, You cannot do that! Because Google will show you and your potential visitors in France different search demand, as different datacenters show different results. The Frech search demand for 'visit Canada' that you see from your place can be several times different from the real search demand in France.

"But", says Mike "I do not care about datacenters! I want to see how many people in France search Google for 'visit Canada'! Google tool has necessary options!"

Sorry Mike, tough luck. Google has too many datacenters. Better take the info from the stars, or go through the meditation cycle, open your chakras and the universe will tell you the necessary search demand.

P.S. So, if the assumption about datacenters is correct, then ladies and gentlemen, your attempts to use AdWords for geo targeted purposes are just a dream.

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

Freetraff,

I'm not doubting your results and am watching this topic with interest. I just have a suggestion to help you focus everyone's attention.

Perhaps you could create a new screenshot of your own results, but make it identical to carpediem's in everything, including which columns are displayed. At that point there will be no distractions to the conversation by people second-guessing your results.

Just a thought.

__________________
Jade Burnside, Ahead of the Web
What good is your web site if no one can find it?
SEO & Optimized Web Site Design

Will do that right away.

BTW, how much do YOU have for:
0 car finance
English
All Countries and Territories
Exact Match

Just interested...

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

As suggested by spiderbait I made a screenshot that has the same columns like carpediem

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

Then it is 3 in my list, different data is used because the replication is not identical.

__________________
Mini Network:: Financial information at your fingertips
Learn object oriented programming where it started

Are the screenshots showing the same parameters?

If yes (I highly doubt you can say 'no')

Then how can the non identical replication be possible?

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions

I can only speculate.

1. There are a global index.
2. Supplied with many local.
3. The local are not an exact mirror of the global and they may vary from region to region.
4. Will such a solution make the index less relevant? No, is my answer. It may make the index more semantic, relevant to the actual region if you prefer that term.

__________________
Mini Network:: Financial information at your fingertips
Learn object oriented programming where it started

I have replicated what has happened FT...

See here my two screen shots showing the two different results that you and carpe diem had. Both are my sceenshots from my pc that I just took one after the other to show how it ocurred....

The first appears to show all countries and territories with the lower search volume....



The second shows all countries and territories with the higher search volume...





It was a simple mistake...as Carpe Diem explained in the previous thread, she had accidentally when taking the screen shot, moved the selection over all countries and territories, whilst the actual search showing was showing the results from United States. This is how I can replicate the screen shot...no photoshopping involved...you can replicate the same...do your search so the results show all countries etc, and then change the selection, this does not change the results until you resubmit...hence the confusion....


Carpe Diem did state in the previous thread

__________________
2009 Hairstyles - Pictures of 2009 hairstyles and a virtual hairstyler demo.
Price Comparison Site - Compare prices of well known brands and products.

Rep. points awarded for a task well done.

__________________
The Penn State Ticket Man http://www.pennstateticketman.com
http://www.happyvalleytickets.com http://www.hounddogtours.com

And I got a reply from Danielle in PM - she confirms that she has the same results as me now.

That is why I think that we can successfully close this issue in 2008 and happily move to 2009. Google AdWords tool seems to be ok!

Happy New Year!!

__________________
Free One Way Links from Inside Real Blog Posts - Free Traffic System
+ Free content for your blogs + Residual affiliate commissions