Google Keyword Tool Exaggerates Search Demand

freetraff · #1 (**permalink**) 12-08-2008, 06:07 AM

This statement is not a copywriting catch.

I have been helping many web sites to get page 1 in Google in different niches. Nothing to boast about, because many WPW users did the same - BUT it was important to mention it, and you will see from the rest of the post.

Since summer 2008 I have been seriously involved into this type of business:

- find keyword with nice search demand (in Google keyword tool)
- build content for it
- build backlinks for it
- grab Google page 1 for it

What I saw is that the search demand that Google is showing for many keywords is often very (VERY) different from what you are getting.

Just one (out of the many) examples.

I picked '0 car finance' with 12,100 average monthly searches and 9,900 searches in Nov 2008. When I am talking about searches, I have chosen the following options in Google keyword tool:
- language - English
- All Countries and Territories

So, this should be a nice global demand for the English searches.

Ok, I did my SEO "homework" and the site in a while got page 1 in Google for this keyword (sometimes even position 1, but the position in not stable - site is quite fresh and jumps position 1, 6, 2, 4 - in any case this is page 1 Google).

I did a quick check via proxy of my positions in US, Europe - fine, page 1. I did not check all possible IPs and countries of course, but I think what I did in terms of checking gives a good hint that page 1 in not just on my IP

And it means that quite a lot of Google datacenters "see" me good enough to have high positions for this keyword.

So, even if we take the November 2008 stats, this should be about 330 people daily who are digging page 1 in Google and searching for good sites about the topic.

I get traffic to the site, targeted traffic. But this is just about 10-20 daily for keyword '0 car finance' (other traffic comes from other keywords in this niche, but this is the subject of discussion for this post). So, even when I have position 1 for this keyword - just 10-20 people from Google for '0 car finance'.

I understand that not all 330 daily visitors are obliged to click on all results on page 1. But my site is exactly about the topic, this is evident from its title and description in SERP, and it is not at the bottom (heck, once again, from time to time it has position 1).

So, I assume that getting at least 50 visitors from 330 daily (claimed by Google) would be more than fair when you are position 1.

= = = = = = = = = = = = =

That was just one example, I have similar situation with ~ few dozens of keywords in different niches.

Moreover, I see that I am not the one who is noticing a mess up with Google AdWords in general. This post is not about keyword research tool, but shows a nasty trend - Google seems to be "doing the best" to make quick buck on its AdWords program.

I know AdWords keyword research tool has never been absolutely accurate (though I always wondered - why?). But when the difference is DOZENS of times, that's too much even for IT!

My assumption - Google is doing that to push AdWords advertisers into the game, making them think "wow, so many searches inside, and the competition is moderate - let's buy more clicks in AdWords for this keyword"

And when they do not see that much traffic, they still can prove nothing, because there are tons of excuses like:
"you ad was not interesting"
"you had to bid more to get a better position"
etc.

Your experiences, ideas, informative comments welcomed.

**crankydave** · #2 (**permalink**) 12-08-2008, 02:54 PM

Are you taking into account all the searches that are being done simply to check rankings and not by a potential visitor looking for information?

Dave

**cw1865** · #3 (**permalink**) 12-08-2008, 04:12 PM

Quote:

Originally Posted by freetraff

This statement is not a copywriting catch.

I have been helping many web sites to get page 1 in Google in different niches. Nothing to boast about, because many WPW users did the same - BUT it was important to mention it, and you will see from the rest of the post.

Since summer 2008 I have been seriously involved into this type of business:

- find keyword with nice search demand (in Google keyword tool)
- build content for it
- build backlinks for it
- grab Google page 1 for it

What I saw is that the search demand that Google is showing for many keywords is often very (VERY) different from what you are getting.

Just one (out of the many) examples.

I picked '0 car finance' with 12,100 average monthly searches and 9,900 searches in Nov 2008. When I am talking about searches, I have chosen the following options in Google keyword tool:
- language - English
- All Countries and Territories

So, this should be a nice global demand for the English searches.

Ok, I did my SEO "homework" and the site in a while got page 1 in Google for this keyword (sometimes even position 1, but the position in not stable - site is quite fresh and jumps position 1, 6, 2, 4 - in any case this is page 1 Google).

I did a quick check via proxy of my positions in US, Europe - fine, page 1. I did not check all possible IPs and countries of course, but I think what I did in terms of checking gives a good hint that page 1 in not just on my IP

And it means that quite a lot of Google datacenters "see" me good enough to have high positions for this keyword.

So, even if we take the November 2008 stats, this should be about 330 people daily who are digging page 1 in Google and searching for good sites about the topic.

I get traffic to the site, targeted traffic. But this is just about 10-20 daily for keyword '0 car finance' (other traffic comes from other keywords in this niche, but this is the subject of discussion for this post). So, even when I have position 1 for this keyword - just 10-20 people from Google for '0 car finance'.

I understand that not all 330 daily visitors are obliged to click on all results on page 1. But my site is exactly about the topic, this is evident from its title and description in SERP, and it is not at the bottom (heck, once again, from time to time it has position 1).

So, I assume that getting at least 50 visitors from 330 daily (claimed by Google) would be more than fair when you are position 1.

= = = = = = = = = = = = =

That was just one example, I have similar situation with ~ few dozens of keywords in different niches.

Moreover, I see that I am not the one who is noticing a mess up with Google AdWords in general. This post is not about keyword research tool, but shows a nasty trend - Google seems to be "doing the best" to make quick buck on its AdWords program.

I know AdWords keyword research tool has never been absolutely accurate (though I always wondered - why?). But when the difference is DOZENS of times, that's too much even for IT!

My assumption - Google is doing that to push AdWords advertisers into the game, making them think "wow, so many searches inside, and the competition is moderate - let's buy more clicks in AdWords for this keyword"

And when they do not see that much traffic, they still can prove nothing, because there are tons of excuses like:
"you ad was not interesting"
"you had to bid more to get a better position"
etc.

Your experiences, ideas, informative comments welcomed.

Is there a difference between the term and 0% car financing?

I've seen this as well where the singular and plural differ. Google will show a result for 'man widget' showing a fairly large volume of searches. Really people are looking for 'men's widget(s)' or 'mens widget(s)'

freetraff · #4 (**permalink**) 12-08-2008, 04:18 PM

Quote:

Originally Posted by crankydave

Are you taking into account all the searches that are being done simply to check rankings and not by a potential visitor looking for information?

Dave

Good point.

But 10 times more - I really doubt.

freetraff · #5 (**permalink**) 12-08-2008, 04:28 PM

Quote:

Originally Posted by cw1865

Is there a difference between the term and 0% car financing?

I've seen this as well where the singular and plural differ. Google will show a result for 'man widget' showing a fairly large volume of searches. Really people are looking for 'men's widget(s)' or 'mens widget(s)'

There was no difference when I was checking this about 2-3 weeks ago. At least, no difference for the position of my site.

P.S. Checked a second ago - no difference. Position 2 both for 0 car finance and 0% car finance.

SIA · #6 (**permalink**) 12-08-2008, 06:12 PM

Think I am going to have to agree that the results are pumped up and down to a certain degree. I have seen items that were listed with no search results which netter my clients quite a few customers (yes I know it depends to a degree on niche, but these numbers were beyond niche) to results that boasted over 2k searches a day that were low touched and received almost no traffic.

I knwo there is a certain degree of specifics involved, but for many years I have taken it as a crap shoot.

**crankydave** · #7 (**permalink**) 12-08-2008, 06:19 PM

Quote:

Originally Posted by freetraff

Good point.

But 10 times more - I really doubt.

Well, if you have 100 sites that only twice a day are checking their rankings, thats 200 searches that aren't going to produce CT's. How often do you check your rankings for the phrase?

I'm not neccessarily surprised at the numbers you've listed.

Dave

carpediem · #8 (**permalink**) 12-08-2008, 09:38 PM

Quote:

Originally Posted by freetraff

I picked '0 car finance' with 12,100 average monthly searches and 9,900 searches in Nov 2008. When I am talking about searches, I have chosen the following options in Google keyword tool:
- language - English
- All Countries and Territories

Hi freetraff,

Please see below and clarify how you received data for 12,100 monthly searches for query "0 car finance". Even with the match type set as a 'broad match' I was unable to arrive at the number.

freetraff · #9 (**permalink**) 12-09-2008, 07:41 AM

carpediem, all I can say is WOW

Please check my screenshot. I guarantee I was not doing any photoshop stuff, and I am sure you did neither.

I am using broad and check what I get

And with the exact match I have 2,900, please check this

freetraff · #10 (**permalink**) 12-09-2008, 08:02 AM

Quote:

Originally Posted by crankydave

Well, if you have 100 sites that only twice a day are checking their rankings, thats 200 searches that aren't going to produce CT's. How often do you check your rankings for the phrase?

I'm not neccessarily surprised at the numbers you've listed.

Dave

Though I am not discharging your point, Dave. I feel the truth is more in the direction of what carpediem is trying to show.

P.S. If what you say Dave is true, than poor AdWords buyers who see this huge demand are really messing up with the wrong numbers.

**crankydave** · #11 (**permalink**) 12-09-2008, 09:14 AM

Quote:

Originally Posted by freetraff

P.S. If what you say Dave is true, than poor AdWords buyers who see this huge demand are really messing up with the wrong numbers.

That's part of my point.

What you are seeing is not neccessarily "demand". All you see is the number of times a query is processed and nothing else. Assuming that those numbers represent potential visitors is a big mistake IMO. The first 2 questions I always ask myself are... Is this what my potential visitors are searching for and is this how they would search for it. Truth be told, over the years I've gotten my better performing phrases from emails, contact forms, and phone calls.

BTW... I get the same numbers that carpediem shows.

Dave

freetraff · #12 (**permalink**) 12-09-2008, 10:11 AM

Quote:

Originally Posted by crankydave

BTW... I get the same numbers that carpediem shows.

Dave

Any clues why I am getting other stats in Google tool? Can IP make THAT huge difference?

I thought by choosing the language and specifying the country you tell Google exactly what you need.

kgun · #13 (**permalink**) 12-09-2008, 11:32 AM

Quote:

Originally Posted by freetraff

Moreover, I see that I am not the one who is noticing a mess up with Google AdWords in general. This post is not about keyword research tool, but shows a nasty trend - Google seems to be "doing the best" to make quick buck on its AdWords program.

I know AdWords keyword research tool has never been absolutely accurate (though I always wondered - why?). But when the difference is DOZENS of times, that's too much even for IT!

See my first post in dburdon'ss thread: Don't trust external data.

Quote:

Originally Posted by freetraff

My assumption - Google is doing that to push AdWords advertisers into the game, making them think "wow, so many searches inside, and the competition is moderate - let's buy more clicks in AdWords for this keyword"

And when they do not see that much traffic, they still can prove nothing, because there are tons of excuses like:
"you ad was not interesting"
"you had to bid more to get a better position"
etc.

Do you mean that Google use AdWords as an intervention variable to manipulate the SERP's?

Quote:

Originally Posted by crankydave

What you are seeing is not neccessarily "demand". All you see is the number of times a query is processed and nothing else. Assuming that those numbers represent potential visitors is a big mistake IMO.

Agree.

Generally: Demand NE Searches AND / OR clicks?

Final question:

Do you use the same reports? The screendumps are so different, so first of all you have to agree on the report you use.

**crankydave** · #14 (**permalink**) 12-09-2008, 12:30 PM

Quote:

Originally Posted by kgun

Final question:

Do you use the same reports? The screendumps are so different, so first of all you have do agree on the report you use.

Personally, I use a variety of reports and sources of data and information, and yes, this is one of them.

@freetraff... okay... I get 2900 using a "broad" match but "exact" match gives 480. Another thing to consider is that when you're using "All Countries and Territories" you've no reliable way of checking rankings in all of them unless you're checking them on IP ranges from all the Countries/Territories. There's a long list of them.

The phrase itself is very regional (local) in nature so I would suspect local results are likely going be preferred. Folks in the US are probably more likely to see US based results as results in Australia are like not going to be of use to them. Compare the CT's numbers you get to volume numbers for the specific regions where you can verify that you rank.

Dave

carpediem · #15 (**permalink**) 12-09-2008, 02:38 PM

Okay I've got the keyword tool to reflect the numbers you got for 'all countries and territories', both on exact and broad. I'm assuming that it was either a PEBKAC issue or the Google tool was on the fritz.

However, freetraff, this begs a new question:

Since the numbers are '480' when under 'United States', and that appears to match up with the number of daily traffic you are seeing?

Quote:

I get traffic to the site, targeted traffic. But this is just about 10-20 daily for keyword '0 car finance'

What are locations for the IP's that you are seeing for your 10-20 daily visits? Are they mostly US based? The numbers would then make sense.

~Danielle

freetraff · #16 (**permalink**) 12-10-2008, 05:09 AM

Quote:

Originally Posted by kgun

The screendumps are so different, so first of all you have to agree on the report you use.

They just look different. And we don't have exactly the same list of columns (for example, I do not have the estimated value per click).

But all other important stuff:

- language choice
- territory choice
- broad/narrow range

... is the same. It is seen from the screenshots.

freetraff · #17 (**permalink**) 12-10-2008, 02:41 PM

Quote:

Originally Posted by carpediem

What are locations for the IP's that you are seeing for your 10-20 daily visits? Are they mostly US based? The numbers would then make sense.

~Danielle

Hi Danielle, this is the screenshot from AwStats directly from my server about the IPs.

As you can see, most of the traffic comes from USA.

And I have one question for you. You mentioned that you managed to get the same AdWords tool stats that I got (I got 2,900 for exact match)

So, did you get 2,900 for [0 car finance] or the same 480 that you had before? (provided English -> All Countries and Territories)

carpediem · #18 (**permalink**) 12-11-2008, 07:51 PM

Quote:

Originally Posted by freetraff

As you can see, most of the traffic comes from USA.

If most of your traffic is US and you are seeing "just about 10-20 daily for keyword '0 car finance", then the Google keyword tool estimate for US seems to match up to your traffic with 480.

16 (avg. daily visits) x 30 = 480

Cheers,

Danielle

freetraff · #19 (**permalink**) 12-13-2008, 10:45 AM

Quote:

Originally Posted by carpediem

If most of your traffic is US and you are seeing "just about 10-20 daily for keyword '0 car finance", then the Google keyword tool estimate for US seems to match up to your traffic with 480.

16 (avg. daily visits) x 30 = 480

Cheers,

Danielle

But that's the biggest catch here.

How come you see just 480 and I see 2,900 under the same parameters [exact match, English, All Countries and Territories]?

I know we are sitting on different IPs. But this has nothing to do with the tool.

The idea of this tool is to give an AdWorder a possibility to see how many people are searching for a keyword:

- on specific territory
- with specific language

So, if I am sitting in Kenia or Iceland, I still can see how many people in (for example) Canada search for "money" in English. And no matter where I sit - Kenia, Iceland, Australia, etc. - the search demand in Canada for "money" in English stays the same. (I do not believe that people in Canada do like this; "ok, we have an adworder from New Zealand, let's search 2 times more this keyword, because we want New Zealand to think that Canadians are searching for this keyword 2 times more")

Then - how come we are checking one and the same keyword, with the same search parameters (and screenshots prove it - you and I have the same search parameters), and Google keyword tool shows:

- 480 for you
- 2,900 for me

???

P.S. I know this can look like a fork from the topic. But quite often a fair discussion opens new - very important - topics. I do not understand how is it possible to show people from different IPs different numbers that should be the same irrespectively of the IP?

carpediem · #20 (**permalink**) 12-13-2008, 01:35 PM

Quote:

Originally Posted by carpediem

Okay I've got the keyword tool to reflect the numbers you got for 'all countries and territories', both on exact and broad.

To clarify further, I got 2,900 (exact) and 12,100 (broad) in English, All Countries and Territories.

The 480 is (exact) in English, United States.

Cheers,

Danielle

freetraff · #21 (**permalink**) 12-14-2008, 06:02 AM

Quote:

Originally Posted by carpediem

To clarify further, I got 2,900 (exact) and 12,100 (broad) in English, All Countries and Territories.

The 480 is (exact) in English, United States.

Cheers,

Danielle

Hmm, Danielle, sorry to keep focusing on this, but your screenshot evidently shows that the choice is:

English, All Countries and Territories, Exact - and 480 searches monthly

I have checked your screenshot dozens of times. Have you been changing something about it in graphical editors?

If not, then it shows: English, All Countries and Territories, Exact - and 480 searches monthly

afftrackin · #22 (**permalink**) 12-19-2008, 04:38 PM

Are you using EXACT MATCH? as when you select that you get a lot lower results. Its important that you bear in mind that this tool was designed for adwords users first and foremost.

I always go by the EXACT, as that really is the competition and the phrase that people would have to be typing in, all else could just be too broad, like " what are the best 0 car finance companies "

Second make sure you are selected the correct search results country

afftrackin · #23 (**permalink**) 12-19-2008, 04:39 PM

Quote:

Originally Posted by carpediem

If most of your traffic is US and you are seeing "just about 10-20 daily for keyword '0 car finance", then the Google keyword tool estimate for US seems to match up to your traffic with 480.

16 (avg. daily visits) x 30 = 480

Cheers,

Danielle

Exactly what I was about to say

afftrackin · #24 (**permalink**) 12-19-2008, 05:44 PM

Also remember these results are based on approximate average for both the search network and the content network, so not all of that volume will be seeing your site listed in google, some will see it on web sites that are in the content network

freetraff · #25 (**permalink**) 12-21-2008, 09:47 AM

afftrackin, please have a close look at the screeshots provided

Danielle has:

- English
- All Countries and Territories
- Exact match

480

I have (I gave 2 screenshots, but one of them is)

- English
- All Countries and Territories
- Exact match

2,900

WTF?! Why do we have different stats under the same parameters?

And approximation from 480 to 2900 (under the same parameters) is a very lame approximation I assume? Google might have done a better job, because data is 6 times different!

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Overture Keyword Selector tool = Which search engine(s)?	littlegiant	Search Engine Optimization Forum	2	10-17-2005 06:39 AM
[Needed] PHP coder - Search Engine Keyword Ranking Tool	TheWebJunkie	Services for Sale/Hire	2	03-29-2005 04:09 AM
Fun With the Google AdWords Keyword Tool	WPW_Feedbot	Search Engine Optimization Forum	0	01-17-2005 08:00 PM
Google Keyword Tool	razsports	Google Discussion Forum	5	09-20-2004 07:40 PM
Google keyword suggestion tool?	bogdan_i	Google Discussion Forum	1	02-28-2004 06:09 AM