Keyword tools... what a disaster!

[Alex Sol]

So, I load up a keyword phrase into the Wordtracker Keyword tool and it shows me 100 monthly searches.

In SEO BOOK Keyword tool it shows 120 monthly searches - so similar to Wordtracker...

And then comes Adwords keyword tool. I make sure to select the [EXACT] search and it gives me... 15,000 monthly searches!

Now, 100 vs 15,000 - there is a big difference, so I really do not know who to trust. I really wish Google's tool is the one... but I really don't know.

[Steven Wagenheim]

Wordtracker's tool reports DAILY searches. That's why the numbers are so
low. You have to multiply by 30.

Personally, I don't use Wordtracker anymore. If Google says X number of
searches, since that's the SE 70% of the people use, I'm going to go by
what they say.

Ultimately, I doubt any keyword tool is 100% accurate, so pick one, get
used to how it works, get a feel for the numbers and eventually you'll get
a feel for the exact traffic. It takes a while, but it does happen.

[warriorprocessing]

I would think google would be the one. After all they have the actual data

[Alex Sol]

Originally Posted by Steven Wagenheim

Wordtracker's tool reports DAILY searches. That's why the numbers are so
low. You have to multiply by 30.

Personally, I don't use Wordtracker anymore. If Google says X number of
searches, since that's the SE 70% of the people use, I'm going to go by
what they say.

Ultimately, I doubt any keyword tool is 100% accurate, so pick one, get
used to how it works, get a feel for the numbers and eventually you'll get
a feel for the exact traffic. It takes a while, but it does happen.

Hey, Steven.

That is my exact problem - I did multiply by 30... and posted the multiplied results - that's what ticks me off.

[Steven Wagenheim]

Originally Posted by Alex Sol

Hey, Steven.

That is my exact problem - I did multiply by 30... and posted the multiplied results - that's what ticks me off.

Wordtracker also has a 90 day backlog. Google's may be more up to date.

Is this a seasonal niche or something new?

You are right though, those numbers are way off.

Yeah, keyword tools drive me batty too.

[Alex Sol]

Originally Posted by Steven Wagenheim

Wordtracker also has a 90 day backlog. Google's may be more up to date.

Is this a seasonal niche or something new?

You are right though, those numbers are way off.

Yeah, keyword tools drive me batty too.

Ok, here is the problem in detail.
The keyword phrase is "Montreal Car Rental". As you can see it is not VERY seasonal, sure there are more searches in the summer. However, there are about 2,000,000 people that live in Montreal plus all the tourists and visitors form around the world... this is why I'd like to believe that Google is giving me more accurate results... but still, I am confused.

[JJOrana]

Hi Alex.

My blog post may help you answer some questions in your mind. Check it here-> Keyword Tools Conspiracy Theory | Taknique

[ildarius]

These tools ARE a disaster, even Google's estimates are just what they are Estimates.

A recent example is a keyword for which I'm ranked on the 3rd page of the SERP, Google says it gets less than a 100 searches per month. I'm averaging 10-15 per day... (3rd page!)

[Alex Sol]

Originally Posted by ildarius

These tools ARE a disaster, even Google's estimates are just what they are Estimates.

A recent example is a keyword for which I'm ranked on the 3rd page of the SERP, Google says it gets less than a 100 searches per month. I'm averaging 10-15 per day... (3rd page!)

Wow...

So I guess the best keyword tool is logic and common sense... Just think what YOU would search for and go for that keyword phrase

Originally Posted by JJOrana

Hi Alex.

My blog post may help you answer some questions in your mind. Check it here-> Keyword Tools Conspiracy Theory | Taknique

Thanks, your post did clear up a few things.

[budfox]

keyword discovery is no better - throw that into the mix and you will get a third totally different figure (reported as a yearly number)

I am pretty much going with Google and dropping the rest of them

[Warrior Markets]

I guess you'd think that Google should be the one to follow, as they have the bulk of the actual results, but that seems almost too easy - otherwise there wouldn't even be a debate; Google should be THE authority.

However, when using the Google keyword tool, do you set it to 'exact'? If so, what can we gain from the 'broad' and 'phrase' match? I mean what do they tell us?

-WM

[Michael Oksa]

Why assume Google's are the most accurate?

I have a theory, but it is only a theory...I will let you connect the dots.

Does Google have anything to gain from inflating the search counts?

What type of people are going to even care about the search counts?

Does Google make money from AdWords?

Would people be more likely to target certain keywords if they thought the search count was higher?

Just asking.



All the best,
Michael

[JackieLee]

Using exact in the google tool is like putting quotes around your keywords in google. It gives you searches in that exact order. Broad will give you results that have your keywords anywhere on the page, not necessarily in the exact order you are focusing on.

[ildarius]

hold on, why put quotes around the keyword?

that would show you the amount of people that search for the keyword in quotes, no?
(literally people that would type a query surrounded by quotes in Google search box--> "keyword here")

[Heman Patel]

Originally Posted by Alex Sol

So, I load up a keyword phrase into the Wordtracker Keyword tool and it shows me 100 monthly searches.

In SEO BOOK Keyword tool it shows 120 monthly searches - so similar to Wordtracker...

And then comes Adwords keyword tool. I make sure to select the [EXACT] search and it gives me... 15,000 monthly searches!

Now, 100 vs 15,000 - there is a big difference, so I really do not know who to trust. I really wish Google's tool is the one... but I really don't know.

Try using the Traffic Estimator in your adwords account. I think that shows the best results but then again, its only limited to Google.com and its affiliate networks.

[SteveJohnson]

Originally Posted by ildarius

hold on, why put quotes around the keyword?

that would show you the amount of people that search for the keyword in quotes, no?
(literally people that would type a query surrounded by quotes in Google search box--> "keyword here")

put quotes around the keywords to search for the EXACT phrase within the quotes, otherwise the big G will search for each word separately and in combinations. has nothing to do with whether ppl used quotes in their search or not.

[Michael Oksa]

Originally Posted by ste25

From my research Google doesnt inflate search volumes. I always set to [exact] when researching keywords.

The MSN keyword tool is pretty good too.

They may not be inflating them at all, I'm just saying that they may have a reason to.

Out of curiosity, what kind of research are you doing to confirm these numbers? I think that would help all of us who are wondering how to find accurate search count results.

All the best,
Michael

[Alex Sol]

Originally Posted by Michael Oksa

They may not be inflating them at all, I'm just saying that they may have a reason to.

Out of curiosity, what kind of research are you doing to confirm these numbers? I think that would help all of us who are wondering how to find accurate search count results.

All the best,
Michael

Michael, that was my concern as well.
If ADWORDS tool tells me there are millions of searches for a certain niche+city yet there aren't many (or any) competitors in the "Sponsored results" I would sure be tempted to run ADS...

I guess the theory is, once again, that Google inflates the results to attract more PPC advertisers... Just a theory.

[kkchoon]

The search count is never accurate, even in Google...

Wordtracker uses a combination of many small search engines, that is a less than 5% searches in the market!

While Google is consider more accurate, but it combines the searches from its' partner, it's for PPC marketer, so you better take a lower number...

Do some PPC testing, you will find the result more accurate!

[Steven Wagenheim]

Look, let's be honest here. The ONLY people who know the REAL truth about
these keyword estimates are the people who are putting them out. Anything
we say is strictly guesswork.

So, because of that, this is how I run my keyword based advertising
campaigns.

1. I choose ONE tool to use and stick with it. It doesn't matter which one.

2. I will run it for a variety of keywords in various niches.

3. I will run my ads, articles, whatever, for each niche and monitor my
results.

4. I will then get an idea of the accuracy of the keyword tool itself and
be able to "guess" based on ratios, what I can expect to get as far as
traffic.

For example, I am going to give you the keyword estimates for 3 keywords
called X Y and Z. Sorry, not giving away the farm on this one.

Keyword X gets 247 daily searches according to Wordtracker. Running
a particular kind of article writing campaign utilizing blog review pages, I
get about 20 visitors a day.

Keyword Y gets 350 daily searches according to Wordtracker. Running
the same kind of campaign, I get 35 visitors a day.

Keyword Z gets 50 daily searches according to Wordtracker. It is more
longtail, at least according to what they're reporting. Well, I consider 50
a day longtail. Some might disagree. Point is, using the same campaign, I
get 10 visitors a day.

Now, if you take the percentages of visitors to daily searches, it's about
10% or the first two and 20% for the third one. These figures are fairly
consistent too.

And many other keywords perform the same way when using Wordtracker.

So, if I find a keyword at Wordtracker that gets 100 daily searches, I
can be fairly sure that I will get about 10 to 20 visitors a day from it, if
I run the same kind of campaign.

So, it doesn't matter if the actual numbers are inaccurate. The only
thing that will be off is the percentage of visitors to daily searches, but
my raw numbers, daily visitors, will be fairly predictable. What do I care
if Wordtracker is wrong and keyword X gets 500 daily searches. I'm still
going to get my 10% of Wordtracker's figure.

The point is, understand the tool that YOU use and you'll be able to get
a very good idea of how much traffic you're going to get.

Does this system take a while to taste accurately? Yes, but it's worth it.
Getting to know your keyword tool is gold.

[Marian Berghes]

another thing that I see alot lately is that people find keyword X that has 10-20k monthly searches and like 4-5k competition on google, they make a website..rank it on the first page but they get only 20-30 visitors a day or something like that. Then they start saying that the keyword tool is rubbish cus they did not get the 500 visitors or something close. In my short IM experience I did learn something that to me is very valuable:

1. Keep in mind that there are days/weeks when people search more for the term or search less
2. Even if you are #1 on google you still get only a fraction of the traffic...I don't have exact numbers but remember that on that page the user has 9 other websites to choose from.

[Warrior Markets]

Originally Posted by Michael Oksa

Why assume Google's are the most accurate?

I have a theory, but it is only a theory...I will let you connect the dots.

Does Google have anything to gain from inflating the search counts?

What type of people are going to even care about the search counts?

Does Google make money from AdWords?

Would people be more likely to target certain keywords if they thought the search count was higher?

Just asking.



All the best,
Michael

Michael, I agree; that's why I said that "you'd think that Google should be the one to follow", but for some reason it doesn't have a monopoly - why? I think you "answered" that in a way.

Originally Posted by Steven Wagenheim

Look, let's be honest here. The ONLY people who know the REAL truth about
these keyword estimates are the people who are putting them out. Anything
we say is strictly guesswork.

So, because of that, this is how I run my keyword based advertising
campaigns.

1. I choose ONE tool to use and stick with it. It doesn't matter which one.

2. I will run it for a variety of keywords in various niches.

3. I will run my ads, articles, whatever, for each niche and monitor my
results.

4. I will then get an idea of the accuracy of the keyword tool itself and
be able to "guess" based on ratios, what I can expect to get as far as
traffic.

For example, I am going to give you the keyword estimates for 3 keywords
called X Y and Z. Sorry, not giving away the farm on this one.

Keyword X gets 247 daily searches according to Wordtracker. Running
a particular kind of article writing campaign utilizing blog review pages, I
get about 20 visitors a day.

Keyword Y gets 350 daily searches according to Wordtracker. Running
the same kind of campaign, I get 35 visitors a day.

Keyword Z gets 50 daily searches according to Wordtracker. It is more
longtail, at least according to what they're reporting. Well, I consider 50
a day longtail. Some might disagree. Point is, using the same campaign, I
get 10 visitors a day.

Now, if you take the percentages of visitors to daily searches, it's about
10% or the first two and 20% for the third one. These figures are fairly
consistent too.

And many other keywords perform the same way when using Wordtracker.

So, if I find a keyword at Wordtracker that gets 100 daily searches, I
can be fairly sure that I will get about 10 to 20 visitors a day from it, if
I run the same kind of campaign.

So, it doesn't matter if the actual numbers are inaccurate. The only
thing that will be off is the percentage of visitors to daily searches, but
my raw numbers, daily visitors, will be fairly predictable. What do I care
if Wordtracker is wrong and keyword X gets 500 daily searches. I'm still
going to get my 10% of Wordtracker's figure.

The point is, understand the tool that YOU use and you'll be able to get
a very good idea of how much traffic you're going to get.

Does this system take a while to taste accurately? Yes, but it's worth it.
Getting to know your keyword tool is gold.

Steven, I totally agree with what you're saying. Irrespective of what the keyword tools show, the results are not affected. So if you make a site, and it gets 20 visitors a day, go back and see what your keyword tools showed. If one showed that you should've got 40 visits, that tells you that, in future, for estimates, use that tool, and then halve the figure.

The difficulty for people is finding that factor (the half in this example) - ie, the number with which to divide/multiply the results in order to get more accurate results. Like you say, with time you will become more proficient in "tasting accurately".

-WM

[ptone]

Originally Posted by Alex Sol

I guess the theory is, once again, that Google inflates the results to attract more PPC advertisers... Just a theory.

I'm going to go out on a limb here and say that this is NOT happening. No matter how much you or I dislike Google, they are not stupid. This would be fraudulent and enough complaints could bring the feds knocking at their door.

Furthermore, if this were true and an investigation were to occur, documents subpoenaed in court could reveal their secrets they hold so dearly to their company.

Adwords is the core of their multi-billion dollar empire and I just can't see Google risking everything to add a few more PPC advertisers to the margins.

Lastly, your comments could be construed as libel which Google might take very seriously. Now, they aren't going to go after you for your comments here at the WF, but I certainly wouldn't start a blog or write a ton of articles explaining your thoughts on this. Personally, I don't care if you do or not, I just don't want to see you get in trouble over this as it would not be worth it...that's why I mention it.

[Li Weng]

If anything I'd trust google, since they're pulling data from their own search engines. It's not 100% accurate only because google calculates averages over a period of time. Other keyword tools obviously get data from google too.

[Jakehyten]

I used to have the worst problems with that. I think seo book gives the best results. Becauuse my clicks match up with the traffic they say.

[seobro]

Guys, I have used Google Adsense and Adwords for years now and the adwords results are accurate, but the keyword tool is way off, like 2000/mo searches average for a keyword where I am the only company making it and get less than 100 clicks/mo. Not off, way off...

[jan roos]

I found a keyword phrase on wordtracker a few weeks ago, Optrimized a blog for it got it to page one and had zero visits from that phrase. Makes me not trust word tracker anymore.

I'll try google adwords next.

Cheers

[Kevin McNally]

The most accurate way to find out keyword volume is to run an adwords campaign using "exact match" keyword and let it run for a while to get a daily average.

No keyword tool is 100% accuate but you can use them in different ways, I often use wordtracker to compile a list of keywords with competition data and then head over to adwords to compare estimated demand.

I was just checking my stats today for one of my sites

Adwords exact match- Average search volume = 6,600 approx
Wordtracker 24 hr google predict = 15 per day

The exact keyword last month was found over 2900 times via the search engines and 90% of this would be via google. The site is ranked at number 1 for this keyword so I am guessing it has 50% CTR via an exact match search so that makes the google data very accurate in this case.

I have other sites where wordtracker is more accurate but the best method is still running an adwords campaign to quickly get the real demand when you are on the first page.

[Kirk Paulhamus]

I would go with Google it just makes sence to me that there info would be the most accurate sence they are the A#1 Search Engine.

[Wendy Maki]

Funny, noone's actually factored in the quality of the page or the way the page ends up listed on the SERPS.

It really doesn't matter how many searches there are for a keyword if the page (as it appears on the SERPS anyways) either seems irrelevant or like junk. Also, where does the page rank in those SERPS?

Just my 2 cents

[Jason Moffatt]

Originally Posted by Michael Oksa

Why assume Google's are the most accurate?

I have a theory, but it is only a theory...I will let you connect the dots.

Does Google have anything to gain from inflating the search counts?

What type of people are going to even care about the search counts?

Does Google make money from AdWords?

Would people be more likely to target certain keywords if they thought the search count was higher?

Just asking.



All the best,
Michael

I wouldn't put much belief into this theory as it really doesn't match Google's model. They are not in the business of duping people or tricking them. If anything, their model is to provide people with the most accurate info as possible.

This is why they implemented quality scores with adwords and decided to quit taking money from people who were buying ads that were not relevant.

Of course I wouldn't trust the data 100%, but I definitely wouldn't think they were consciously trying to skew any results for monetary gain. They make their money by providing the most relevant and up to date info they can, and they strive to continue on by that model.

That's my theory.

[Michael Oksa]

Originally Posted by ste25

Seobook is a complete joke! Honestly the numbers are so far out. It's ok for long tail research but forget the numbers. With some keywords they might match up but with some they're so far out its just ridiculous.

Heres just one example (I could post 1000s) of what I mean. Take the keyword phrase:

business accounting software

Google tool: 165,000 searches month average [exact match]
Seobook: Predicts 8 searches per day on Google

It's a joke!

But how do you KNOW that Google ISN'T the joke? What I want to know is what methods you are using to arrive at your conclusion. You may be right, but how do you know for sure that SEO Book isn't right and that Google is the one that's way off?

Just curious, because the answer to that question could add a lot of value to this discussion.

All the best,
Michael Oksa

[Gary McCaffrey]

I have a site that is #2 in google for a term seobook says gets 3000+ google searches per day.

Its bringing in about 5 visitors per day!

[JJOrana]

This is funny as a a lot of people suggesting using that damn tool. The same tool that made me create my own keyword research system because seobook tool is flawed.

People are in confusion because they are mesmerized with the features these tools are giving, without really researching where these tools get their data.

- Jon

Originally Posted by Gary McCaffrey

I have a site that is #2 in google for a term seobook says gets 3000+ google searches per day.

Its bringing in about 5 visitors per day!

[Michael Oksa]

Originally Posted by Jason Moffatt

I wouldn't put much belief into this theory as it really doesn't match Google's model. They are not in the business of duping people or tricking them. If anything, their model is to provide people with the most accurate info as possible.

This is why they implemented quality scores with adwords and decided to quit taking money from people who were buying ads that were not relevant.

Of course I wouldn't trust the data 100%, but I definitely wouldn't think they were consciously trying to skew any results for monetary gain. They make their money by providing the most relevant and up to date info they can, and they strive to continue on by that model.

That's my theory.

I like your theory better, at least it is more sound than mine.

It makes sense that Google's main purpose is relevancy...and purposely skewing their numbers would alter that relevancy.

Thanks for the different angle.

All the best,
Michael

[Michael Oksa]

Originally Posted by ste25

Because Ive compared SEObook's keyword tool data and data from the Google keyword tool and Google gives fairly accurate estimates whilst SEObook doesnt.

When I look at the number of visitors I get coming to my site for a certain keyword phrase its about what I should be getting according to Google.

Having said that some (emphasis on the word "some") keyword phrases are about right with SEObook (ie if SEObook predicts 200 searches per day in Google for a keyword and I run and Adwords ad (using exact match) I get around 200 searches whilst other keyword phrases are so far out its just ridiculous, predicting several thousand searhces per day when in fact there are only a handful and then doing the exact opposite and saying there are 5 or 6 searches per day when in fact there are 1000s daily.

The only useful thing SEObook is for is getting ideas for long tail keywords. I dont trust the search volumes for every keyword phrase.

EDIT: What I like to do is compare Google's data against SEObook's data. If Google says a keyword gets an average of 6000 searches per month [using exact] and SEObook predicts 200 per day on Google then the keyword phrase does get around 6000 searches per month.

So, SEO Book isn't all bad, and Google isn't all good.

Thanks for explaining, I think that helps clarify it nicely.

All the best,
Michael

[Kurt]

Originally Posted by Steven Wagenheim

Look, let's be honest here. The ONLY people who know the REAL truth about
these keyword estimates are the people who are putting them out. Anything
we say is strictly guesswork.

Originally Posted by Jason Moffatt

I wouldn't put much belief into this theory as it really doesn't match Google's model. They are not in the business of duping people or tricking them. If anything, their model is to provide people with the most accurate info as possible.

Hey Steven, Jason and All...

Here's the problem with the above theories...I don't believe even Google knows the real numbers.

My theory (and it's more fact than theory) is that there's a number of hard core SEOers that have tools designed to skew the results. It's very easy to query Google for certain search phrases to artificially inflate search stats, and I personally know SEOers that do exactly this.

If you are in one of their niches, your search numbers are just about meaningless. They inflate numbers for keywords that are NOT their best, to draw attention away from these phrase.

They are already making money in these niches, so they don't need these stats.

I've posted many times that EXTREME SEO means you go after ALL keywords that can bring traffic. Now, you can use your own stats to see which keywords you should focus on in the future.

With this strategy, you toss out keywords with very low search counts, as:
1. Zero and very few searches can't be faked by SEOers.
2. Therefore, you can be pretty sure that these keywords truly don't drive much traffiic.

Take the other keywords and pound away at them with content, links and articles. Then let your own stats tell you what's best.

Use any keyword research tool as a <very> loose, general guide, good only for getting you started. You do have to start somewhere, use the research tools for this, and only this. But they are far from being gospel.


My point being, with the hard core SEOers skewing the results, I doubt even Google really knows what's going on. As keyword research becomes less and less accurate, you should put less and less time into keyword research, choosing to build resources instead. These resources will provide your research in the future.