how accurate is google keyword tool?

Google keyword tool is very good to use, but still you need to check for the competiton on keywords.

But, I believe in using this keyword tool as it is worth to use.

Thanks

Comparing with all the other tools, I find Google tool is the most accurate but I am not getting anything close to the search volume even for my top ranking sites...

Agree, in some cases wildly inaccurate. I recently found a longtail keyword showing 60K results in exact match in US only. I had my doubts but wanted to test this out, so I sat up a website,filled it with content and added high pr backlinks. In few weeks I hit the #2 spot and are enjoying 2-5 visitors/day.
Why GKT is showing these numbers I have no clue.I checked today and they still do.

Hi kenj & claytons,


I don't think your are correct in implying that the AdWords Keyword Tool's data is inaccurate. Perhaps you do not have a clear understanding of the data that is being presented.

The AdWords Keyword Tool reports the number of searches, not the number of click throughs. Not all searches result in a click through. Often folks will search with a term that is too ambiguous and they will then follow that with a term that is more concise.

So, certain keywords will have dramatically lower CTR than other highly specific terms. But even when a search is performed that returns highly relevant results, the user must take that next step of clicking a listing to result in traffic to those pages listed. People are often fickle and change their minds, get interrupted or distracted before they click through.

The fact is that nearly half of all searches never result in a click through. The average is around 46% abandonment for all search queries.

Another thing to take into consideration is that the tool shows you historical data, which often varies a great deal from future search traffic. If you look at the trend graph on your keywords you will notice that there are frequent spikes and dips for many search terms. What happen last month, or even the last 12 months may not always be indicative of what you will see next month.

Perhaps the data is accurate and you have not interpreted the data correctly? If you want to be certain, you can always run an ad campaign with your budget wide open and your bids set high enough to always display impressions. There really isn't any other way to measure search traffic.

The tool is a bit hit and miss to be honest.

I have a site ranked 4-5 (in all countries where the searches come from) for a 27,000 / month exact search keyword that gets 2-5 hits a day from that keyword. Trends indicate

Then I have a site ranked 6 for a 8000 / month exact search keyword that gets 40-50 hits from that keyword a day.

I don't trust the tool at all, but I still go by it when looking for niches as the other options are worse.

Yes it is quite accurate, although it does not show you the long tail keyword, those that are low in searches.

I would always use it first to check if there are good traffic for the niche I am going into.

Then use Keyword Research Pro: Keyword Research Tool to find the relevant long tail keyword to target the niche, so I can get organic search with good free traffic.

Hi claytons,

I'm sorry I cannot agree, you cannot measure searches by getting your page to #1. You are assuming a 40% CTR when the reality could be anywhere from 1% to 90% based on a number of factors. And how are you measuring search abandonment rates for your targeted term? You can't because you cant see that from your website stats.

The only way to measure that number is to run an ad campaign that displays on every query and measure your impressions.

You say it's overstated, yet I have ran many campaigns with many thousands of keywords and have found it to be highly reliable data. Could it be a case of misinterpreting data?

Hi claytons,

I think we mostly agree, we at least agree that an adwords campaign can be used to verify. I just don't think that you can accurately project traffic based on the data since there are many other factors that will vary a great deal from keyword to keyword and from page listing to page listing.

In case I wasn't clear in where my disagreement was, please let me clarify.

When you do a lot of AdWords advertising on the Search Network you quickly realize that CTRs will vary a great deal based on the individual keyword. This means you can't rely on an overall average since most keywords will range way outside the average. On top of those variations you may also get a wide range of CTR based upon the ad text. Those same wide ranging variations can and will be seen for organic listings based on Page Title, Meta description and the wide ranging CTRs for individual keywords.

It just doesn't seem fair to blame the tool that reports the data when it's much more likely to be another reason for the disparity. Again the tool doesn't report traffic, just searches. If those searches do not result in a click through there is no traffic associated with that search. Since the difference between search queries and traffic will vary a great deal based on the above mentioned factors, you simply cannot use your traffic level to measure the accuracy of the keyword tool's data.

The data could be 100% accurate and you may see a wide range of traffic based on these various factors affecting CTR. Even if the data is inaccurate using your traffic level to determine that would be a highly unreliable method at best.

Hi claytons,

Sorry to come across as condescending, please accept my apology.

I can see that I fell short again in making my point. I meant to use the AdWords information to point out the parallel with organic listings. Similar to AdWords, your traffic from organic listings will only come after a click on your listing. I meant to point out that your CTR will vary for your organic listings and unlike your adWords campaigns you do not have access to impressions data so there is no mechanism to measure impressions from organic traffic.

In an earlier post in this thread you said "getting to #1 and seeing if you get somewhere around 40% of that traffic" and this where I specifically disagreed with you.

The best data I have been able to find (leaked AOL data) that shows CTRs for organic traffic indicates, on average, the #1 position gets a 23% CTR. While that represents about 40% of the click through traffic, it is still only 23% of the total searches. This is because about half of all searches are abandoned without a click through.

Again, we are talking averages here. There seems to be a very wide range of CTRs for organic traffic as well and this makes the use of organic traffic a highly unreliable mechanism to measure the accuracy of the Keyword Tool's data.

Again, I'm sorry about the confusion, I hope I have made it clear that it's the organic traffic that is unreliable not AdWords impressions. I agree with you that it's impressions that are reasonable measure of search volume, and I stand in disagreement about the use of organic traffic as a useful measure to determine keyword data accuracy.

Hi claytons,

Sorry to come across as condescending, please accept my apology.

I can see that I fell short again in making my point. I meant to use the AdWords information to point out the parallel with organic listings. Similar to AdWords, your traffic from organic listings will only come after a click on your listing. I meant to point out that your CTR will vary for your organic listings and unlike your adWords campaigns you do not have access to impressions data so there is no mechanism to measure impressions from organic traffic.

In an earlier post in this thread you said "getting to #1 and seeing if you get somewhere around 40% of that traffic" and this where I specifically disagreed with you.

The best data I have been able to find (leaked AOL data) that shows CTRs for organic traffic indicates, on average, the #1 position gets a 23% CTR. While that represents about 40% of the click through traffic, it is still only 23% of the total searches. This is because about half of all searches are abandoned without a click through.

Again, we are talking averages here. There seems to be a very wide range of CTRs for organic traffic as well and this makes the use of organic traffic a highly unreliable mechanism to measure the accuracy of the Keyword Tool's data.

Again, I'm sorry about the confusion, I hope I have made it clear that it's the organic traffic, in my opinion, that is unreliable not AdWords impressions. I agree with you that it's impressions that are a reasonable measure of search volume, and I stand in disagreement about the use of organic traffic as a useful measure to determine keyword data accuracy.