30DC: Is The GTrends Technique Really Reliable?

Someone asked on the 30DC forum about whether the Google Trends technique is a reliable source of determining Google Search traffic. This made me think that it's possible that maybe Dan's method may not be so accurate after all.

Here is the argument I made on the forum:

Quote:

Originally Posted by learningcurve Hi everyone, If you have Google Analytics set up for your site, you can track ACTUAL VISITS to your site by KEYWORD...so you can, in fact, determine how much traffic you're receiving per keyword per day (or any other period of time, for that matter.) This is how you'll be able to determine actual traffic numbers for your own keywords and replace MYI with your own phrase as your benchmark. So, if Dan says he's getting 500 visitors a day on the keyword "male yeast infection" - I'm pretty sure he knows that for sure. Suzanne

Google Analytics is a great tool and you can check the ACTUAL VISITS to your SITE by KEYWORD. For those who don't know what Google Analytics is, Check out the product tour on their website:

http://www.google.com/analytics/

Suzanne you are correct about being able to see the actual visits for your site. However, your conclusion is incomplete.

THEORETICAL EXAMPLE
Let's say I own a domain called www.mynicheproducts.com. I set up Google Analytics and rank #1 in the natural search of Google.com. I also heavily optimize my site for the keyword "sausage making in my basement".

For the sake argument, let's say that 100% of the time that someone types in "sausage making in my basement" in Google, they ALWAYS click on www.mynicheproducts.com. Let's say that Google Analytics reports that I have 500 clicks per day for "sausage making in my basement". Therefore, it would probably be safe to assume that I can go into Google Trends and just match the numbers with what I have on Google Analytics. Also, the line graph in Google Trends should be the same graph in Google Analytics.

REALITY
The problem is that for the #1 rank you may not get 100% of the clicks every time someone searches for "sausage making my basement". You may get 40%... 60%-80%... who knows... Marketing is almost an art because the science of consumer psychology are not always exact numbers.

A Guess at Dan's Methodology
1.) Take the average of the #1 or #2 spot (i.e. 40% of searches are click throughs) <<< This is where the problem is, how do you determine the average?
2.) Write the average of daily clicks over a week or 2 week period (i.e. 300 clicks daily)
3.) GTrends Baseline = (1 - [% of searches]) * [average daily clicks]

CONCLUSION
I really do like the technique, however, I think it's important to understand how reliable this technique is. The issue lies is how to determine the baseline of a Google Trends line graph.

RESOURCES
Here's an article that may help define the average of different ranking of sites:
Click Through Rate of Google Search Results - AOL-data.tgz - Want to Know How Many Clicks The #1 Google Position Gets?

So what is the Clickthrough Rate of the #1 and #2 spots on Google.com?
The following image is from Redcardinal.ie:



It gives an interesting statistic:

• The #1 ranked site receives ~42% clickthroughs of search results
• The #2 ranked site receives ~11% clickthroughs of search results

This means that the #1 and #2 spots receive HALF of the search results for a keyword. I don't know how this data was generated but it could be representative of the search results we see on Google.

30DC Criteria

Since our criteria for a keyword is at least 80-100+ searches per day, we should have the following:

• If we had the #1 spot for a keyword, we should be getting about 40-50 clicks per day
  
• If we had the #2 spot for a keyword, we should be getting about 8-10 clicks per day