Wednesday, October 28, 2009

Google Trends for Websites - do you trust it?

Google Trends for Website - competitive intelligence for freeGoogle Trends for Websites is one of the free competitive intelligence services provided by Google. It is pretty easy and straightforward to use: just type in domain names of sites you want to compare and it gives you traffic trends over time (for the selected period) and some additional data. Unlike Compete free service, Google Trends for Websites offers worldwide coverage with the ability to segment results per country. Cool, isn't it?
Compare your site traffic vs. competitors
Simple but valuable
While simple, it is a very interesting service as you can compare traffic trends of your site vs. your competitors. You can learn a lot like the impact of competitor advertising campaigns (were they successful?), product launches, announcements... You also get the key search terms used by visitors and the sites they also visited (and find out who are your biggest rivals). Super-cool isn't it?

See what sites your visitors also visited
(Note: for competitive search terms analysis, I strongly recommend Google Insights for search, another cool free Google service. See the practical example I did earlier this year)

For me the main limitation is the fact that data are only available for sites with traffic above a certain threshold (Google gives no details on this limit). Practically, it means that it is only usable for big sites and big countries.

Still, even if the provided information is limited, it can be very valuable and, hell, it is FREE (and ones knows that it is important nowadays :-)).

For more examples or details on Google Trends for Websites, read Avisnash's excellent post on that topic.

"But are Google data reliable?"
This is the question that usually comes when I show Google Trends for Websites. For quite some times, Alexa is offering similar service but it has received a lot of critics regarding its accuracy and reliability. Alexa collects data mainly from the Alexa toolbar - not the most used toolbar to be honest. The consequence is that the sample is rather limited (from a size & geographic perspective) and therefore it is quite biased in my opinion.

For Google it is quite different. Google has access to a huuuuuuuge amount of information from a variety of sources, such as aggregated Google search data, aggregated anonymous Google toolbar /Google Analytics data, consumer panel data, and other third-party market research. As Google claims in Google Trends for Websites help section - the data is aggregated over millions of users. So that is how Google can get information on your website traffic - even if you don't measure your site activity with Google Analytics.

Ok sounds good but still is it reliable? To answer the question in a simple way, I decided to do a simple test: compare our own data against Google Trends for websites data. I took the traffic trends for 2009, for 2 markets with high traffic volume, no Google Analytics on it! I put everything in Excel, I adapted the scale to get as close as possible to the scale of the Google tool and this is what I get (without any data tweaking, I swear!):
Own data vs. Google Trends data - test 1wn data vs. Google Trends data - test 2
I think the result speaks for itself (a picture is always better than long explanations :-)). Spooky, isn't it?

Maybe I was lucky with my 2 examples but for me it is a simple way to show that Google data is certainly not bad quality and that somehow I can trust it. Therefore I usually propose it as a data source for benchmarking in appropriate cases - taking into accounts its limitations.

I would be really curious to know if anyone did similar comparisons and what was the outcome. Did it work out for you or Google Trends for Websites results were miles away from your own data? What do you think about this Google service? Like it or not? Please share your experience.

Cheers!

(A little side note: surprisingly Google Trends for Websites can give you trends for Yahoo.com or Bing.com but not for... Google.com (see here). Doh!)

Related posts & resources:

7 comments:

  1. Wow Michael! Is this the post you were alluring to on Twitter yesterday? If so, I can guarantee you this post will be a huge hit! :)

    This is a very interesting study, which reinforce my love/hate relationship with Google. Great to have a free service like Google Trends, but also bye-bye to keeping such valuable competitive information to ourselves...

    Would be nice to see how Alexa ranks against Google Trends and your own data :)

    Stéphane

    ReplyDelete
  2. Hi Stephane,

    Thanks for commenting. Indeed it is quite impressive to see what Google can do.

    For example if you look at the stats from www.toyota.de. You will see in the top visited websites: toyota-europe.com. Does that means that most German visitors go to our European portal site? Not really - reason is that many German visitors on our site uses the Car Configurator that is a centrally hosted tool, hosted on toyota-europe.com domain...

    Regarding comparison with Alexa, it is a very interesting idea but difficult to do as Alexa stats are based on % of global activity (i.e. vs. all the rest of Internet) and segmentation per country is rather limited :-/

    Cheers,

    Michael

    ReplyDelete
  3. Yeahh kinda spooky. It's just the data pool that Google controls. I love insights and google gives them to me. So rock on...

    ReplyDelete
  4. Interesting! I always wanted to do this comparison. I just checked it for a big website (> 100.000 visits a day).

    Google Trends for this site is not as accurate as it was for the sites you showed in your examples.

    Maybe you have lots of traffic from Google on those sites.
    The higher the percentage of traffic from Google is, the better the estimate.

    ReplyDelete
  5. @Peter: thanks for sharing your experience. But when you say not as "accurate as my examples" - do you still see similarities for main trends & movements such peaks/drops?

    To answer your comment, both get more than 25% of traffic from search - one getting significantly much more than the other :-)

    Don't know what you have in your case?

    Cheers,

    Michael

    ReplyDelete
  6. Interesting !
    For smaller websites, it would be interesting to develop the same kind of analysis with Ad Planner. I have talked to other webmasters and it seems that Google public figures are really close to the ones with get (through Google analytics).
    Can we have an idea of the amount of trafic you were referring to in the 2 examples ?

    ReplyDelete
  7. @TribuGourmande.com:
    Thanks for sharing your input and what you have heard from others. It is true that it is frustrating for small sites or markets not having such data.

    Regarding your question, I can tell you that both sites are above 250,000 visits per month.

    Michael

    ReplyDelete

Feedback Form