A while back I stumbled on a new part of the Google website called “Google Trends”. You can view this website for yourself here.

In Google’s words, the website allows users to “compare the world’s interest in your favorite topics. Enter up to five topics and see how often they’ve been searched for on Google over time. Google Trends also displays how frequently your topics have appeared in Google News stories, and which geographic regions have searched for them most often.” Basically it’s a service that allows you to see the number of searches on various keywords over time, broken out by geography, language, etc. All the values are normalized relative to the total number of searches – while that allows meaningful comparisons over time, it does remove the user’s ability to get at raw data, which might be moreuseful from a data analysis perspective. But it’s a free service, so I’m not complaining.

For example, the this link shows search patterns on the term “Six Sigma”.

I’ll freely admit I was blown away by the distribution with respect to geography and language, and I suspect others in our profession in North America might be as well. For those of you outside of North America, this is probably old news. I wasn’t too surprised to see India on top in terms of searches by country, but I certainly wouldn’t have predicted that Thai language searches for “Six Sigma” outnumber those in any other language. I and definitely wouldn’t have predicted that the USA would rank a mere seventh! (Admittedly Canada, where I live, doesn’t even rank in the top ten.) As with many things, it’s one thing to know what’s going on anecdotally, but quite another to be confronted with the data.

You can also use the site to compare searches on two or more terms, which lends itself well to all sorts of correlation/causation games. If you’ve ever been in a Six Sigma training course, you’ve probably heard some variation on the notion that “correlation does not necessarily indicate causation” (although it can certainly be a big hint!). One popular example of this fallacy is that sharks must like the taste of sunscreen because sales of sunscreen correlate well with the occurrence of shark attacks. The correlation turns out to be true, and apparently interest in sunscreen is not a bad predictor of reports of shark attacks. See for yourself here.

On a less frivolous note, this resultshows searches on the term “Six Sigma” compared to “Lean Sigma”. It’s worth playing around with the “regions” and “years” drop down menu to see what has happened in various parts of the world over time. The trend for “Lean Sigma” itself is actually very interesting too – herewe see Washington, DC on top, which I theorize may reflect the recent interest in “Lean Sigma” by military and government agencies in the USA.

Anyway, there’s lots there to play with, and it puts some data behind what had been largely anecdotal up to this point. Of course, any data-hound worth their salt will already be wondering where the data is coming from, what it really means, how it was gathered, what sort of processing done, etc, etc. These are all valid concerns, especially since Google takes the time to point out the following:

“Google Trends aims to provide insights into broad search patterns. As a Google Labs product, it is still in the early stages of development. Also, it is based upon just a portion of our searches, and several approximations are used when computing your results. Please keep this in mind when using it.”

Elsewhere they go into a little more detail:

“Google Trends is a Google Labs product, which means that it’s still in an early stage of development. The data Google Trends produces may contain inaccuracies for a number of reasons, including data-sampling issues and a variety of approximations that Trends makes use of. We hope you find this service interesting and entertaining, but you probably don’t want to write your PhD dissertation based on this information. We’ll aim to update the information provided by Google Trends monthly.”

So have fun, but consider yourself warned! If you are interested in Google trends, you can find out more about it here.