Wednesday, December 10, 2014

GNU/Linux and Google Trends


I noticed something interesting looking at Google Trends -- it looks like GNU/Linux searches are decreasing over time:



From:
http://www.google.com/trends/explore#q=%2Fm%2F03x5qm%2C%20%2Fm%2F0fpzzp%2C%20%2Fm%2F02wxtgw&cmpt=q


Though this strikes me as odd... since as of 2014, GNU/Linux probably holds the market shares (roughly):

  1. desktops:  1.5%
  2. servers:  40%
  3. smart phones/tablets: 84%
  4. super computers:  97%
From:

Also, most of these market shares are steady or gradually increasing.  Even on the desktop -- the smallest usage -- the market share has slowly inched up from 0.5% over the past ten years.  If there are 7 billion people on Earth, and about 5% own a computer,  that would mean roughly 5 million people are running Linux on the desktop now.  So why would Google trends be decreasing?  This seems to contradict all the other trends I've seen.

My guesses (based on my own search patterns)

1.  The numbers mean something different than what might be expected.  According to Google Trends, "A downward trending line means that a search term's popularity is decreasing. It doesn't mean that the absolute, or total, number of searches for that term is decreasing." So plausibly, now there are more people online searching for other things.  

2. The need for searching has changed.  About 15 years ago, when I started using Linux, I'd have to make a lot of continuous tweaks to get everything working the way I wanted.  Almost every installation resulted in multiple hardware or configuration problems.  Now, I install Linux and everything "just works" for me.  So I do very few searches now compared to ten years ago.  

3. General search terms are not needed.  I don't usually search for a broad term like 'linux' anymore.   I don't need to include it.  Usually I'll search for something much more specific.   

4. The web itself has become more specialized.  I'm more likely to check specific websites (like stackexchange, or distro forums) for answers rather than searching over the entire interwebs.   

So, if I restrict the trends to look at just YouTube stats, the trends look a bit more like what I'd expect from other stats -- a stable or slightly increasing line:


So, using  Google Trends (on the entire web) is not  a very accurate measure of overall interest or market share of technology.   Generally the numbers don't mean what might be assumed, since the search terms are competing against all popular searches.  For example: Linux vs Pop Celebrity of the Week or the latest Playstation Games :-).  

What might make the numbers more useful is to use one term as a base line, and normalize all other numbers relative to it.   For example, adding Windows as a baseline, and restricting the category to "Software > Operating Systems," it generally looks like searches are becoming less "techie" over time.  



Overall then, Ubuntu seems to have matched Linux in general search interest.  Also the gap between Windows and Linux is generally decreasing.  Which makes more sense.

In summary, Google Trends can offer some information, but it's fairly easy to misinterpret.  For example,  "android" has spiked in searches (see first graph), but "android" is really just a more specific type of "linux" search.  

No comments: