Who reads this blog, anyway?

Who are you?

When National Geographic branded Science Blogs, they also deleted our big giant collective Google
Analytics Account and gave us each individual Google Analytics accounts. The bad news is that I then lost the ability to analyze use of my blog over the last few years, which was only of marginal value anyway, but gained a better way of accessing data. Having each individual blog be on its own account makes it much easier to use G.A.

Which Operating System Is Most Common?

Now that it has been a few months, I’d like to review some of the data, which I warn you are not shocking or anything, in who you’all are, or in some cases, what you’all seem to prefer with regards to technology. Let’s start out with Operating Systems running on devices used to access this site. Changes in this number can reflect changes in behavior of individuals, such as a shift towards using a hand held device for a period of time, instead of a desktop. Or, they can reflect changes in what people generally own or use. Or, they could reflect changes in the readership of this blog. The shifts from June to August shown in the following table are very small and clearly are not statistically significant. (When I look at operating system data for the entire period up to today, half way through September, the relative position of Windows, Mac, iOs, Linux and Android seem to alternate or at least shift around randomly.)

June August
Windows (56%) Windows (55%)
Macintosh (16%) iOS (16%)
iOS (13%) Macintosh (15%)
Linux (8%) Android (7%)
Android (6%) Linux (6%)

The bottom line is that most of you are accessing this site via a Windows machine, with Apple operating systems holding a strong second place, and Linux-based systems the usual third place. So, these numbers should be regarded as two samples that show the same thing.

More interesting is comparing operating system use from a few years back to the present. Digging back into the past I found some comparative data from July 4th, 2008.

OS 2008 Internet 2008 Scienceblogs 2008 This Blog 2012 This Blog
Windows 91% 78% 76% 55%
Apple (all) 8% 17% 19% 30%
Linux (all) 1% 5% 5% 13%

Here, you have two trends. One is less to more smart (the world at large vs. science blog readers) and the second is across time (2008 to the present). The uptick in Linux use is entirely due to the addition of Android devices; desktop Linux is holding steady at about 1% for everybody, 500% greater than that for us. Similarly, the increase in Apple use is mainly due to hand held devices.

In other words, Windows does not do servers well, and it does not do hand held devices well, and the one area it kinda works is becoming a smaller relative part of the regular end-user market. One thing you’all have told me in the past (some of you, anyway) is that you use Windows at work because you are forced to but Linux or a Mac at home. I’d bet that is still true. I would love to look at OS against time of day to see if this shows up as a pattern, but I’m not sure I can with G.A.

Which Browser Is Preferred and How Has This Changed Over Time?

Regarding Browser use, there seems to be something of a trend over the last few months, and the numbers are a little impressive, and in my view, expected.

Browser June August
Firefox 32% 25%
Chrome 28% 28%
Safari 13% 20%
IE 12% 14%

A shift towards Safari or a shift away from Firefox could explain the drop in use of Firefox. In any event, the once “best” browser because it was Open Source and always seemed to work better than Internet Explorer has finished its course (of beating up on IE) and has now become the browser, in my experience, that usually hangs, usually does not work, usually annoys me. All that annoyance for me is on Linux, as I don’t use Windows and have not bothered to install Firefox on the iMac (why would I? is it even possible?). Because of rounding you can’t see that Chrome has actually gone up a whole percentage point across this time period, which probably means nothing.

Again, digging out some old data, I can construct the following table:

Browser Use on This Blog

Browser December 2010 December 2011 August 2012
Firefox 42% 32% 25%
Chrome 17% 26% 28%
IE 22% 20% 14%
Safari >1% 13% 20%

Just so you know, I use Chrome and Safari on my iMac and Chrome on my Linux Desktop and Laptop. Chrome does a great job keeping things synced between browsers and lets me use Chrome apps for things like Google Tasks (more about that later), while Safari is great for accessing Apple related sites. Also, I have two distinct Googly Identities (e.g., G.A. for The X Blog vs. this blog) and I keep one on Safari and one on Chrome. So far this has not been as clumsy as it might seem. My next project will be to figure out how to make Safari on the iMac and Safari on the iPad sync. They probably already are syncing because that is the kind of thing Apple would automate, right?

Moving on to the people rather than the machines…

What country are you from?

Most of my readers are from the US, but a fair number are not, which is something I truly appreciate. Americans, as you all know, suck, so it is nice to have some friends who are not Americans (though some of you suck too, I’m sure). In the past I’ve had more African readers than I do at the moment, based on my recollection (can’t find any old G.A. navel gazing posts) and that is probably due to the fact that I intentionally posted African (mainly Congo-related) news for a period of time to raise awareness of what was going on there, plus, of course, the widely read Congo Memoirs and Lost Congo Memoirs. I hope to return to some very interesting Africa-related topics soon, so that may shift.

When I look at trends over the last few months, I see no meaningful shift over time, so I’ll just give you the data running from June 1st to the present. First, let me list the countries that have no representation according to G.A. among the readership of this blog:

  • French Guiana
  • Western Sahara
  • Mauritania
  • Niger
  • Chad
  • Central African Republic
  • South Sudan
  • (French/Brazzaville) Congo
  • Turkmenistan

There are a few things that tie this list together. One is langauge; Even though this blog is read in many countries that are not mainly English, these are smallish countries with small populations and mainly non-English speakers. In some cases, the countries probably either have no effective Internet or the Internet activity that occurs there actually routes though a neighboring country (I’m thinking this may be the case, for example, with Western Sahara).

I must say that I will be very disappointed if I don’t eventually pick up some traffic from South Sudan. I’ve never been there, but the region is very important to me. It is historically where many of the ancestors of the folks I’ve lived and worked with in the Congo come from. Also, I remember hearing the secret night flights overhead when relief supplies were being dropped there in the 1980s. I hope this newest African country gets its feet on the ground, and I really want to go there and check it out.

Anyway, enough of that. What about countries that do have readers of this blog? Here’s a simplified table:

Country %
United States 64%
Canada 8%
United Kingdom 8%
Australia 3%
Germany 1%
India 1%
Philippines 1%
Ireland 1%
New Zealand 1%
Netherlands 1%
France 1%

DR Congo is 141th and South Africa is 18th. Just so you know.

It is also worth mentioning that within the United States, readership more or less follows population size, according to my eyeballs and not referring to any actual data. California, New York and TExas are the top three states. Minnesota is 8th which I think brings it above expected rank based only on population.

That’s all for now. We’ll have another look in a few months.


  1. #1 Markk
    September 15, 2012

    If Google Analytics hasn’t fixed the issue with Google Reader then everybody who reads your site with Google Reader won’t be counted in your numbers. Do you know if that has been fixed? I have to believe that is a non-trivial number.

    By the way, it is funny, I could replace Firefox with Chrome and vice versa in your story about your experience and it would describe me. We have exact opposite experiences with browsers. I’ve left Chrome behind now. It is so good that there is a choice.

  2. #2 Greg Laden
    September 15, 2012

    On what operating system?

  3. #3 Greg Laden
    September 15, 2012

    RSS feeds are still not counted, and that is because RSS readers strip out a lot of code. Also, since page views are used to determine the “value” of a page over time, which in turn is determined by ad impressions (or theoretical impressions) that would require that the numbers be split out.

  4. #4 kevin R
    September 15, 2012

    I read science blogs on a PC, an iPad, a Kindle Fire, or a Samsung Galaxy phone depending on where I am and what I am doing. One reader, multiple devices. I am not sure that looking at browsers or devices tells you much about your readership.

  5. #5 Greg Laden
    September 15, 2012

    Yes, it would definitely be misleading to assume that 80% device A and 20% device B means for every 100 people there are 80 A users and 20 B users. The best way to use the data is probably to look at changes over time. A 20% shift in from one browser to another is probably not a bunch of people adding a second machine with that browser, unless it correlates with a similar shift in OS that is, say, to more mobile devices. (Thus the safari increase we see)

  6. #6 Michael Haubrich
    Central Coast
    September 16, 2012

    As far as browsers go, I am surprised that Opera for Linux doesn’t have much traffic. I like it, but the only reason that I use Chrome more is because of the App Store. The Opera extensions aren’t very useful for me.

  7. #7 Greg Laden
    September 16, 2012

    Opera is good. One reason for resistance to it may be that it is not classic GPL/OpenSource, I think.

  8. #8 Ian Kemmish
    September 17, 2012

    Might one infer from the fact that nobody has spotted that “>1%” so far, that “scientists” pay nor more attention to tables inserted in the middle of text than anybody else does?

  9. #9 Laurent
    September 18, 2012

    French Guyana is probably counted as French traffic, wouldn’t it?

  10. #10 Greg Laden
    September 18, 2012

    I don’t know. It’s kinda far away from France, I would think it would be differentiated, but that could well be the explanation.

  11. #11 Laurent
    September 19, 2012

    Do you want me to connect at a certain time (if I give you IP#), so that you can tell if Guadeloupe is counted as French traffic (should roughly be the same as French Guyana)?

    Maybe we should try two times (H & W), because the mother research institution is located in France so probably the traffic is coming from there…