Google changes how Android stats are collected

Google changes how Android stats are collected

Summary: Google changes the way it collects version information about Android, give Jelly Bean a significant usage share boost. However the aging Gingerbread continues to be the most popular version of Android.

SHARE:

Search giant Google has changed the way that it collects statistics on Android device usage in an attempt to give developers a better picture of the version landscape.

Google used to collect information on the versions of Android in use by keeping a tally of automated check-ins to its servers, but this month, that system has been replaced with one where only devices accessing the Google Play store are counted.

News of the change was broken on Google's Android Developer Dashboard page.

Beginning in April 2013, these charts are now built using data collected from each device when the user visits the Google Play Store. Previously, the data was collected when the device simply checked in to Google servers. We believe the new data more accurately reflects those users who are most engaged in the Android and Google Play ecosystem.

This change in how the statistics are collected has given the most current versions of Android, 4.1.x and 4.2.x, otherwise known as Jelly Bean, a significant boost, up from 16.5 percent last month to a whopping 25 percent for the 14-day period ending on April 2, 2013.

(Image: Google)

Even under this new data collection regime, the most popular version of Android continues to be Android 2.3.3 to 2.3.7 Gingerbread, a version first released back in February 2011.

So, are these numbers twisted in Google's favor, or are they a more accurate representation of the Android versions in use?

My take on this is that the data on the Android Developer Dashboard is aimed at helping developers know what versions of Android are most popular in the Google Play store so they can write apps that target these platforms, and that this data was never meant to provide a metric as to Android usage. With that in mind, restricting the data collected to devices that actually access the store gives developers a fairer view of their potential audience.

However, by hanging the way the data is collected, Google has made it harder to get a full picture on what devices are in use, and the fact that the new collection method seems to skew the data in favor of the later versions of Android will no doubt take some of the pressure off Google to do more to get Android updates into the hands of users.

Topics: Android, Google, Operating Systems, Smartphones, Tablets

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

26 comments
Log in or register to join the discussion
  • geez!!

    such a complex system of different version and with so many hardware vendors how would one know where it's getting delayed to reach a device? of course a user wont wish to delay any update once it is available to him. one of the bitter benefits of an OPEN system..
    dugbug11
  • Just curious wrt Android updates

    Do devices running Android Gingerbread continue to get security updates?
    Rabid Howler Monkey
    • Most phone models running ICS and JB don't continue to get updates.

      I would think GB is all but forgotten unless you root your phone and do it yourself.
      Bruizer
      • yup

        as a GB user that was promised (lied to about) ICS updates I can tell you, yup, GB users have been forgotten about
        dosmastr@...
    • It's (almost) irrelevant

      Old or new versions allow (with some reserves) to install fake applications that can be malware.
      Phones have a shorter life than traditional PCs, upgrades are not that relevant as some try to make it.
      AleMartin
      • Yep, and...

        Android had support for things like Dictation all the way back in 2.2 and LTE since Gingerbread so, sometimes other companies may get more updates but their updates still leave their devices lacking when compared to older versions of Android.
        slickjim
        • I agree

          Even if new devices on sale are normally one version behind (sometimes even 2), android is so much advanced compared with competition that they can still compete with success.
          AleMartin
          • says the man with a new/expensive phone

            oh ok then, trade phones with me and see how advanced my GB phone is.
            dosmastr@...
    • Yes and No

      Gingerbread has gotten multiple point releases and last I used it, I think it was up to 2.3.8 and all of these after 2.3.3 were for bug and security fixes.

      Most vendors will push out the minor patches but, not all of them. However, you asked if they were updated for critical updates and at this point they are beyond the two year support cycle but, given the install base, I would bet really critical updates would still be rolling out.
      slickjim
      • But they don't get updated. It is that simple.

        The vast majority of Android handsets simply do not get critical security updates.
        Bruizer
        • much as i'd like to agree...

          LG promised and never delivered ics to my c800, but I did get two updates since this time last year, both GB and one was as recent as Jan 2013
          dosmastr@...
  • The answer is neither...

    "So, are these numbers twisted in Google's favor, or are they a more accurate representation of the Android versions in use?"

    From a developer stand-point, this will help make a decision on what you need to code/test for but it will show a bias toward newer devices and lower the count for old devices. It will show a slightly better picture of OS adoption but this is a side effect.

    It will not give an accurate representation on Android versions in use but it will give an accurate representation of usage.

    The key point, however, is this line Google:

    "We believe the new data more accurately reflects those users who are most engaged in the Android and Google Play ecosystem."

    Android has always had an engagement issue. It has very low engagement on web browsers and applications when compared to other smartphone platforms. This is the first time Google has actually acknowledged this.
    Bruizer
    • about engagement

      A lot of people just use their smartphone for a few basic stuff : as a phone, check emails/facebook and take pictures (maybe)

      why do you think it's a problem?
      Jean-Pierre-
    • That's not true

      Throwing bad stats around about engagement issues doesn't make it true.

      Android just has more advanced browsers that can report themselves as anything the user chooses. Mine right now is being reported as a desktop OS and I have the options of iPhone, iPad, Android, Desktop, or Custom (I can the web page I am on IE if I want.

      As for the Play store, a large number of countries see significantly more engagement within the play store than Apple's App store and these stats are out there.
      slickjim
      • Throwing bad excuses around is simply silly

        The vast majority of Android users (as in 98%) do not know how or care to change a user agent string. Likewise, this does nothing to explain the huge app discrepancies between iOS and Android. The sad truth is Android lacks engagement with its users and this creates a very weak Eco-system. Making up reasons why almost every study shows substantially less user engagement with both the web and apps with Android is simply silly.

        I know of no country where per capita engagement on Android is higher than either iOS or WP. There are countries with significant installed base and market share but per capita engagement is still very low.
        Bruizer
    • browser stat skewed

      I have 6 Android devices in active use and all of them use browsers with desktop user agent. None will get reported in stats as Android. When I use safari on my Mac, my iPad which has icloud tab syncing and my iphone also registers usage without me using it. All three Apple devices on different IP addresses.
      And the zealots shout about iOS browser usage stats
      warboat
      • are you claiming to be typical????

        good for being a geek but how many android users or even smartphone users out there are like you? i doubt its over 10 or even 5%
        dosmastr@...
  • Still though, the 4.x series as a whole is now larger than 2.x. Frankly there isn't a ton of difference between 4.0 and 4.2. The platform is really plateauing - a very solid stable mature platform at this point and so the issue of fragmentation is lessening. You really only have to target 2.3 and 4.x now.
    drwong
  • Correction: not most popular,

    most common. People on gingerbread are on gingerbread because the devices they bought either can't support a later version, or the carrier they bought it from abandoned them.

    I doubt a lot of people are willingly choosing to stay on gingerbread.
    baggins_z
    • gingerbread

      A very large proportion of that gingerbread stats will be just the Galaxy S2 phones.
      It got a million versions of ICS from Samsung to choose from but very few carriers pushed any of them to users as an update.
      frankly, they don't have a reason to update, their S2 runs fine without it.
      Most normal people just don't have update-itis.
      warboat