Open Signal Open Signal notes the diversity they see in downloads for their app.
When we think about Android, one of the first things that comes to mind is variety. While a nice luxury to have, it also brings a wealth of other problems in what we like to call fragmentation. So many devices, so much variety, and it doesn’t stop at hardware.
Open Signal has just released their Android Fragmentation Visualized report, at it has some interesting statistics. The info provided by Open Signal shows eight different Android iterations in use, whereas Apple has three iOS variants. Of those three, one has one percent of the overall share, leaving the other two to take the lions’ share. Of the remaining two, iOS 6 holds a staggering 95% of devices, meaning nearly everyone is up-to-date.
While Android diversity is a bit like the never ending cycle of a snake eating its tail, hardware is as good a place to start as any. We’ll start with a bit of a disclaimer on this whole report, though: these numbers represent Open Signal Maps’ own unique statistics, and are not in any way official Google figures. While not official, Open Signal Maps is a widely used app (nearly 4 million downloads and counting), and therefore is often looked upon as a good barometer for what is really happening with fragmentation. Their numbers also coincide with official Google statistics on fragmentation. In equity, Open Signal decided to sample 682,000 devices for this report, which is the same number they sampled last year.
When it comes to device fragmentation, Open Signal Maps has some very clear evidence that the situation is getting more diverse on a hardware level. They saw their app realized on 3,997 different devices in all of 2012. That’s an interesting enough number, but when you account for the 11,868 device already in 2013, the task of adopting Android across the board starts to become clear.
Open Signal The diversity of Android is contrasted by the simplicity of iOS.
With so many devices utilizing the app, you can imagine the screen sizes vary a bit. That’s probably putting it far too mildly.
The screenshots above show another way in which Apple has it right. Developers for Android have to get their app to work on a variety of screen sizes, while the Apple landscape is much easier. It may seem like extra work is needed for all those dissimilar screen variants, but that’s not always the case.
I reached out to my friend Samuel Johnston with Open Source to find out how developers successfully navigate the amazing breadth of screen size variants:
We tried to do two things: come up with a type of layout that scales well, secondly make sure all the images in that layout stay really sharp. Some things look stupid when scaled while others don't - e.g. if you blow up a button, it will look like it was made for a giant, but if you blow up a dial it will look like you've just made it clearer. In general as you move up screen sizes, the white space should scale up more than the size of UI elements.Samuel JohnstonOpen Signal
Open Signal This chart shows Samsung’s 47.5% share of Android.
One of the sleepy contributors for fragmentation lies with OEMs. If an OEM like HTC or Samsung decide they no longer wish to support a device, that device is no longer able to get an upgraded OS from the manufacturer. This can be problematic in regard to reports like this, but the decision to “abandon” a device is often due to hardware limitations. As Android gets more robust, it asks more of the hardware, and newer versions of Android on older hardware may task the device too much. Open Signal notes in their report that limited hardware for some Android devices has the ability to reach developing markets and countries with economic concerns, as the devices can be had at a lower cost. Lower cost is due to lesser hardware, and that limits the device’s capability to run the newest version of Android.
As Android gets more robust, it asks more of the hardware, and newer versions of Android on older hardware may task the device too much.
From a manufacturer’s perspective, if they have a skin (like Sense or TouchWiz) that has already been developed for an older version of Android, working that onto a low-end device for an emerging market can be a winning proposition. In many cases, the hardware is probably very similar to the device they originally designed that skin for, meaning their work is done, and another market has access to Android. It is also a good option for established markets, as it opens Android up to a whole new market segment for those that may not be able to afford an HTC One or Samsung Galaxy S4.
When we examine the device fragmentation chart above, we’ll see that Samsung holds the lion’s share of Android diversity. According to Open Signal’s stats, Samsung has roughly 47.5% of the Android space. Sony is a distant second with 6.5%, Motorola at 4.2%, and HTC with 3.9%. If that seems a bit off, remember that these are worldwide statistics. While Open Signal didn’t track the location of each device, they tell me that the top reporting countries are the United States (25%), Russia (5.5%), ITaly (5%), Germany (5%), Brazil (4.5%), and the UK (3.5%).
Open Signal Android and iOS fragmentation are highlighted here.
Open Signal sees eight different versions of Android running their app, which presents as much problem as benefit. When I prompted Johnston to tell me at what point they give up on a version of Android, his answer was a bit surprising to me. They rarely do:
Generally, when a new API is introduced, you have to make sure your app is compatible as any changes Android make are likely to be carried forward the next API level after that. You definitely can't skip an API level. On the other hand, we did give up on Cupcake when its market share became so small Google stopped publishing stats of how many people have it. When a new API level comes out, there's a mixture of feelings: you might have to give up on one of your features (e.g. the ability to control airplane mode in apps was removed in 4.2), you might have to make some small rewrites, but also there'll be a lot of cool new things you can do as well.Samuel JohnstonOpen Signal
Android is open source, and the folks at Open Signal appreciate and embrace that. While diversity can create a unique set of challenges for building a great app, the rewards are there. Johnston mentioned to me that good reviews led to increased downloads of their app, which led to a higher ranking in the Play Store. That can only be accomplished by diligently adopting many versions of Android, and appreciating the benefits of the operating system in the face of all the work it takes. This has a lot to do with the diversity of devices they now see, but it’s also a good barometer for where Android is in regard to such variety of operating systems in the wild.
Like this post? Share it!
Those charts and graphs… O_O
Agreed. Completely amazing.
Diversity is a good thing. I can get a cheap tablet running v4.1 for $99 at any retail store, will it ever see an update in its life; NO! But is it worth it for, depends on your needs. And even that $99 tablet can be rooted and have custom ROM installed which may be based off of the latest version of Android. No big deal.
but not for developer , if you are developer with small investment its going to be hard.
that’s okay for both parties. when we get further down the line, we will have enough depth in the play store that someone else will come up with free apps for most core functionalities. that’s the beauty of android where there are apps like notifications toggles, helium backup app, mxplayer free, etc etc – all of them are ad supported, or in some cases – simply free. there used to be a time when the bar was low – so core functionalities like a torchlight used to come with a price tag. its no longer the case, and developers will have to move on in such cases. it’s like winzip, no one bought them, then 7zip came along.
as regards paid apps – yeah, it’s jolly good time people paid up for good products. you can still accomplish a LOT without any paid app, and with a $99 4.1 tablet.
Lets not forget that Google is moving the APIs and such to its Google Play Services and pushes updates regularly, the devs can still benefit by only having to program for one set of APIs and the enduser gains improved functionality even if their manufacturer/carrier is holding a version update. Android Fragmentation is way over blown and really a non issue at this point.
It just occurred to me looking at the Brand Fragmentation image that part of that is due not to Android or the OEMs but the carriers that want THEIR version of a particular Model. How would that image look if that distinction is removed?
aah the carrier beast. that problem not going to be solved anytime soon, unless people stopped their buying pattern. just pony up the money and buy the damn phone like people do around the world. USA and a few select market are completely trapped by these carriers. I cannot fathom how they got themselves there in the first place!
Sony/Sony Erisson much bigger than i expected. i thought they far behind HTC, LG, Huawei and Motorola
they’ve been doing really good since 2012. that’s because they’ve been trying very hard with good product. if only they matched that with a marketing budget like samsung.