Splintered and fragmented? How to maximise your Android app
I’m Peter, one of the Android developers here at LiveStyled. I’ve been with the company for 6 months now, and working with Android for more than 6 years. I’m thrilled to see how far Android has come in that time, and now I’m excited about the improvements we’ve been working on that we’ll be bringing to our platform soon. Android is sometimes seen as playing second fiddle to iOS, but new Android devices being sold well outnumber those running iOS, so it’s really important to LiveStyled, and it should be to you too. But what I want to talk about in this blog is how we can maximise the benefits of Android - revenue, design and features.
One of the commonly cited drawbacks of the Android platform is fragmentation - how do you overcome the abundance of different Android versions, and what do they all mean? Why are they all named after candy? Most importantly, who is using what the most and how are they spending? If we can answer these questions from looking at the data we can increase revenue from the LiveStyled platform, and cut ourselves loose from the restrictions of backwards-compatibility with versions nobody uses anymore.
First off, a quick explainer. Every year around Autumn time, Google introduce the latest version of their operating system Android, and every year it is named after a candy from the next letter of the alphabet. This year it was the turn of Android Oreo, but let’s have a brief look at the last few iterations to get some context:
Android 7.0 N (2016): Android Nougat introduced a more enhanced notification system for users, multi-window support, and improved the doze mode from Android Marshmallow.
Android 6.0 M (2015): Android Marshmallow seriously overhauled permissions, which are now requested while the app is running, in context, rather than prior to install as it was before. It also introduced doze mode which allows the phone to sleep when it’s not being used and save a lot of battery power.
Android 5.0 L (2014): Android Lollipop brought in the material design guidelines we still follow today. This hugely improved the UI for Android users, and gave app designers a consistent way to create their apps.
Android 4.4 K (2014): The last brand tie-up was Android KitKat, which improved the phone app, had a smoother multi-tasking experience, and gave us ‘Ok Google’.
If you were following the list above you’ll notice we went right back to 2014. LiveStyled apps currently support devices from as long ago as 2012 but this increases the time to build apps, QA them, and ultimately get them on the store.
Here at LiveStyled we make apps which provide a great user experience and this always turns out best on the latest devices. We want to be able to generate the most revenue and provide the best experience for your customers - that means that we want to use material design, beautiful animation, and other tools and techniques available on the most up-to-date devices.
However, Android has a lot of older phones still in use that need to be catered for; and because of the huge array of devices and manufacturers, each has their own policies and variants with their own unique considerations. Couple this with a lack of incentive for manufacturers to update their devices, and the result is that usually only new devices will carry the latest Android version.
So we realise we need to support older versions of Android to have a viable user base, but how big will that user base actually be? Can we look at the data to find a compromise between supporting older devices and letting developers loose with new features? If we want to maximise the revenue we can get from our users, we need to look at the data for insight.
First off, can we see any trends in average revenue by Android version? This can give us an idea of which versions of Android have the highest spenders using them at a glance and whether or not users will spend more when they upgrade their device.
The above graph shows an upwards trend of revenue by version for all Android apps measured in GBP, with users of Android 7.1.1, one of the latest releases, coming top with an average order price of around £19.60. These are averages, not total spend, so this isn’t taking into account the number of devices in use on each version, which may explain some outliers (4.1.1 for example).
Below is a pie chart showing the percentage revenue in GBP for all versions of Android in the last 90 days (from 08-Jan-2018) as recorded by LiveStyled.
The first thing you’ll notice is that the vast majority of payments are made from Android devices running version 7 (Nougat). This was released back in 2016 and is now the most commonly shipped version of Android with new phones. Many users may also have upgraded to Android 7 on their old devices bringing them up to speed. Devices running Android 7 and above account for 74% of revenue in the last 90 days, which shows that most paying customers are using newer versions to spend their money.
But if 74% of revenue is coming from newer phones, what about the other 26%? That’s still a sizeable chunk of revenue we wouldn’t want to lose, so where do we draw the line? This chart breaks things down a little more succinctly, showing the stark drop off in revenue below Android 4.4:
As you can see, over 99.5% of 90-day revenue was taken from devices running Android 4.4 (KitKat) and above. KitKat was released way back in 2013, and is already over 4 years old. One of the problems with KitKat is that it doesn’t follow the material design guidelines, and newer apps which do follow these guidelines need to use support code to keep the apps running on KitKat, which makes designing and developing them more difficult and less efficient. So what happens if we take KitKat out as well, and go straight to Android 5.0 Lollipop, where material design was introduced into the Android platform?
Just 2.4% of revenue came from devices running anything below Android 5.0. When you factor in the cost of supporting these older versions, and the limitations imposed on designers and developers because of it, is it worth supporting such a small subset of devices, a base which is only shrinking all the time?
Another interesting fact is that the data in these charts is based on purchases made in British Pounds (GBP). Our Scandinavian clients in Norway and Sweden have a revenue rate of 0% for devices running anything below Android 5 - 100% of revenue comes from customers using Lollipop and above. Scandinavia is well known for its early adoption of technology, but this is still a surprise given that we support down to Android 4.1. It only strengthens the case for being forward thinking and moving support away from older versions.
QA should also be considered when trying to have such a broad reach. Testing for older devices adds time to the QA process; and it increases the potential for problems when a greater spread of devices need to be able to run the app. It may not be worth the investment in backwards-compatible QA if the vast majority of revenue is coming from more modern devices.
Ultimately we in the Android team here at LiveStyled feel that the data demonstrates it’s time to leave anything below Android 5 behind. We think that to appeal to the most affluent users and maximise your revenue, you need an experience that runs best on Android 7 and above, which is much more effective when your app is looking to the future.
Join our mailing list here: