The Varieties of Android Experience
Last Friday we snapped the photograph in this post. We posted the image online and it spread virally to reach unanticipated global fame. Much was said about Animoca and about our excessive QA practices. Even more was said about Android and fragmentation. Today we would like to give you Animoca’s point of view on these issues.
A little background
After the photo generated some buzz on social networks, TechCrunch featured our image in a top-trending article titled This Is What Developing For Android Looks Like. The responses to that article ranged from enthusiastic agreement to the predictably vitriolic. Some opined that the writer, respected tech reporter Kim-Mai Cutler, was denigrating Android to exalt iOS (we disagree).
The TechCrunch article explained how Animoca chooses to perform QA of Android apps across as many devices as possible – it did not suggest you have to do the same, or even that our approach is necessarily the best. That said, keep reading to find out why this method works for us, and why Android fragmentation isn’t always a bad thing.
Do the diversity embrace
Here at Animoca we are Android fans, and not just because Android is our primary platform for game development and distribution. We like Android’s potential, and we appreciate the flexibility and freedom it offers to users and developers alike. Apple famously takes a “one size fits all” approach (and does it well), whereas Google’s platform offers a product range that proclaims reassuringly: “we have something in exactly your size”.
Android fragmentation is a bit like multicultural diversity: it may cause friction from time to time but a community that embraces it ought to be richer and more vibrant. Android users have a truly impressive choice when it comes to smartphones and tablets, and no other mainstream mobile operating system provides the extensive functionality and potential of Android at such a variety of price points. That’s why we embrace the diversity of the Android ecosystem.
What is QA, really?
To Animoca, QA means quality assurance in the most literal sense. It’s not simply compatibility testing. It’s not simply bug-finding. QA at Animoca is about making sure that as many users as possible are able to enjoy the best experience that an app can deliver on a specific device, taking into account the restrictions imposed by hardware and software variations – not to mention the requirements of different app stores (Amazon, T-Store, etc.) and numerous SDKs (AdMob, Tapjoy, etc.).
Someone using an ARM6 processor ought to enjoy our apps as much as someone on an A10 or dual core processor. While the newer, more powerful device will probably have HD graphics and other attractive features, we believe that all Animoca users – including those on low-end devices – should be offered the best possible app experience. To do otherwise hurts the consumer and, by extension, our reputation as a developer and publisher of Android games.
Animoca products must not only work in accordance with generalist compatibility tests; our products must work well. That’s why we test some apps across hundreds of devices.
The low-end is not the bad end
Some developers choose not to support low-end devices, while others explicitly support only some popular models. At Animoca, providing support across different devices whenever possible is part of our mission. In a previous post, we listed the most popular devices used to play our games; a glance through that list will reveal that half of the top 10 models turned out to be low-end devices like the Galaxy Ace or the HTC Wildfire.
Developers who focus their attention on more expensive devices like the Samsung Galaxy SII or HTC Desire are by no means wrong to do so – we use a similar strategy for certain games that require more processing power, like Ultimaze or Forest Defense. But remember that when you exclude the low-end, you also exclude millions of potential customers who own those inexpensive Android devices. It is a decision that must be weighed carefully.
Thankfully, we don’t live in a world where all smartphones cost USD 500 or more. Our opinion (and hope) is that the digital divide in many societies will be bridged by the kind of low-cost personal computing devices that are becoming abundant thanks to the rise of Android. Product diversity also means cost diversity, and that opens up interesting opportunities (we have teased this concept in a recent post; more to follow).
What if consumers only had expensive Android devices to choose from, instead of the entire range? The digital divide would become an impassable abyss. The advantages of owning a smartphone would lie beyond the reach of most consumers.
One of the mottos of Commodore founder and Atari legend Jack Tramiel (who passed away on April 8, 2012) was: “computers for the masses, not the classes”. Android’s low-end not only represents a substantial market, but it also has the potential to deliver affordable personal computing to a vast number of people on the less fortunate side of the digital divide. We like that.
It’s not Google
We can probably all agree that the annoying thing about Android fragmentation is the vast number of different versions of Android in active use. Why aren’t all devices updated to Ice Cream Sandwich, or at least a reasonably modern version of Android?
This problem has little to do with Google and much to do with the constellation of Android device manufacturers and vendors. In order to reduce Android fragmentation, manufacturers must make stronger commitments to deploy operating system updates to their devices more regularly. Providing operating system updates is increasingly essential because Android – like any operating system – is evolving all the time. Instead, today vendors apply Android updates sporadically, if at all. Fortunately, this state of affairs is set to improve substantially as Ice Cream Sandwich continues to roll out.
How global is your market?
We suspect there may be some more Morbus silicii vallis making the rounds and preventing some readers from seeing the wider picture. Animoca’s market is international. North America only represents 27.5% of our global audience (details here). QA’ing for the entire world requires us to test our apps on many more types of devices than developers who are interested in fewer markets and specific geographical zones.
We find that some of the most interesting Android markets are in Asia: South Korea and Japan are quite ahead of the rest of the world, tiny Taiwan produces a disproportionate amount of cutting-edge mobile technology, and China is the single biggest smartphone market. Mention the word “gaming” and you might think of Valley-based companies like EA and Zynga instead of the larger, more profitable and better performing Nexon or Japan’s DeNA or GREE.
Asia has the advanced technology, some of the biggest gaming companies, and the greatest number of mobile users. There’s also the fact that the majority of Android devices are of Asian origin (eg. Samsung, HTC, LG and Sony), which means Asian consumers also enjoy a broader range of products. These are, clearly, very attractive markets for app developers.
So consider South Korea: everyone knows Samsung and LG, but few know of Pantech, which is the second largest handset maker in the country (ahead of LG). It has revenues of USD 2.8 billion and is definitely not a small player. We have a few Pantech Android phones and some of them are really quite good. Deciding to exclude Pantech devices because they are not well known outside of Korea would be like refusing to test for Motorola’s Droid Razr or the Xoom: we would probably regret it later.
Some developers who focus on single or a few markets can afford to limit the QA process to the few devices identified as sufficient in their target markets; Animoca’s global strategy requires us to act more comprehensively. That’s why we do QA for our apps on a multitude of devices. We’re not being wasteful – we are attempting to cater to a variety of different markets across the world.
The tyranny of compatibility
Some posters responding to this issue offered views on compatibility targets, with a few arguing that 90% compatibility is good enough and that what Animoca is doing (QA on hundreds of devices) must be overkill.
Let’s look at some numbers. In our year and a half of existence, Animoca apps have been downloaded 70 million times and we have always striven for high compatibility (achieving well over 99% in some cases). If we had taken the approach that 90% compatibility is good enough, we’d be lacking support for 7 million of those downloads. Several millions of consumers would have had a bad experience as a result of our decision, and our app revenues would probably be short by around 10%.
Ok, you say, 95% may be a more reasonable target. Yes, but 5% of 70 million users is still 3.5 million; would you give up 3.5 million potential customers/downloads if you could avoid it?
To a certain extent it is a question of scale and distribution. Animoca invests in QA because, ultimately, it yields returns on the scale and distribution at which we operate. Some commenters suggested that we can afford to be frivolously extravagant in our QA process because we are funded by Intel Capital and IDG-Accel. However, our QA philosophy predates the funding event in question (November 2011), and our QA practices are intended to protect the investments we have received.
The Varieties of Android Experience
Higher device compatibility means a bigger audience and that translates to higher total income derived from that audience. There have been some reports that Android monetization is inferior to iOS monetization. We have disagreed in the past, and continue to disagree. At Animoca, monetization of Android apps provides yearly revenues in the double-digit millions – something that we are not alone in experiencing. We have to wonder if the low monetization some sources have reported for Android may be partially due to problems of product compatibility.
Today, developers who choose to support only a limited set of devices are automatically limiting the ecosystem in which they do business. That may be the right decision for them but we maintain that our broad approach makes sense for our company and its strategy.
So, after writing at such length, what (you may ask) is Animoca’s opinion on the Android fragmentation issue? The varieties of Android experience are best seen as a mixture of perils and opportunities; from our humble vantage point, the perils appear manageable and the opportunities extraordinary.