A new survey of mobile developers conducted by Appcelerator and IDC has turned up some interesting results. No doubt spurred by the death of mobile Flash, developers are rapidly embracing HTML5 for building apps. Similarly, developers seem to be embracing Google’s social efforts more than, perhaps, everyday users might be. However, another finding jumps out: Developer interest in the Android platform seems to be slowly eroding. Apple’s iOS remains the top target of mobile developer efforts, with 89 percent of respondents indicating they’re very interested in developing for the iPhone, and a nearly equal number (88 percent) saying they’re very interested in the iPad.
Android holds a strong second-place ranking: some 78.6 percent of developers say they’re very interested in developing for Android phones, and 65.9 percent say they’re very interested in Android tablet. But here’s the thing: the Android phone figures represent a 4.7 percent quarter-to-quater decline. Similarly, the Android tablet figures are 2.2 percent lower than from the last survey.
Fewer Android developers means fewer top-drawer applications for Android — and could mean users will eventually begin leaving the platform for one with better software. Why are some developers turning way from Android, and what can be done to keep them on board?
Is Ice Cream Sandwich melting on the shelf?
The results of the Appcelerator survey seem to fly in the face of comments from Google chairman Eric Schmidt, who forecast that the release of Android 4.0 Ice Cream Sandwich would spark developer interest, and perhaps make Ice Cream Sandwich some developers’ preferred deployment platform by mid-2012. As the first release of Android to be optimized for use on both tablets and smartphones, ICS was supposed to overcome longstanding concerns of Android platform fragmentation and give developers a solid, sure target for their apps.
Android’s future would seem to be rosey: In late February at Mobile World Congress, Google’s Android chief Andy Rubin claimed that there were more than 300 million Android devices out in the wild, including 12 million tablets. Furthermore, Google was activating some 850,000 new Android devices every day. That would seem to be a pretty ripe plum for developers looking to get their mobile apps in front of a large audience.
However, developers seem to be taking a cautious wait-and-see approach to Ice Cream Sandwich. Google’s latest and greatest Android operating system was initially only available on a single device — the Samsung Galaxy Nexus — and while device makers and carriers are gradually getting ICS out to a number of recent devices, progress has been slow. A surprising number of Android devices on the market are still running Android 2.2 (Froyo) or earlier. Many will never get upgraded to Android 4.0 Ice Cream Sandwich: Some users will only upgrade when (and if) they buy a new Android device. Right now, the most widely installed version of Android seems to be version 3.0 Gingerbread, which had a slower rollout than any previous version of Android. So far, Ice Cream Sandwich seems to be experiencing an even slower introduction — check out Chris Sauve’s analysis of Android versions and fragmentation for an idea. When building an app for Android today, developers still face a serious question: Do I target Ice Cream Sandwich and hope the market catches up with my app? Or do I target Gingerbread (or perhaps even earlier versions of Android) to reach the most people?
Still, Android doesn’t have it so bad. The same survey found developer interest in Research in Motion’s BlackBerry platform has dropped from 20.7 percent last quarter to 15.5 percent this quarter. And, despite underwhelming sales to date, developer interest in Microsoft’s Windows Phone platform is holding steady at 37 percent. That basically puts Microsoft’s nascent smartphone platform in the number-three position, albeit distantly. Furthermore, an essentially equal number of developers report being very interested in making apps for Windows 8 tablets.
As intriguing as the survey results are, it’s important to note they don’t represent scientific fact. The results are based on surveying some 2,173 Appcelerator Titanium developers in late January 2012, and then following up with 484 in late February. There’s no way of telling if developers who use Appcelerator’s development framework are in any way representative of Android developers in general. Both Appcelerator and IDC note that the figures for the erosion of developer interest in Android aren’t far from the margin of error for their survey. So it could all be a hill of beans.
To market, to market, to buy a fat pig…
Android fragmentation is likely only part of some developers’ concern over building Android apps. Another appears to be workable revenue models for Android applications. Simply put, making a quality Android app that generates significant revenue seems to be significantly different than making a quality app for iOS that generates significant revenue. Developers prefer to design and build one application, then bring versions to different platforms. However, the reality is that an app design and revenue model that works for iOS may not work for Android, and vice versa. While not yet common, some developers are finding themselves in a position where it makes sense to drop their Android development efforts — a recent example is Mika Mobile. And doing business with Google’s Android Market, now Google Play for Android, can be quite a lot different than doing business with Apple: Game developer Retro Dreamer says it doesn’t build paid Android apps because the international tax situation is too complex — an issue it doesn’t have with iOS.
Last November, Piper Jaffray analyst Gene Muster raised eyebrows by opining that the overall revenue generated by Android applications was a mere seven percent of that generated by iOS apps. Many criticized Munster’s analysis as incomplete, noting the figures did not seem to account for so-called “freemium” games that are free to download but that require payments to full activate or that make additional features or content available through in-app purchases. A key difference in the Android and iOS app revenue models is that Apple’s iTunes App Store is built on Apple’s iTunes Store, which has been accumulating accounts (and associated credit cards) for nearly a decade. Apple had (and retains) an enormous number of customers who are accustomed to paying for digital content a la carte. Google didn’t have that, and many successful Android app developers have eschewed the traditional pay-to-download model and embraced freemium and in-app purchases as a way to generate revenue. Of course, a tremendous variety of applications also derive their revenue from in-app advertising — another revenue source that wouldn’t have been factored in to that dreary seven percent solution.
Nonetheless, the revenue disparity between Android and iOS apps seems very real. Last December, Flurry Analytics found that if an app is available for both iOS and Android, the Android version will generate on average less than one quarter of the revenue of the iOS app. Flurry attributed the discrepancy to an Android users’ ability to pay: With Apple, nearly every user is a potential paying customer, where Google Checkout is used by far fewer Android users.
Free apps that rely on in-app advertising to support their development also have their downsides. A recent study from Cambridge University found that 80 percent of free applications in the Android Market were supported by in-app advertising. More than 70 percent of free apps request “dangerous” permissions to access things like user’s messages, email, location, contacts, phone number, and unique device identifier. The price of “free” apps in many cases seems to be a loss of privacy and disclosing personal data. As consumers become increasingly aware of online and mobile privacy issues, the free-with-ads revenue model may be inhibiting consumers adoption of Android apps.
Can Google and developers cope?
Android is not facing an imminent crisis amongst developers. But, looking out over the next two years, Android (and Google) are clearly going to have to move application development and revenue generation to the same priority level as OEM adoption and device activations, or face a stagnating software and content ecosystem. Here’s where we expect Google to focus its attention:
Currently, most in-app purchases on mobile platforms are limited to games, and are fueled by a handful of dedicated gamers buying virtual goods and content add-ons. Google only enabled in-app billing features in March 2011: Both Google and developers need to leverage this capability to create revenue models that work beyond games. This will have to include support for subscription billing — something Google has yet to enable — so users can subscribe to regular content updates from magazines, newspapers, and other publishers. Moreover, subscription billing will have to be as painless as possible to consumers. If it’s a difficult chore, it won’t be embraced.
Put user privacy ahead of in-app advertising
A tremendous number of companies offer in-app advertising services for Android (and iOS) applications: Developers typically sign up for the service, implement the ads as drop-in libraries, and hope for a check. Although there are ethical ad networks out there, in most cases app developers aren’t aware of what information ad networks are gleaning from their users’ devices, and how that information is being used. Google cannot police third-party advertising networks. Heck, Google has exhibited very little interest in policing its own app marketplace. However, Google can build privacy controls into Android that put users in charge of how much information about themselves they are (and are not) comfortable disclosing to apps. Does a comics app really need to have access to location data? Probably not, but advertisers dearly want to know where people are located. Putting users in charge of their personal data may put a crimp in the existing ad-supported app model, but it will mean consumers will be far more comfortable with that model going forward. Privacy safeguards could have a similar effect for so-called “incentivized installs,” which lure users into downloading advertisers’ apps in exchange for in-game or in-app content.
One reason consumers may not be embracing Android apps as strongly as iOS apps? Google’s market is (still) a mess. The unmoderated nature of the Android Market has always attracted a large number of apps that infringe on copyright and trademark, as well as instances of outright malware. Google is attempting to address the threat situation with Bouncer, a new automated system that attempts to reject malware and other threats before they ever reach consumers. But it doesn’t take more than a search for “Pokemon” or “Batman” to see some rough edges. Google has attempted to expose reputable apps and publishers through its editorial choice lists and charts ranking top apps, but there’s always a sense with Android Market — sorry, Google Play — that you’re just a tap from a mistake. If users have confidence in a marketplace, they’re more likely to become paying customers.