In our last post about the Android Market, we dispelled the reports in TechCrunch by mobile research firms Distimo and research2guidance that the Android Market is set to surpass the App Store in size later this year. Distimo counted 206,143 Android applications as of March 2011; In research2guidance’s May report, they counted 294,738. Shortly after research2guidance’s report, we got some authoritative data to set things straight. From Google IO’s conference on May 10:
The projection that Android Market is going to surpass the App Store (now over 425,000 available apps) in size in late summer is wrong, or at least misleading. As is a separate data point by Distimo that there were more free apps on the Android Market than the App Store as of March 2011, though that will happen soon. The Market, although growing at a faster rate than the App Store, is still far behind in the number of available apps and will need more time to catch up.
Let’s recap the numbers. As of today, June 7, 2011, there are approximately 206,000 available Android apps in the Market. As mentioned above, there are over 425,000 iOS available apps in the App Store. So the App Store is still more than double the size of the Android Market.

App Attrition

Why were Distimo and research2guidance so far off? Most likely they failed to isolate the apps that are actually available to consumers in the Android Market, reporting instead the number of Android apps that have ever been created.* The Android Market has a relatively high rate of apps that were once available for download and have since been unpublished. Out of roughly 300,000 apps that have ever been on the Market, over 95,000 (32%) are no longer available.

In contrast, out of approximately 500,000 apps that have reached the App Store, about 80,000 (16%) have been unpublished. So in Android’s briefer existence, it already accounts for roughly as many unpublished apps as the App Store.

*They may also be including beta apps or private apps that were never intended to be published in the Market. The Amazon App Store is one such “private” app that is not distributed in the official Market, but it is an exception to the rule. In any case, there are also a considerable number of “private” apps for iOS, distributed through alternative, jailbreak-only stores like Cydia that would also need to be counted.

Moreover, as some device manufacturers and platforms begin to adopt Android as an OS while distributing their apps through their own stores (see the Blackberry Playbook), there are indeed more Android apps not distributed through the Market that need to be accounted for. But this number is highly unlikely to bridge the gap between Distimo/research2guidance’s numbers and ours/Google’s.

Pulling the plug

Why unpublish an app? The primary reasons are listed below, along with our hypotheses for how they have contributed to attrition rates:

  • The app was largely exploratory and not intended to make an impact. The lack of a review process for the Android platform provides little disincentive to publish, then unpublish an app. A first-time Android developer may just want to explore the experience of having an app on the Market and pull it so as to spare his reputation a mediocre app with few downloads and a few, poor reviews. Android also offers tools like App Inventor that greatly reduce the barrier of entry, even among non-developers. These “explorers” may also publish an app out of curiosity.

In contrast, Apple’s review process, stricter guidelines and $100 annual developer membership fee, represent a greater investment for developers, making it costlier to change course. Put otherwise, it’s economically wise for an iOS developer to set a clear, sustainable vision for his app offerings.

The difference in barrier to entry is an important point to consider when comparing the sizes of the two stores. With a lower barrier for entry to publication on the Android Market, there are likely a considerable number of “throwaway” apps created for Android that were, in fact, thrown away. Then again, there may also be many “throwaway” apps for Android that were never thrown away and continue to idle on the Market.

  • The return is not worth the investment of time, money, or attention. Google’s initial promise of openness and monetization has given way to the reality that profiting on an Android app is an uphill battle. A greater percentage of Android apps are free in comparison to those the App Store – for various reasons that don’t bode well for Android developers looking to make money. And advertising via AdMob is an underwhelming revenue model. As some developers – even those in the middle echelon of app quality – have found, maintaining a published, functional app in the Market may not be worth it. To be sure, maintaining an unsuccessful app in the App Store is more expensive due to Apple’s minimum $100 developer fee. It’s possible that more apps have been removed from the App Store because the expenses of maintaining the app exceeded revenues.
  • The developer changes course, opting to offer only a free or paid version instead of both, or a sufficiently distinct version of an app that warrants replacing the original. The inability to switch an Android app from paid to free (and vice versa) is also a cause of apps being discarded, especially considering the struggle it takes to settle on a solid revenue model on Android.
  • The app violates Android’s TOS. While Google’s lack of a review process means that any developer can publish a pornographic or otherwise mature/offensive app, Google’s TOS disallows such apps. Users will report these apps and they’ll eventually be unpublished from the Market. In this respect, whereas iOS takes a preventative approach, Android governs its store remedially. This would logically result in more apps being published and then unpublished on Android compared to the App Store, but this is likely a minor cause.

Size doesn’t matter

Key takeaways:

  • While increasing sales of Android devices are indeed a reason to be enthusiastic about the platform, reports of Android’s explosive growth have been exaggerated in some outlets.
  • There has been greater attrition on Android than the App Store in spite of the fact that it is more expensive to maintain an app in the App Store than in the Market (although this trend may not necessarily continue).
  • A lower barrier of entry to the Android Market than the App Store suggests that Android’s long tail offers less to consumers than the App Store’s, though perhaps greater accessibility will result in more diamonds in the rough.
In general, we find the comparison between the sizes of the Android Market and the App Store to be relatively unimportant. The numbers of applications available in the Blackberry App World and on the Windows 7 platform are salient because their stores are currently too small to satisfy a large number of customers. But comparing the sizes of the App Store and the Android Market is a flawed way to measure either’s success. At the very least, if you insist on comparing the size of the respective stores and projecting their futures, be sure to include their rates of attrition and patterns of growth.

As you’ve all seen with App Trends, Appsfire is looking at any and all signals that indicate trends in the App Store. Keep an eye out for more on that front in the near future, as Appsfire introduces new and ever-more accurate tools for measuring relative trends and indicators of health between the App Stores.