Categories

Android developers are more prolific but less persistent than iOS developers

According to 148apps.biz, the App Store has seen over half a million apps since inception. The number of available apps, according to Apple, is now 425k. (148apps claims 402k apps available in the US store.) The history of the App store catalog is shown in the following chart (showing both US and World-wide measures).

In an interesting new post by Appsfire, APPtrition – or why app store size does not matter that much… Ouriel Ohayon makes a good point: available is very different than accepted. When comparing catalogs it’s important to distinguish between these measures. Apps are published and then unpublished for various reasons. He calls this app attrition and details the reasons it might happen.

What makes this interesting is the contrast between attrition rates on Android’s Market and those on Apple’s App Store. About 80k (or 16%) of Apple’s catalog has been removed while nearly a third of Android apps have disappeared from the catalog. (By another measure, the submission total is tracked here and shows nearly 375k apps while Ouriel points out that there are 206k available. An apparent attrition rate of 45%.)

We can compare how the two marketplaces are adding apps here:

(The iTunes data is US only)

This data would indicate that the iTunes App Store has plateaued at about 20k apps per month while Android is still growing and adding at about 30k per month. The challenge for an observer is to reconcile the growth in Android submissions with the apparently far higher attrition rate.

In other words, Android developers and producers seem to be more prolific but less persistent.  With few if any constraints on submission it’s understandable that there will be more apps thrown into the catalog. But why would they be so rapidly removed remains a mystery. Maybe economics, maybe policing, maybe something else altogether explains this.

Comments

28 Comments so far. Leave a comment below.
  1. Outcoxed,

    The most likely reason is that Google periodically stages a vast improvement of the underlying core of the operating system despite leaving it open to developers to modify. Let me give you an example: prior to Android 2.2, Task Killer apps were very popular. If your phone was running slow you simply killed all the background apps and then it was fast again. In Android 2.2 this changed, the battery life improvement enhancements made it more efficient technically to not use task killers anymore.

    Finally, Google deliberately broke all the task killer apps in 2.2. They don’t work anymore. You can’t run them.

    So here’s an advantage: if Apple rewrites their DRM model you couldn’t write an app that used internal API’s so it’s simple for them to support your app going forward. If Google changes their core services it’s much harder to avoid breaking apps.

    • CndnRschr,

      There are still 634 task killer apps in the Android market. Seems there remains a lively need for such functions. Whether they all work or not is moot since they are *still* in the Android Market and hence counted (quite a few people still use Android 2.1….).

      Since it is the norm for apps to be revised and for the revision to not count as a "new" app, it's not easy to reconcile the fact that many apps are deleted from the Android Market. Something else is going on – not sure what though.

  2. the lower entry barrier in Android allows experimentation and fast iterations, no need to go through the Apple approval process to test and try new things. I do not see it as a negative fact, actually it is the opposite in my opinion.

    • definitely not a negative, but it would account for more unfinished / unpolished apps making their way to the store

    • Shrike,

      The lower barrier of entry also allows other things that are negatives: this already been 2 instances of DroidDream infections, plenty of unpolished and unfinished apps, app clones and copies, and various sundry things like that.

      Amoral things always cut both ways. Do the positives outweigh the negatives? The market will tell. The data we have today is saying that the Android Market isn't as good a place for app developers as the App Store in terms of revenue.

      Maybe the Android Market will catch up soon, but certainly not today.

  3. hoomy,

    Is the graph representing the App Store's submissions or approvals? If it's approved apps then the plateauing of Apple's rate is perhaps evidence that this has become roughly the "assembly line" constant in their approval process. Going forward I would expect Android Marketplace to demonstrate far more erratic peaks and troughs in its addition rate compared to the App Store, strictly due to the absence of a vetting process.

  4. JaWahl,

    Do ringtones and wallpaper actually count as apps in Android Marketplace?

  5. timnash,

    The catalogs cover 3 markets for apps – Paid for, Freemium and Free with ads. Apple's App Store continues to lead in the first two as the $2.5bn earned by developers (WWDC announcement) shows. As Google makes its money from serving ads, the Free with ads category is more interesting and the Android market reflects this – no entry fees, no curation or quality control – and clearly leads in numbers in that category.

    Developers however are interested in making money from their apps and since there is less to be had through Android, there is less incentive to update apps there than on iOS. So to judge the health of the app ecosystems, it's probably best to look at how many of the apps have been updated since they first appeared, and how recently, as this is a better indication of committed developers who will help to grow the platform.

  6. Gromit1704,

    I am guessing, but maybe Apple update iOS less and the updates do less damage to apps. Android appears to be a work in progress with new versions and tweaks happening every five minutes. As mentioned above, some of these changes break or kill the app and possibly there is little incentive (because they are not making the expected money) to spend time repairing a broken app so It dies. Apple updates, unless they are full 3 to 4 tend to be minor, just bug fixes, and do not add or take away functionality. They do less damage to the existing apps.
    It will be interesting to see if Android submissions plateau like the iOS ones apparently have, and where that figure eventually rests.

    • Hamranhansenhansen,

      iOS leads Android in releases by about 2:1. Wikipedia has their release histories.

      Apple's apps are native, while Google's run in a VM, so it is Apple's apps that are theoretically more breakable across versions. But Apple has been doing the work to support their app platform.

      • Gromit1704,

        Just had a look myself and you are correct. I was surprised because it is more like 4:1. IOS for phones has had 41 versions in 4 years (not including unreleased or tablet versions) and Android 11 in 3.5 years. In an attempt to desperately contradict my original answer, maybe it is lack of OS support that is making more Android developers give up?

      • Shrike,

        I don't think I saw anyone mention the fragmentation word, but the app statistics could be symptomatic of fragmentation. It all depends on how apps are counted.

        Are device/OS specific versions of the same app counted twice? Is Angry Birds for a Droid/Eclair and Angry Birds for a Thunderbolt/Froyo and Angry Birds for a Droid Pro/Froyo? Is the Angry Birds Donut version counted as a different app?

        Now look at the various Android versions in the wild: 1.6, 2.1, 2.2 and 2.3. Depends on how you count, there could be higher attrition rates and new app rates because developers are removing 1.6 (Eclair) versions of apps and submitting new apps for 2.2/2.3.

        Who knows.

  7. CndnRschr,

    iOS app developers make money from their apps up front – they presumably an obligation to update these apps so as not to piss off their user base – good business. Free apps have less stickiness to the people who download them, less ownership, less commitment from the developers. There is far less of a legacy feel to Android apps. App development is still the wild west (both iOS and Android) but Apple is laying asphalt.

  8. Jimjeffers,

    Are you taking into account that the android market is open, not curated, and thus vulnerable to a lot of spam apps. Or potentially that apple could potentially be regulating the volume of additions they make to the app store? Or that simply the curation is a bottleneck? It seems the comparison is apples to oranges when I look at it this way. Did I misunderstand the metrics?

  9. Shrike,

    I would profile what types of applications are being submitted in both stores, and it'll tell you why the Android marketplace is currently at ~30k new submission/month. There are whole classes of Android "applications" that don't exist in the App Store.

  10. Can't policing explain the high app attrition rate? My understanding (not my personal experience!) is that the Android Market is full of copied apps.

  11. Here are a few advantages Android programming has over iPhone programming:

    1) Does not require an Apple computer.

    2) Often does not require learning a new language.

    3) Android devices are more affordable in many parts of the world.

    4) The Android Market registration fee is 1/4 that of the Apple fee.

    5) Less hoops to jump through before publishing an app.

    These factors make it easier for anyone to jump in, write an app, and upload it to the market.

    They also give educational facilities good reasons to base a mobile programming course around Android.

    The flip side is that someone who has bought an Apple computer, learnt Objective C, paid for an Apple mobile device, enrolled in the iOS Developer Program, and appeased the Apple reviewers, is much more committed to trying to recoup that investment and so might stick around for longer.

    • Hamranhansenhansen,

      Most app developers know C or C++ which are both supported by iOS but not by Android. Very few app developers know Java for Android. Very little client code is written in Java. So Android is not appealing to most developers who have both C knowledge and C code they can immediately apply to iOS and easily monetize.

      Also, most developers already have Macs. The Intel Mac dominates high-end PC's as much as Windows dominates low-end. Most of the PC's at Google are Macs.

      • Lb51,

        IMO, if you know C andor C++, then learning Java would seem trivial since it is based on C. Mainly inspired by Objective-C, which is an older derivative from C than Java.

  12. poke,

    If the Android Market is anything like the non-curated markets for other phones I've seen, then they're probably locked in a cat and mouse game of copyright infringement and removal. A high percentage of those "apps" are probably things like active wallpapers, widgets, themes, etc, using artwork, photography, logos, etc, that the developers do not have the rights for. No doubt that stuff is constantly getting removed and re-uploaded.

  13. Jimmy,

    If I might propose – I think that Android has a lot of hobbyists developing for it. Since they aren't as "serious", they would be more likely to abandon a project and move on to something else. And as mentioned, it is easier and cheaper to program for Android and so easier to take on as a hobby.

  14. Ted_T,

    One word: malware.
    Just the latest of many infestations: http://www.linuxinsider.com/story/Androids-Dream-
    "The latest malware attack has claimed between 30,000 and 120,000 victims, the mobile security firm estimates.

    Like in the first incident, the list of infected applications in this latest attack includes apps with sexy names such as "HOT Girl 4" and "Sex Sound: Japanese," as well as seemingly useful apps such as "System Monitor," "System Info Manager," "Quick Uninstaller," "Brightness Settings" and "Volume Manager.""

    How many malware apps has Google had to pull out of the marketplace that Linuxinsider didn't write articles about?

  15. Kizedek,

    Oh, you think the disruptions are coming from the Android developers? And not the iOS developers because iOS developers are apparently so hampered by the curated Apple approach?

    Interesting. Because any disruptions produced by mobile app developers sure seem to be by iOS developers, precisely because of Apple's approach…

    There are actually two sides to Apple's approach. The second should not be forgotten. Moreover, in combination, they together produce more possibility of disruptions:
    1) the curation and great user experience for customers that you cite
    2) the integrated approach of Apple, with hardware and software, and now adding cloud services to that.

    There are several reasons that Apple's conscious decisions regarding its approach(es) will and ARE helping iOS developers create disruptions. These include such things as:
    • the user experience and curated app approach engender trust among customers, and, increasingly, the enterprise. Customers are willing to pay for this.
    • this in turn inspires more iOS developers to go to greater lengths to create great apps.
    • Apple is putting out some phenomenal api's and features for its developers to use, and increasing their access to core features.
    • Over time, Apple sees how things pan out and relaxes — hence the announcement this week about the relaxing the in-App purchase policies. By contrast, Google is INCREASING its control.
    • Android might have a longer spec sheet in some ways, but the actual capabilites of the iPhone and iPad are incredible… and they are being exploited by developers in INCREDIBLE ways, and in an accelerating number of industries, that not even Apple foresaw: use as menus, flight charts, interactive charts for doctors to discuss with patients, etc. etc. — where is similar Android adoption and the creation of such disruptions in multiple industries?
    • Hardware can be created and added on by iOS developers: blood analysis for diabetics, credit card swipers, car unlocking for rental agencies, etc. etc. — Again, where are the innovations and disruptions created by Android developers?

  16. I see that people think I am being anti-Apple in my comment when in fact the contrary is true. I use an iPhone and develop primarily for iPhone.

    Why I prefer iPhone has very little to do with Horace's question, which is: "Why are Android apps so rapidly removed from the Android Market relative to iOS apps from the Apple App Store?"

    I tried to answer the question by looking at the kind of people who might write Android apps, and the psychology behind them doing so compared to iOS app developers. I also tried to take the point of view of someone who does not live in the USA.

    Your points are valid within a certain context, but irrelevant; they do not address the question.

    Maybe I shouldn't be surprised at how poorly some people read.

    • Steven Noyes,

      I see the perceived entry barrier as just that. Perceived. I think the high Android Attrition rate is a result of the non-Custer aspect of the Android Market resulting in significant gaming of the Market.

      On any new App I submit, the first couple days, result in huge sales. I propose that Android developers actually pull an App and then repost it 1-2 weeks later to get another whack at the "new" category. The complete lack of a curated market keeps gaming of the system present creating huge churn rates of poor performing apps.

      • That is a good possible answer to the question :-)

        I just find it disturbing that most of the explanations in the comments for why Android apps are so rapidly removed relative to iOS apps tend to focus on negatives and would like to assume some kind of "foul play" is taking place.

        Both of our explanations are interesting theories, but one feature that distinguishes asymco from many other commentators on the mobile scene is his use of hard data made digestible using compelling graphics.

        I am interested in seeing what real facts Horace Dediu can dig up with regard to this phenomenon.

  17. Hamranhansenhansen,

    It's much more disruptive to enable consumers to safely install their own selection of native C apps as on iOS than it is to provide yet another malware applet farm, all in disrepair, as on Android. Consumers choosing their apps is changing the economics of software. $10 Keynote or $5 iMovie on a $499 tablet is disruptive. People choosing to give up XP/Office for iPad with $30 of iWork apps is disruptive. The power of the Apple native C app frameworks is disruptive … it gives developers a much broader palette than on Android, even when restrictions are factored in. The reach of the platform is disruptive: Angry Birds released for 100 million phones at once is disruptive.

  18. yet another steve,

    This is rather shocking. It costs more to keep in Apple's app store than the Android marketplace… so WHO is pulling so many apps? Hobbyist/Novice would be LESS likely to pull an app than a pro building a reputation or a brand.

    The only explanation here in the comments is the blatant copyright violation cat and mouse game. If that is what's going on THAT IS SERIOUS NEWS because it means the "size" of the Android marketplace is grossly inflated. I wonder if there are any serious journalists left to investigate this.

    If blatant copyright violations are behind this than the ascension of the platform's app count would be a total myth, and the reports of a lack of profitability in Android development would actually understate the case. It's also a conclusion that even a fanboy like me hesitates to jump to though.

    But wow.

Add Your Comments

Disclaimer
Your email is never published nor shared.
Required
Required
Tips

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <ol> <ul> <li> <strong>

Ready?