The iOS App market sales run rate is $4 billion per year

At the 2012 WWDC, Apple released new data concerning its App ecosystem. Namely:

  1. 30 billion apps downloaded to date (excluding updates)
  2. $5 billion paid to app developers
  3. 400 million iTunes accounts

These three data points allow us to update our picture of the app economy. First, the app download rate.

As would be expected from an expanding user base, the app download rate has been increasing. It is now at about 49.5 million apps downloaded every day. The history of this rate is shown in the chart below:

I added the same data for iTunes songs and book downloads for comparison. Note that although music and books are available to the same user base (actually higher due to Macs and iPods which do not run apps,) apps are being downloaded at a far faster rate–at least four times faster.

In terms of total, cumulative downloads, the comparison is even more stark:

It won’t be long before total app downloads will be double total song downloads. Put another way, in the time it took the App store to reach 30 billion, the song store sold only 3 billion.

On average, songs are more expensive than apps ($1.29 vs. $0.24). What is the difference in value, not just units?

The data shows that apps have very likely overtaken songs as “gross” income. Apple does not book all app income, only the 30% share they keep, but in terms of transaction value, clearly apps are now a bigger business.

What about consumption? Are individual iTunes account holders buying a constant or variable number of songs or apps?

Here again apps show a distinctive pattern of out-performing songs. As far as digital media is concerned, apps are taking the download business by storm.

What about pricing? Is the expansion of the user base into the hundreds of millions leading to a collapse in app pricing?

No. You can see the price as the ratio between the downloads and the revenue to developers. If downloads scale more rapidly than payments then pricing is decreasing. As it turns out the average price of an app has held steady since approximately 1 billion downloads were recorded. That price is now about 24 cents per app (including free).

Finally, how about the content owners? Are payments to the owners of the media beginning to favor developers over music publishers/promoters?

Developers still have a long way to go. We can’t be sure about how revenues are shared with music labels, but it’s probable that they are still receiving bigger checks than developers. However, under some scenarios this will change quite soon.

The bigger story here is that the iOS app economy has reached a run rate of $4.3 billion per year from a standing start four years ago.

How big will the TV app economy be four years from the launch of an Apple smart TV?

  • Great analysis as always, Horace. I’d be very interested in a couple things:

    1) Overlaying key dates in the average sale price graph that could drastically change the market (launch of the iPad, which introduced many popular apps selling at $4.99+, free on contract iPhones, which we worried would drive down the average sale price, and launch in China and other growing markets)

    2) Making some assumptions based on the Top 300 paid, free and grossing apps in a number of markets to determine how in app purchase has shifted the model for many developers.

    3) Some analysis on “true” downloads per device per month once you remove some devices based on a fair depreciation schedule.

    Let me know if I can help in securing any of the data if you’re interested in looking into the above.

    • Hitendra

      If you could share the data, that would be helpful.

  • obarthelemy

    If we were to apply Facebook’s lofty “per account” valuation to Apple, not even taking into account the fact that Apple do have their customers’ credit card info, and should be more “sticky” (iTunes and all) how much should Apple be worth ?

    • Joe_Winfield_IL

      That’s a fun academic exercise, but I’m not sure Facebook makes for a very good baseline in any valuation comparisons. Perhaps the opposite would be more valuable…

    • Tatil_S

      Apple’s account holders interact with mainly with Apple, not with each other, (although iMessage and FaceTime has been muddying the waters lately) so Apple’s numbers do not have that network effect. You can switch to another provider by yourself, you don’t have to wait for enough of your friends to make the move before you can realize the value proposition offered by a competitor.

      • obarthelemy

        What “other provider” ? Who else is allowed to offer apps, services (backups, cloud, sync…) remotely as integrated as Apple’s, even content ?

      • theothergeoff

        metcalf’s law not withstanding… I think the key here is ‘per account’ valuation is critical to Apple in it’s marketing to record labels, publishing, and developers. If you have an AppleID/CC on file… you are 2 clicks away from fulfillment/sales-booking.

        Specifically, gaming. I’m curious if GameCenter will evolve, or if Apple is really just going to let all the Multi-Player game companies use the ‘agency’ model to have Apple just play bill collector for them for the 30% fee.

        Same for TV content. I think ‘AppleTV’ is more about bypassing current revenue channels (networks/cable cos). It’s a rehash of DVR+OnDemand+rental… but if Apple can pull off the “network of ‘me'” (after all isn’t that what iTunes is selling…. this is ‘my album’ ‘my mix’ ‘my radio station).

        What I read from these slopes is… AppleTV may be a ‘hobby’, and iTunes may be ‘break even’ but the AppStore… The whole design of this walled garden is to build the ‘cathedral’ that every one is compelled to buy from. As HW margins drive to zero… getting someone to spend $100 a year on apps and inApp purchases [$10 a month], over say a 500 million ‘locked seats (where else can you get your apps). at $30*500 Million = 15Billion in pretty much profits. No wonder the WWDC has become the WWiOSDC.

        And from a stockholders perspective, ‘per account’ becomes a measure of growth… if what Horace plots out becomes maintains that trajectory… which it should last until smartphones saturate the phone market, and iPads level off against Laptops/netbooks/ultrabooks/desktops… Say What… 5 years? More? Less

  • Aaron Pressman

    Fascinating charts but I feel like the free apps are skewing the comparison with music. Almost no music is free and the music labels pushed tghrough a huge price increase per track in 2010. If the average price paid per app download is 24 cents then at least 3 out of 4 apps, in reality probably even more, are free apps since the minimum app price is 99 cents. And if you subtracted all those free apps from the app/music charts, apps growth rate look pretty much like music.

    • This is why I compare value in the Monthly Income chart.

      • Great analysis! iTunes does have an active market for free content: podcasts. Although these are not hosted on Apple’s servers in the same way as songs are – and so have a limited impact on the company’s infrastructure costs – Apple does manage access to this content. It’s an imperfect comparison to free apps, but including numbers on podcast growth would be an intriguing data point in future discussions. Regardless of its relative popularity, it does say something about the health of the iTunes platform.

    • DJ Bean

      At the same time there’s probably a fair bit of ad revenue generated from those free apps that isn’t reported in Apples numbers….

  • vastaman

    Do we know that “downloads” only counts first downloads and not app updates? The latter could easily explain the non-linear growth and I suspect it’s what’s being reported.

    Songs rarely get redownloaded.

    • Having followed Apple for 34 years, I am quite sure that downloads are app purchases and not updates. Apple doesn’t massage numbers — and a CEO giving out misleading numbers is subject to all sorts of litigation.

      • vastaman

        I’ve dowloaded over 200 apps, but I’m technical. You believe an average of 82 downloads per iOS device sounds right? Seems a bit high to me.

      • Average apps sold per device sold is 77. You can obtain that by dividing 30 billion apps by 390 million iOS devices sold (through end of Q2).

      • vastaman

        I got 82 per device sold from the 365 million total iOS devices mentioned by Tim Cook in the WWDC keynote. An irrelevant difference if you are correct per the Steve Jobs 2009 precedent.

      • 365 million was as of end of March shipments. I assume that their download data is as of June.

      • Walt French

        But consider the alternative: I’ve updated some apps MANY times. At ~ 80 downloads per device, that’d only run to 20 apps per device. In the US anyway, that number seems far too low.

      • vastaman

        you are absolutely right!

      • “Downloads” here means first time purchase (regardless of price) only.

    • Yes, we do know that downloads figures exclude updates. It’s been known since Steve Jobs stated it in 2009.

  • beidaren

    are the “in-app” purchases included in the app income total?

    • App income data is based on “payments to developers” so I assume it includes in-app payments.

  • Oak

    I can’t speak about the record labels’ contracts, but as an indie musician I can tell you from experience that Apple keeps 30% per song sold, as well. Also, I’m going on record as changing my mind about an Apple television: There will be no screen, and I the Apple TV solution is already here. Set-top-box + content deals (+ Airplay, as desired). The iDevice+Siri is the remote, and also the simple UI that SJ said he “cracked.” So all that’s missing is content arrangements, and maybe some very easy hardware/software refinements.

    • RobDK

      PLUS Apple will expand the iPad line with an iPad HD with 13″ and 17″ screens in 16:9 aspect ratio. Carry it around. Stand it on the table. Hang it on the wall. Take it on vacation….

    • AirPlay, YouTube and a large screen TV is an amazing way to spend a lazy afternoon…

  • It looks to me like the figure titled “Total Apps and Songs downloads per iTunes account” is implying that between 2008 and 2011 that iPhones were being re-used, but that currently (2012) there is something like a 1-to-1 match between iPhones and iTunes accounts. Asa a sanity check, is the source of your data for 2008-2011 different than the source of the 2011-2012 data?

    Thanks for the useful info!

    • The data is the same. I don’t see the implication of re-use. It simply says that the number of songs was constant for any iTunes account. On the other hand, the number of apps per iTunes or iOS device has been increasing steadily.

      • To be clear, I like and agree with your conclusion that the number of apps per iTunes account or iOS device has been increasing steadily, in contrast to the number of songs per iTunes account which is relatively constant.

        My question is about a more minor detail; I was just trying to figure out what it meant that the apps per iTunes account is different than the number of apps per device between 2008 and 2011. For example around Dec 2009 your curves shows that this difference was large, while currently (your most recent data points in the plot) the difference is small. Do you have an explanation for this?

        My assumption was that the total number of apps (the numerator) has to be the same for the two curves, so the thing that can vary between the two is the denominator; Is this assumption somehow incorrect for your data? If my assumption is correct, then for the most recent data we can see that the number of iTunes accounts and iOS devices is about the same. Looking at Dec 2009, we get a much different conclusion.

      • As the number of iOS increased the number of iTunes accounts that were not attached to an iOS device shrank as a percent of total. In 2008 iTunes accounts were used with iPods (or just with PCs and Macs). That total was substantial but now the number of iPhones in use dwarves that. Therefore devices attached to iTunes are nearly equivalent to total iTunes accounts.

      • got it, thanks!

  • Just_Iain

    Another point that will add uncertainty to the calculations is that a lot of iPhone Apps also work on an iPad. I suspect that 20% of the apps on my iPad meet that requirement.

  • Greg Lomow

    Apple has taken an Integrated approach to hw/sw/maps/cloud/ecosystem but seems to be taking a Modular approach to search and social media with a best-of-breed approach by region – Twitter, Facebook, Bing, Google Search, Baidu, Sina, Youku and Tudou.

    Do you see any implications from a Disruption Theory perspective, or do you think this is simply happenstance (Apple just wanted to rely on Google less)?

    • Joe_Winfield_IL

      I’d love to see a thorough discussion on this topic! I can’t imagine this strategy is in any way accidental. Allowing rival search providers is just good common sense; some users prefer different search engines, and Google is not allowed to feel comfortable with its position as the default. Far more interesting to me is Apple’s choice of integrating other web services directly into core phone functionality, removing the need to use a browser altogether.

      Google sees the social networks as a major threat, as their algorithms cannot crawl the data being gathered by Facebook et al. Apple sees this as an opportunity; they can integrate these offerings directly into the phone in a way that Android never can or will. This improves the iOS user experience while also reducing the type of web traffic that Google needs to monetize its mobile efforts.

      Services like Yelp and Open Table are where things get really interesting. Apple is completely removing the “search” step from the process of getting things done. I imagine that the company is charging to direct traffic into the waiting arms of its chosen providers, which happen to be best in breed with large existing user bases. The company is clearly choosing winners and losers, but the hundreds of millions of iPhone users won’t care as long as Siri produces accurate results. Google CANNOT follow this strategy while upholding is primary business model for PC search. It would damage Google’s profitability while simultaneously drawing huge anti-trust scrutiny.

      Finally, the addition of a Siri button in vehicles is a mind-boggling coup for Apple. If Siri becomes the equivalent of “Made for iPod,” it will be game set match. Once it works in a few vehicles, it will quickly become a necessity in all vehicles. It’s almost impossible to buy a stereo today without an Apple licensed dock connector, and just like the dock connector, Apple won’t soon be sharing Siri with its competitors. As Joe Biden would say, “this is a big f—ing deal.”

  • jp

    hi, i think there’s a story to be learned in the scaling up AAPL is performing in Retina display sizing…the 4Q10 Phone 4 at 3.5 inches, the 1Q12 ipad at 9.7 inches, and now the 2Q12 15.4 inch macbook pro retina…seems like retina scaling is possible for a ~40 inch tv in ~3Q13…seems they are scaling up at .55428 per calendar quarter

  • fiftysixty

    Horace, am I correct in my calculations here? You posted on Twitter that average monthly downloads per user are at the level of 5 to 6 downloads per device, and average developer payment per download is $0.166 so the yearly developer revenues per device are $10 to $12, is that right? Has that number changed over time, and is it likely to change?

      • fiftysixty

        Yes, I had a faint recollection of you talking about developers thinking about iOS users as revenue stream. In your linked article, you come to the conclusion that one device equals $0.5 per month, but I arrived at roughly double that figure, so where is the discrepancy? Also, is there a difference between a user and a device in this calculation?

  • Horace, I think you made a mistake about the average song price being $1.29. This is actually the price for the most expensive songs (since a few years ago, when they changed from the fixed price of $0.99), but there are a lot of songs priced below that, and when you buy an album you also get a much better deal per song. So I guess the average price should be somewhere around $1.

    • Although you are right about the average song price, I still think the average song _revenue_ is closer to $1.29 as that is what the most popular songs are priced at.

  • I think a key factor on the music side is that there are other venues for buying songs, whereas apps can ONLY be obtained through Apple: AmazonMP3 (which labels did their utmost to promote as a challenger to ITMS hegemony), eMusic, ripping from physical CDs, and, lately direct purchase from (mostly indie) labels or bands (not to mention the ubiquity of illicit sources). None of that helps Apple’s bottom line of course, but it does suggest the total ‘track economy’ is bigger than the part Apple gets a cut of.

  • Shiraz

    Important to note also that music isn’t available everywhere apps are.

  • This is great. A comparison with the Android marketplace would put things into greater perspective.

    • twilightmoon

      Which Android Marketplace? There’s no singular unique store on Android, plus you can side-load Apps on that platform.

      • For a start even aggregated download numbers would be interesting to look at.

      • twilightmoon

        From what source? How do you get aggregated download numbers for Android that have any meaning whatsoever?

    • It would indeed. If Google published information about Android then we would be able to do this.

  • Pingback: Apple’s Retina MacBook Pro not for the DIY set, plus 4 other Apple stories to read today()

  • Pingback: Apple’s Retina MacBook Pro not for the DIY set, plus 4 other Apple stories to read today | iMicroid()

  • Habeeb Hashim

    You have to take into account availability as well.

    Free Apps are available everywhere.
    Paid apps are available in the vast majority of the countries that Apple operates in.
    Music is offered in very few countries due to licensing restrictions. I’d love to buy music and books but not offered here in Singapore.

  • Pingback: 世上的光 » Asymco数据告诉你苹果应用的现状()

  • Horace critical precision missing in your post: those are only revenue from transactions (paid apps and in app purchase). but the iOS app market is way bigger when you include ad revenues and off store subscription (Eg spotify). The title of your post should be more the iOS paid app market is …or the transaction app market is..

    • This calculation is based on Apple’s reported “payments to developers”. I don’t know if they include payments from iAd.

      • Horace iAd is a tiny part of the multi billion dollar app marketing supported by advertising. Think how many ad networks pump money into apps: admob, millenial, tapjoy, Flurry, SponsorPay, Greystripe, Appsfire and more…Apple definitely does not report that

        Neither the revenues generated off store by online subscription (Spotify, dating sites,..)

  • Pingback: iOS app economy worth US$4.3 billion a year - Macworld Australia()

  • Pingback: Asymco estiment le chiffre d’affaire de l’App Store à 4 milliards de dollars/an()

  • Pingback: Aplicativos para iOS movimentam US$ 4,3 bilhões por ano | Leandro Freitas()

  • Pingback: Jon Lund » Blog Archive » Træthed i App-land()

  • Pingback: Aplicativos para iOS movimentam US$ 4.3 bi ao ano, diz relatório | Consultoria Jr()

  • Pingback: July Hitlist | sayem islam / blog()

  • Pingback: What will Apple do in travel? More questions than answers | Global HDS | Global HDS()

  • Pingback: The brutal truth? Most apps sink without a trace | Richard Hartley()

  • Pingback: iOS App Domination | sayem islam / blog()

  • Pingback: Asymco:数据告诉你苹果应用的现状 « 好应用,好产品,互联网最新产品动向()

  • Pingback: Decoding Apple and Google’s Duopoly in The Lucrative Apps Market | Business()