Gartner predicts the consumer apps’ market will grow to $29 billion by 2013, fulled largely by free apps growth. This however is in the context of yet more news of the online content industry moving towards paid content models, signaling the slow death of free.
Gartner sees advertising sponsored apps growing to 25% of total apps revenues by 2013. I don’t have a view on advertising growth but I think Gartner could well be underestimating future, overall apps revenues if we take into account potential growth in areas like the auto, however wrong they seem to have got the revenue model. We have referenced before though, considerable informed scepticism about the value of apps with Tomi Ahonen appearing to dismiss them as irrelevant. Here are some Gartner extracts:
Consumers will spend $6.2 billion in 2010 in mobile application stores while advertising revenue is expected to generate $0.6 billion worldwide, according to Gartner, Inc. Analysts said mobile application stores will exceed 4.5 billion downloads in 2010, eight out of ten of which will be free to end users.
Gartner forecasts worldwide downloads in mobile application stores to surpass 21.6 billion by 2013 (see Table). Free downloads will account for 82 per cent of all downloads in 2010, and will account for 87 per cent of downloads in 2013.
“As smartphones grow in popularity and application stores become the focus for several players in the value chain, more consumers will experiment with application downloads,” said Stephanie Baghdassarian, research director at Gartner. “Games remain the No. 1 application, and mobile shopping, social networking, utilities and productivity tools continue to grow and attract increasing amounts of money.”
I took part in an exercise recently to estimate the potential scale of the enterprise apps market over a similar timeframe. The figure was roughly double the Gartner estimate for consumer apps. Neither take account of how an open source movement, see for example Michael’s post this morning, around apps might affect the market. Of course issues like app discovery and the contribution of apps to store holder revenues remain problematic – the proportion of free apps will increase according to Gartner from 80% today to 87% and we’re looking at a tenfold increase in downloads.
This is the area where I think Gartner are most off the mark. I think we are seeing a decisive move against free – with one more newspaper converting to paid content: The New York Times. How long before a normally functioning economy impacts the web and the mobile apps scene?
I’d be interested to see if Gartner have any serious numbers behind that headline data. It looks like they’re making the same mistake as a recent GigaOM article that got corrected… if there were 2.5 billion downloads and 80% of them were free then we’re only talking 500 million paid downloads for $4.2 billion of revenue… or more than $8 per app average???!!!!! I don’t think so – particularly not when they predict advertising will only grow to 25% of revenue by 2013.
Or are there some other major monetisation methods beyond paid apps and advertising that I’m missing?
If the Yankee Group reckons $343 million for all apps sold in all app stores then we’ve a long way to go to get to $4.2 billion for this Gartner estimate or there’s a lot of room for error in estimation of something that’s already happened?
I reckon Yankee Group are right for 2009 – but 2009 data is dominated by the iPhone You’d have to assume that other apps’ initiatives will get their act together,and you’d expect the arena to broaden considerably (and I’ve seen enterprise apps charged out at Euro 80 per pop – so I am more confident about enterprise apps). But yet I think you have a good point – you can’t be free and make a ton of money.
Oh yes, I agree, there’s going to be massive growth in this area. I’d just like us to present some realistic data to potential app developers.
This looks like a classic case of garbage in, garbage out to me. Take a figure that’s an order of magnitude out now, apply some seemingly reasonable but somewhat arbitrary growth curve to it and produce a figure that’s still about an order of magnitude too big for 4 years time…
Personally I’m a bit more optimistic on the growth curve than Gartner, although extremely skeptical of their input data (or basic math) and agree that you’d expect a significant uplift from the other companies getting their acts together. However I suspect $29 billion is still too high for 2013.
Regarding 80 euro per pop enterprise apps, I’d agree that there’s likely a lot more value in individual enterprise apps, but the target market and total number of apps is likely to be smaller, so I’m not so sure it’ll scale that fast. If you include bespoke development I might concede your figure. It’s worth noting that the iPhone app store figure includes a very decent chunk from the likes of TomTom navigator at around $99 (IIRC) though and allegedly iRa Pro (video surveillance app – you probably want to count that as enterprise) that was about $900 took a reasonable chunk by revenue too.
>>I’d just like us to present some realistic data to potential app developers.<<
Well I guess when you have "garbage", or guesswork, out there the best route to quality is this kind of debate, or dialectic, Mark. It would be great to be resourced enough to produce that data but as we are not then we're here in the spirit of trying to fathom out what we should all believe and trust in.
I wouldn't be so swift to label it garbage though – defining what will happen in any market is a tough business. yes, better if Garner acknowledged that and maybe invited more discussion.
>> Well I guess when you have “garbage”, or guesswork, out there the best route to quality is this kind of debate, or dialectic, Mark.
Indeed, which is why I thought I’d wade in.
You very appropriately phrased your title as a question so I thought I’d point out that the basis for this particular piece of 3rd party analysis is fundamentally flawed.
My “garbage” label refers only to the relatively well known GIGO term used in mathematical modelling. I have a relatively high degree of confidence that the Gartner 2009 revenue figure is very wrong because of the (probably approximately correct) 2.5 billion downloads, more than 2 billion of those are from the Apple store. Although Nokia’s Ovi is probably number two now on a daily download basis, for most of 2009 that place was held by GetJar, which is 100% free. Everyone else adds almost nothing to the figures. I’ve seen loads of “average” prices for Apple App Store paid downloads and they’re all under $3, some very significantly so. It looks a bit like someone forgot to factor in the 80% free figure…
Yes, predicting the future is very hard and very smart people regularly fail to get close. It’d be nice if they all sanity checked their figures before publishing though.
I’m still wondering where this dubious $343 million number comes from. Did you get that number from Tomi Ahonen’s blog??
All I was able to find is a *prediction* for 2009 by Yankee Group of $342 million in sales and for the United States only (Apple, RIM & Android.)
So I think it’s fair to say that this number is far off the real sales for “all” app stores world-wide in 2009. I’d even say it’s far off the real sales for the U.S.
Really, think about it for a while. $343 million in sales for a full year!?
343 companies in the mobile app economy world-wide producing sales of $1 million on average? Come on. Doesn’t make sense.
By the way, Gameloft just reported 10 million paid downloads in the AppStore since it’s launch 18 months ago. That alone might well be $33 million a year. Gameloft claiming 10% of the the mobile app sales worldwide!? No way.
There are other estimates about the sales generated by apps in 2009. Those figures are a lot higher – into several billion US$.
From my personal experience, those figures seem to be closer to reality. $4.3 billion for 2009 by Gartner as you quote? Yes, perfectly making sense to me.
Re: the $8 per app question, check out the ranking by revenue for the Apple AppStore. Average price for an app in the current TOP 10 in Germany: ~ EUR 25,-.
Hi Jan Ole – I think as remarked above that the data is skewed by people thinking APP = Apple App Store and the Yankee Group figures seem to me to have some credibility in terms of Apple. I am however repeatedly amazed by how US based analysts believe the mobile space is America and its leading edge is the west coast.
I have a question while I am on the way to responding – will you write abut Gravity for us? I’m really keen to get as many Symbian developer stories into our record here as possible and would love to hear about Gravity and the others you’ve created – and of course any you have in the pipeline.
How can we usefully shape this discussion? It seems to me that the enterprise scene will be larger than the consumer scene but that the consumer scene will be big. People spend on aggregate huge sums on ringtones so why not on Apps? Has though the Apple model spoiled people so they now expect free?
How could we model that type of argument and its influence? On consumer apps, will the reversion to paid content online INCREASE the proportion of paid apps so we’ll be looking at 75% free not 87% as per Gartner.
Will apps stores learn to protect average revenue from their store products – ie act like a responsible store?
Apart from shee volumes these are important questions shaping revenues.
This is from today’s techcrunch, by the way:
There is no doubt that the iPhone and iPod Touch are glitzy game machines (just check out the ads for the IPod Touch). But don’t underestimate the reach and earning power of apps based on the lowly text message. New York City-based Snackable Media sold $170 million worth of text-based games through carriers in 2009, according to CEO Eyal Yechezkell.
Hi Haydn,
> This however is in the context of yet more news of the online
> content industry moving towards paid content models,
> signaling the slow death of free.
This is a potentially naive assumption (who can say at this point?). The alternative is that companies who put themselves behind a pay-wall will find that with the likes of the BBC and various news-wiki sites they simply won’t get enough paying customers to support their business model. It’s a completely different discussion as to whether this is a good thing or not.
As for apps I agree with the folks at the Stammtisch who made the comparison with PC apps. In the mid 90s downloading PC apps was a huge business. These days how many people actually install any paid-for apps for their PCs (games aside)? It seems to me that most things people used to pay for either became part of the OS, now get shipped free on with the hardware (Office) or have been largely replaced by freeware.
Yes – there is plenty of potential in bespoke enterprise software but I think there is a fair chance of the app market for mobile becoming very flat in after the hype of, say, the next 5 years as most of the important functionality ships as standard with the devices.
Oh course this still makes the next 2-3 years rock ‘n’ roll time for people producing apps…
Or apps purchasing could become as automatic and trivial as ringtone purchasing or apps functionality could become progressively more innovative and interesting. That’s the guesswork isn’t it David….. what we’d like to find are some less dispensible parameters to attach to these arguments.
@Jan Ole – for the record, I also think the $343 million figure is too low, although if it’s purely for the the US that makes sense.
Following my very basic calculations from published figures and others estimates for average downloaded app cost, I’d say the figure is likely to be somewhere between $1-1.5 billion (~500 million paid downloads at $2-3 average price).
Gameloft’s 10 million paid downloads is 2% of the total at probably slightly above the average price, so they maybe have ~3% of the global market. As one of the world leaders in a the number 1 paid download category, that doesn’t sound unreasonable to me.
The average price of the top 10, 25 (as per GigaOM), or even 50 by revenue isn’t likely to reflect the overall average price well (because of the truly immense volume of $0.99 downloads) but it does tell us that charging more for a quality product is likely to get you more money than selling cheap rubbish. I’m not sure whether we have enough data to say it makes enough to justify the extra costs involved in building a quality product, but my instinct (and hope for the future of the industry) is that it does.
[...] Yesterday, Gartner released their projections on mobile application stores. They say there were 2.5 billion downloads in 2009 and there will be 4.5 billion this year and 21 billion in 2013. Remember, these are only forecasts. Even as a mobile developer, I find 21 billion very hard to believe. This seems high to me. Meanwhile, Haydn on the Symbian blogs thinks this value is too low. [...]
[...] were obviously huge in 2009. Analysts seem to think they will still be in a few years but on the other hand web browsers are getting better, so we may see a swing back towards the web. [...]
Quick follow-up on the numbers. Here’s the link to the $343 million cap of the app market:
http://www.yankeegroup.com/ResearchDocument.do?id=52164
It’s a forecast for the U.S. only dated September, 10th 2009.
It’s a pity that this number has been used in a “misleading” way on another blog.
Anyway, it shows that numbers are sometimes used to spark interest or to prove a point.
For Symbian, there are two aspects that I’m missing:
1.) There has been a 3rd party “app economy” around for years – long before the App Store was opened.
2.) There is no centralized app store necessary in the Symbian ecosystem. Anyone can sell directly. You can even open your own “AppStore.” That’s a big advantage over the iPhone ecosystem.
Of course, this makes “number counting” a lot more difficult – with all the consequences of scarce press coverage, less user attention and fewer developer projects.
It would be nice if the Symbian Foundation could come up with some estimates on how big the Symbian 3rd party ecosystem really is. Would that be possible? Asking some of the research companies to have a real look at the Symbian ecosystem?
[...] were obviously huge in 2009. Analysts seem to think they will still be in a few years but on the other hand web browsers are getting better, so we may see a swing back towards the web. [...]
There has been a 3rd party “app economy” for Symbian, but for a number of reasons this has dwindled enormously. A couple of reasons (not exhaustive):
1) rampant piracy, including Nokia’s own Mosh becoming the biggest crackware distributor in a very short time.
2) greedy resellers. Handango was moving to a 70-30% split, with 30% for the developer.
3) UIQ 3.x failing. Most apps were sold on UIQ 2.x, which was at some point in time the best selling platform on Handango. WHen 3.x failed, that market evaporated.
4) The end of S80. S80 was the second biggest seller for Symbian apps, even though in device numbers the platform was tiny.
5) The lack of a proper brand. By the time S60 got some traction in the market we also got the Symbian 9.1 and people could not move their apps to the latest devices, reducing the stickiness of the S60 brand.
6) Famously hard to program for.
7) Rubbish on-device app store clients.
Happily Apple came along at the right time with their SDK fixing all those issues (apart from the piracy).
I also wouldn’t call having multiple App Stores a boon for the Symbian platform. The more App Stores, the more work it is to get your apps in those stores and the more red tape there is to cut through before your app is in those stores. Consider for instance a “Tell a friend” function, where you can email your friends and colleges about a very good app. If you have 10 app stores, they will all want a version of your app that has a link to their App Store in that email.
Hi Sander – those are important observations. Would you consider writing a blog post for us to expand on your experience?
Jan Ole – I’ll raise that internally. I’m sure seen as a service to the eco-system it should be something that gets a bit of take up!
Haydn, this is all water under the bridge, apart from the App Store bit. I have already proposed there should be an App Store API that hides the issues related to having a lot of App Stores, so you can buy an extension of tell a friend about an app and the API filling in the App Store details.
The current really important issue is that there’s going to be a split between the API’s for plain Qt on the one hand, and the Maemo 6 and possibly Symbian^4 versions on the other hand.
With Maemo 6 the idea is to replace all high-level Qt classes like QApplication and QWindow with Dui variants (DuiApplication, DuiWindow), you’ll see immediately that this destroys source compatibility, and if the Symbian^4 proposal goes the same route, then we will have three trivially different sets of high-level API’s doing the same thing. This is exactly what happened with Uikon splitting into Avkon/S60, Quikon/UIQ and S80/S90, and as a result it became much harder and much more expensive to write Symbian apps.
If Symbian and Nokia are smart, they keep plain Qt classes for everything except for those UI components that are unique for the platform itself. So If I want for instance to get a number from the user in a dialog, I have a class that reads numbers. At the API level I don’t care about the looks of that number editor, looks is something for the platform UI to worry about.
Having a unified API will ensure that Symbian^4 developers have a straight route to Maemo 6, and vice versa. It will also enable them to take advantage of other Qt enabled platforms, and this makes Symbian more attractive.
[...] Haydn touched upon this issue last week by referencing what I call “platforms” as “Type 1” propositions and “infrastructure” as “Type 2“. This distinction is quite important because the audiences for platforms and infrastructures are quite different and entail different strategies in the operating system and the brand projection (see also the comments sections on the blog here and here). [...]