Is Investing in iPhone App Development a "Fool's Errand"?
The iPhone economy may have generated more than $1 billion for developers, but when you break down the numbers as Tomi T. Ahonen has done, it’s clear that creating software for Apple’s mobile devices is a losing proposition for most developers — “a fool’s errand,” he calls it. After reviewing Ahonen’s analysis, which cuts through rhetoric with simple math, I’m convinced that most developers would be better off targeting more mainstream devices such as feature phones and services such as SMS. And while Ahonen’s argument revolves around iPhone development, it applies equally to other smartphone platforms as well.
In many cases, effective software can make up for the shortcomings of today’s mobile browsers and the connectivity they require, but that trend could easily shift as mobile broadband matures in terms of speed and coverage and phone browsers gain offline storage features and other advanced functions. For now, web apps are generally less capable than software solutions, so developers are focusing on mobile software programs. But as Ahonen points out, those doing so for the iPhone are only targeting 13 percent of the worldwide smartphone market:
So worldwide there are 80 million iPhone compatible devices today in use. That seems like a big number. Except, that compared to just the installed base of smartphones at the end of 2009, it is 13%. So if you do any kind of free iPhone app, and intend it to be a mass market media vehicle to reach the pockets of the total population, you have abandoned 87% of all smartphone users in the world today. That seems like a poor idea to me.
Indeed, although Apple has sold tens of millions of iPhones since June of 2007, the company is still a small smartphone player when compared to Nokia or Research In Motion. And Google is fast catching up to Apple thanks to more than 100,000 Android activations daily. But the more important point is how small the smartphone pool is to begin with. As Ahonen notes:
The world has actually ‘featurephones’ which do apps fully well, using Java and Brew and Widgets etc. How many of those are out there in the wild? Try 2.1 Billion. So that moron who approved the ‘brilliant’ marketing idea to develop a cool app, and did it only for the iPhone eco-system, conveniently spat in the faces of 96% of the population with reasonably advanced phones – that all could have easily taken that free app, and engaged with your brand, through a mobile phone app.
In other words, the number of smartphones still pales in comparison to the number of feature phones. Yes, the trend is for smartphones to take more of the market — over half of those in the U.S. are expected to use a smartphone in lieu of feature phone by the end of 2011 — but on a worldwide basis, the feature phone still rules.
Perhaps even more important than the device is the services such a device is capable of using. Nearly every feature phone I know of includes SMS or text messaging, and as Ahonen points out, in 2009 the U.S. version of “American Idol” generated $500 million through SMS voting. With “Idol” available in another 50 countries, Ahonen’s example makes a good case for developers to focus on the lowest common denominator of simple services across more devices: SMS, WAP, and HTML, to name a few.
If Ahonen’s rationale isn’t enough to convince developers to stop working on iPhone apps, perhaps his math is. Using a combination of data points from Apple, Distimo, the Yankee Group and other analytics firms, Ahonen makes the case that for most developers, iPhone app activities are anything but a money-maker. (Related: The Apple App Store Economy) I recommend the full walkthrough Ahonen provides, but here’s a summary of the logic:
- The iTunes App Store has generated 5 billion downloads and $1.4 billion in revenues, which works out to 20 cents per downloaded app for a developer, after Apple’s 30 percent cut.
- After considering that 73 percent of all apps are paid, each of the 164,000 paid apps in the App Store generate an average of $3,050 for a developer in a year.
- Although there are more paid apps than free apps, 85 percent of all downloaded apps are free. Even with an average of 94 apps installed, each Apple mobile devices only generate $14 for developers annually.
- With the low estimated development cost effort of $15,000 it can take a developer up to 22 years to recoup costs from creating a mobile app, using the median revenue of $682 annually.
Certainly there are plenty of success stories in this app economy — between ads and in-app purchases, Tap Tap Revenge was earning $1 million a month in 2009. But I’d love to hear the reaction of some smaller development shops to Ahonen’s insights, especially those that are eschewing iPhone development in order to focus on services that leverage feature phones and simple revenue generators like short messages.
Related content on GigaOM Pro (subscription required):
Something not touched up in this story is app discovery, and how that impacts customer adoption. Its very hard for an application to reach an audience, due to an archaic Appstore search and the shear mass of applications. Simply creating a quality app isn’t enough these days, and you can quite literally see single digit downloads per week without some marketing effort to back the app.
So, why haven’t developers jumped on the feature phone? – too hard to get installed? – too hard dealing with the phone manufacturers? – too hard dealing with the carriers?
How does the math compared to writing programs for desktops/laptops? – Bet it isn’t much different. So, no matter the device/platform, the vast majority of developers aren’t going to make any money (even feature phones). And something like 90% of all new businesses fail.
Anyone considering iPhone development needs to study the situation carefully. No amount of “running the numbers” in a general sense is going to tell you about how your particular situation is going to pan out.
This is all a moving target. There is so much innovation not even tapped yet. Plus, iAds is a new factor and Apple will certainly improve app discoverability ahead of any mass exodus from it’s app store economics. I find the math in both these articles to be completely unrepresentative of the real world because the real world is much more complicated. Of course, don’t wear rose-colored glasses, but do the research and the hard work.
If you’re just going to be a quick hack, whether you’re writing blog articles that repeat information, or an app that repeats existing functionality, you’ll get out of it what you put into it.
Kevin;
Calculating averages in this case doesn’t help, nor does merely calculating ‘median’ as in Ahonen’s longer post. IF the statistical distribution is skewed (not uniform, not gaussian, maybe rayleigh?), the top few percent (that is still 1000s of developers) may be making significantly more money than “average”. That would be the interesting stat — e.g. 874 (random number) developers make more than $50K every year, and the mid + lean part of Long tail doesn’t really make more than $1000/year or something like that.
Granted, you (or Ahonen) may not have the stats and only Apple does… but then, why put out a honeypot headline if you dont have the right stats for the right analysis.
The cost of supporting a large number of Java and BREW devices is quite a lot higher than supporting iPhone. J2ME is extremely fragmented across manufacturers and carriers, and requires extensive testing across a carrier’s handsets. Because of the wide range of phone capabilities, when designing a Java or BREW app, you can’t assume that the phone will have an accelerometer or compass, you can’t assume access to the camera, you can’t assume network access, you can’t assume screen size, etc. So, you play it safe and design a lowest common denominator game.
iPhone represents one of the first smartphones with a large enough install base with a common feature set, there’s a lot of fragmentation even in Blackberry handsets.
This type of fragmentation will likely creep up on the Android side as well — even now AT&T’s Android phones are crippled, and feature crippling has long been a problem with Sprint and Verizon handsets.
@Mike you bring up a great point which Kevin C. Tofel completely left off (poor tech journalism) which is the cost of MAINTENANCE! Applications are not static monolithic bundles of code, they are morphable and require the ability to adapt to the dynamical changing world (customer’s needs / desires might change, etc.). Who the heck has the time / money to maintain the fragmented slew of J2ME Brew devices?
Kevin, don’t get trapped by the marketing presentations of all those non-Apple feature-phone manufacturers:
Yes, there are billions of feature phones out there, but …
… you cannot sell apps to those billions of phones, because there is no AppStore for that … you need to make contracts with hundreds of operators, web based app-shops, mobile portals … so forget it.
… even if you find some operators who will sell your app … don’t even dream of a revenue share of 70% for the developer … the operators will laugh at you.
… these billions of phones have different operating systems, screen sizes, buttons, GUI- and navigation concepts and even if all of them are Java enabled phones … you need optimized code for every model.
But I don’t need to convince anybody … I just say it’s perfect to develop for and sell apps on that platform ;-)
The articles premise is flawed, as the writer seems willing to ignore market trajectory. Nokia bricks are a ghost town and Apple can’t manufacture iPhones fast enough.
Forget the misleading snapshot of today’s marketshare. What does he think marketshares will look like next year, or in the years to follow?
Business isn’t about blind analysis, so much as trend recognition.
I don’t think Kevin or anyone else is confused by the demand-side of this equation. There is a large and growing over-supply of developers. Certainly, there will still be some (many?) who hit one out of the park and make five or six figure returns – but there will be many more whose ROI is paltry.
@Governor — the more I think about this article and the responses in comments to it so far, the more I am convinced that Kevin C. Tofel wrote this article as a troll … he’s just trolling for information to see what people would come up with in response in the comment section (trolling for insights). Don’t give away your insights to Kevin C. Tofel and friends!
Well, Governor, I guess that since the iPhone’s global market share has stayed static at about 16% of all smartphones for the last three quarters and about 5% of all web enabled phones we can pretty much see where it’s going.
One issue with pursuing the statistically larger market share with an app is that other platforms have been incapable of monetizing them well. Hopefully this changes but iTunes gave Apple a significant advantage in having a comfortable purchasing mechanism that customers already accepted and used. Until that is matched in some way the larger marketplace is still a fragmented mess.
I couldn’t agree with you more Bob. The issue of the platform’s size pales in comparison with the issues surrounding distribution of applications for feature phones.
Here you go, Kevin, some more food for thought just posted…
“Android Market Payouts Total 2% of App Store’s 1B” found at…
http://larvalabs.com/blog/android/android-market-payouts-total-2-of-app-stores-1b/
Tomi seems to associate market share with app revenue which is in my opinion a flawed assumption. There may be billions of feature phones but I don’t know anyone who has installed an app on theirs, let alone paid for one (japan and Korea excepted). The story is similar for the smartphone platforms with the majority of marketshare – symbian and RIM. Everyone and their grandma has an s60 phone in Hong Kong, Malaysia, Singapore, etc (Nokia 6700 slide is free when renewing you re contract) yet I don’t know of anyone who has installed an app. Blackberry users probably grab the free Facebook, yahoo messenger, AIM, MSN messenger clients and that’s it. Android has a growing catalog of apps but in general I’ve found them lackluster compared to their iPhone counterparts. I’m sure they’ll improve but at present most of the android users I know have far fewer apps installed that the avg iPhone users.
I agree with his conclusion that it’s a fools errand to develop for iphone though. Too many apps jockeying for attention.