FernBlog

Application development for the Apple iPhone and Facebook has skyrocketed. In a July 14 press release, Apple revealed that the App Store contained over 65,000 applications with over 100,000 developers participating in the iPhone Developer Program. The Apple press release further noted that 1.5 billion applications have been downloaded. Meanwhile, Facebook reports that there are 350,000 applications currently available in the Facebook Applications Directory with over 950,000 developers currently working on application development.

What are the implications of this significant application development for Apple and Facebook and the myriad of developers creating new applications for these platforms?

In many ways these figures are fantastic news for both Apple and Facebook. They suggest a thriving market of developers providing complementary software that will increase the value proposition of these companies’ respective core products (the iPhone/iTouch and Facebook) for target customers. An increase in the number of developers and applications promotes variation in the types of applications that are developed, which in turn increases the probability of an application being developed that is a homerun (i.e., the application gains broad adoption). This variation also suggests that applications will be developed that address the needs of the long tail of consumer preferences. There are indirect benefits as well. For instance, the significant level of application development has generated tremendous free publicity for both companies.

If this application development is good news for Apple and Facebook, what are the implications for application developers? Will a number of developers survive and prosper in these two markets?

There is precedence for this fervent entry and development in new product markets. From the perspective of technology evolution, this stage of development is known as the “fluid phase.” In this fluid phase there is substantial uncertainty about the technology and its market. A significant number of new entrants enter the market and experiment with different product features to assess market response. This process continues until a dominant design is established and a shakeout occurs. During this shakeout period, a vast majority of the entrants are forced out of the market. Some are acquired, some turn to other industries, but a vast majority exit.

As an example, let’s consider the automobile industry. According to research by Dobrev, Kim, and Carroll (2002), over two thousand producers of automobiles entered the industry from 1885-1981. These producers experimented with a variety of product configurations and features, with considerable variation in the source of power (steam, electric, combustion), the systems of steering and control, modes of stopping, and so forth. There was a dramatic increase in the earlier period of the industry, with the number of producers increasing significantly from 1893 to the mid-1920s (Suarez and Utterback, 1995). After the late 1920s there was a sharp decline. This decline was marked by some consolidation, but a preponderance of the startups disbanded. In the end, the American automobile industry consolidated to just the big three automobile companies: General Motors, Ford, and Chrysler.

iPhone and Facebook applications are not automobiles, and thus this analogy can be taken only so far. The barriers to entry for developing an iPhone or Facebook application are much lower than producing an automobile, both in terms of capital investment and technological uncertainty. It takes just a few hours or days to develop a new iPhone or Facebook application and it’s relatively easy to apply existing software development skills towards application development on these platforms. The low barriers to entry explain in part why there are 100,000s of application developers, and not just a few thousand as was the case in the automobile industry. This also explains why developers flooded the market at a much quicker pace than they did in the automobile industry. It took a number of years for the automobile industry to takeoff as technology and infrastructure improved. In the case of Apple and Facebook applications, much of the technology and infrastructure is already in place.

Despite these dissimilarities, what’s clear is that like new entrants in the automobile industry, a vast majority of Apple and Facebook application developers will disband because they are unable to develop competitive products that support an ongoing business. How many firms will ultimately survive and prosper? If we define a business as a company with at least one full-time employee (as opposed to someone developing Apple or Facebook applications as a part-time hobby), I suggest the figure is exceedingly low. Consider the fact that if just 1% of current Apple and Facebook application developers survive, this amounts to 1,000 and 9,500 companies, respectively.

To shed further light on the issue, let’s conduct a rough scenario analysis about application development for the iPhone. First, let’s make a few assumptions. Assume 1,000 application developers are competing in this market, 1.5 billion applications are downloaded annually, consumers pay a price of 99 cents per application, Apple gives developers 70% of sales, and revenue generated is equally distributed to all developers. In this scenario, each developer would generate $1.05 million in annual gross revenues. $1.05 million is interesting, but not particularly exciting, at least from the standpoint of a high-growth technology venture.

Revenues are never equally distributed and further many iPhone applications are being given away for free. Accordingly, let’s apply two additional reasonable assumptions. Assume that a handful of top developers eventually gain a non-trivial share of the market (e.g., 25% of all downloaded applications are created by 10 firms). Furthermore, assume that half of applications are given away for free, a generous assumption given that the top 16 applications in the App Store are currently free. With these assumptions, the top 10 firms would generate an average of $13.125 million and the remaining 990 developers would generate an average of $398,000 in annual revenues. In this scenario, a few firms generate a bulk of the revenue and the remaining firms have revenues that would support just several full-time employees. Recall, these numbers are based on the assumption that just 1% of developers persist over time.

If we push the number of developers to 10% (10,000) of the current developers (100,000), then the annual revenue for the average developer (not in the top 10) drops to $39,800. At 100% (100,000), the annual revenue for the average developer would equal $3,980.

In short, the economics of the iPhone application market in its current form could not support more than 1% or 2% of the existing 100,000 developers. To support a greater number of developers, application downloads would need to increase substantially (e.g., if Apple licensed the iPhone OS to other hardware producers) or consumers would need to pay significantly more for each downloaded application.

Anecdotal evidence bears out the challenging economics of competing in iPhone application development. As of May 24, 2009, the company Stromcode developed 20 iPhone applications. Six of these applications were ranked in the top 100 in various categories, such as social networking and board games. Stromcode is charging 99 cents for two of the six applications, and the rest are free. One of the paid applications is generating $20 a day and other $4 a day, yielding a total annual revenue of $8,760. This makes for a nice hobby, but certainly not a business, much less a high growth venture that might interest a venture capitalist.

In conclusion, while a very limited number of companies will find financial success developing applications for these two platforms, a vast majority (on the order of 98% to 99%) of developers will be unable to develop and support viable businesses in these markets. The 1-2% that succeed can expect an upside of around $10-20 million in annual revenues. This isn’t quite a lottery, but it’s depressingly close.

Update on Thursday, August 13, 2009 at 3:15PM by Michael J. Fern

I have received a variety of comments both here and elsewhere regarding this blog post. Let me expand on my post in several ways, drawing on some of these comments.

First, it’s been suggested that my post misses the point that not all developers for Facebook and the Apple iPhone should be classified as new ventures. I agree, and I could have been more clear in my post. There are at least three types of developers: a) hobbyist/part-timers that are developing applications in their spare time; b) established firms that are porting existing applications over to Facebook or the iPhone for the purpose of expanded distribution or marketing; and c) new ventures that have been founded to develop Facebook and iPhone applications. My post speaks most directly to this last type, new ventures founded to primarily focus on Facebook or iPhone application development. These are the firms that are depending on generating enough revenue from their Facebook or iPhone applications to support and grow a software business. Given the current economics of the Apple App Store (and probably the Facebook platform) I think the outlook is bleak for most of these firms. Of course the outlook would improve if customers’ willingness to pay for applications increased, if paid application downloads increased, or if application revenue was augmented through other sources, such as advertising or in-application purchases.

Second, some have expressed concerns about the scenario analysis. The analysis is admittedly rough, as noted in my post. I based the analysis on several assumptions, most of which were informed by publicly available information; e.g., the July 14 Apple Press release, App Store revenue allocation, the App Store top selling application lists, etc. Some of the assumptions are rough educated guesses: the percentage of apps that are free versus paid (I assumed 50%, which is a very generous assumption) and the eventual concentration of the market (10 firms controlling 25% of the market). I believe the scenario I presented is plausible, if not optimistic, in view of current information. Of course there are many variables to consider and the landscape is evolving quickly, so the scenario I presented might miss the mark considerably. For instance, as noted above, application downloads might increase significantly, willingness to pay might increase, etc. Despite this uncertainty, I believe the analysis is informative because it shows how the current economics of the App Store can not possibly support anywhere close to 100,000 developers, or even 10,000, unless a vast majority of these developers are hobbyists or established firms not relying on these applications for revenue generation. Furthermore, it suggests that very few, if any, new ventures focused on iPhone application development will realize the top line revenues that will satisfy a professional investor. In the scenario I presented just 0.01% (i.e., 10/100,000, or one hundredth of a percent) of current developers participating in the iPhone developer program will realize $10-15 million in annual revenues. This is striking given the number of venture capitalists that are financing companies dedicated towards iPhone (and Facebook) application development. Admittedly, venture capitalists will fund the more competent ventures that have a higher chance of success, but what VC would have guessed that the top free and paid applications for the iPhone would include titles such as iFart, the Moron Test, and Mirror Free.

Let me be clear, I do think there are opportunities in these markets, especially if we move beyond just the iPhone and Facebook and consider the larger mobile and social networking markets. We have to remember that the iPhone and Facebook control just a portion of their total respective markets. Rather than focus on any one platform, the new ventures most likely to succeed over the long-term are those that develop excellent software that are platform agnostic. With this approach, new ventures will avoid the risks of pinning their fate to any one platform. Furthermore, these ventures will launch and distribute products to a much larger addressable market.