AGILEVC My idle thoughts on tech startups

September 23, 2008

This is part #2 of a three-part series which is also being featured as a guest column on VentureBeat.

In my first post, I explored the question of whether a fragmented developer ecosystem is ultimately in the best interests of both developers and platforms like Facebook. I wanted to follow-up with some of the implications of a cottage industry-like development community for all parties involved (platform owners, users, developers), as well as a deeper exploration of the evolution of other software platforms.

What’s in the best interest of platform owners like Facebook (f8), Apple (iPhone App Store), and MySpace (based on OpenSocial)? The short and simple answer is presumably “that which maximizes the value of their platforms.” The long and more complicated answer is… well, complicated. One consequence of a largely fragmented developer ecosystem is that platforms have to better police the applications themselves for potential issues of privacy or perceived suitability. Obviously the more fragmented, the harder to police effectively.

Perhaps the most critical implication of cottage industry development for platforms is attracting investment in applications. It seems ludicrous, perhaps, to suggest that Facebook isn’t attracting investment in its platform when companies like Slide, RockYou, and Zynga have raised tens of millions in venture capital funding at significant valuations. And certainly in aggregate, the investment in both financial and human capital in Facebook applications is nontrivial. But the investment in any given Facebook application, whether measured in dollars or software engineers, is fairly small today.

I don’t think the value or success of a given application is based solely on how much was spent to develop it. But if you look at other types of platforms, their success was typically predicated not on how many applications are available for it but rather is there a critical mass of great applications for it. Look no further than the game console realm, where new console generations succeed not because there are hundreds of games available but because of a handful of wildly popular launch titles. And as anyone who follows the gaming business knows, big games that wow users require significant investment to development (example: EA’s development budget for Spore is estimated at $35M).

The aggregate investment in third party applications also matters, to the extent it stagnates or shrinks over time. Whereas Microsoft let lots of large independent software vendors flourish from the Windows platform, Apple often thwarted the efforts of Mac focused developers in the late 80’s and early 90’s. Joel Spolsky has blogged about this, and others, like Randy Komisar, have written about the experience of companies such as Claris that failed on Mac. It obviously took many years for Apple to recover, and ultimately they failed as a personal computing OS and succeeded (wildly so) instead as a multi-product company based upon their design leadership. Windows didn’t become indispensible because of applications from developers we’ve never heard of… it did so because companies like Intuit, Adobe, EA, Symantec, Autodesk, et al built applications lots of people wanted to use. Without a clear opportunity for large independent developers to exist, platforms themselves can fail.

How much do users care whether development ecosystems for the platforms they use are fragmented or concentrated? Directly, not much. But they undoubtedly feel the effects. Consistency of experience across applications certainly impacts users… there’s a reason that if you use Quicken you can usually pickup TurboTax’s user interface pretty quickly. All of us have experienced the effects of cottage industry development on Facebook, whether it’s apps that don’t scale (let’s just say that when I gave up my modem more than 10 years ago, my patience for 30 second page loads decreased) or message spam of various forms. Anything that contributes to a user feeling “application fatigue” can’t be a good thing.

What about developers? Well both large and small ultimately share many objectives: The long run growth and health of the platforms they develop for, transparency between platforms and developers, and the ability to build a business for themselves on a given platform. All want access to and support from platform owners, and obviously the bigger the platform owner gets the more resources they can potentially devote to developers.

It’s interesting to note that most of the companies doing large scale development on Facebook have direct or indirect ties to the company. There are of course exceptions like RockYou, but companies like Slide (Peter Thiel is a board director of both Slide & FB), Zynga (chief executive Marc Pincus is an individual investor in Facebook), Project Agape / Causes (Sean Parker was an early Facebook exec, Joseph Green was Mark Zuckerberg’s former roommate), SGN (backed by Facebook investors Founders Fund and Greylock), and Buddy Media (Pincus and Thiel are investors) all have some form of tie back to Facebook. I’m not trying to be critical here… many of the folks involved in these companies are my friends, former colleagues, or current collaborators — and smart people congregate around exciting ideas. My point is simply that if there were true transparency and encouragement of larger scale development, you’d think there would be more big developers that were wholly unconnected to Facebook. We’re still in the relatively early days of platforms like Facebook, so perhaps in time more large independent developers will emerge.

It seems to me that few parts of the platform ecosystem really benefit from a largely fragmented community of developers. Which begs the question of how could things be different? I’ll address that in my third and final post in the series, tomorrow.