AGILEVC My idle thoughts on tech startups

September 24, 2008

This is the final installment of a three-part series which is also being featured as a guest column on VentureBeat.

My first two posts in this series examined the fragmented nature of the development community that has evolved around Facebook and other platforms, and then addressed the implications for platform owners, users, and developers. In this last post I wanted to touch upon what Facebook might do to foster larger scale developers and how things might evolve in the future. My focus on Facebook here isn’t so much to pick on them… platforms like the iPhone, perhaps Twitter, and others may face similar issues someday. But Facebook is the clear leader at present among emerging software platforms (Google hasn’t counted as “emerging” for a long time), and Facebook has the stated ambition of becoming an “operating system” for the internet it’s worth focusing on them.

One idea would be for Facebook to take algorithmic-based approaches to the promotion of various applications, whereby those apps that end up with the most usage are highlighted in app directories or news feeds, instead of simply selecting “Great Apps” by whatever process is used today. Most usage could be measured in a lot of different ways (active daily users, installs, time spent interacting with app, growth rates, etc) and it’s reasonable to debate which might be most appropriate. Google has obviously taken this approach to search… the collective input of users is algorithmically processed for everything from PageRank (links users create on the broader web) to AdWords placement (which ads users click on most frequently). Of course Facebook app developers might try to game algorithms just like web developers try to game Google through search engine optimization techniques. But it would seem that an algorithmic approach would be both fairer and more importantly would ensure that the best apps (in the eyes of users) rise to the top.

Facebook could work with large developers differently than smaller ones. Again, I’m not “anti small developer” but Microsoft doesn’t treat very large Windows developers in the same way it treats very small ones. In many cases large developers enjoy greater access to new technology, more integrated marketing support for new products, and other benefits. Microsoft still spends resources on both large and small developers and more importantly lays out its various programs with reasonable transparency. Mature as the Windows platform may be, there are still thousands of small developers today right alongside the large ones. So while they encouraging large scale developers that are critical to the success of Windows platform, they still support a vibrant network across developers of different sizes.

There is one key difference worth noting between platforms like Facebook versus software operating systems or gaming consoles. Both OSs and console revenue models were built largely (if not solely) on software licensing revenues as opposed to being ad-based. In the case of OSs, the fact that people bought apps from Intuit, Adobe, et. al. meant that in the long run more Windows licenses would be sold. For gaming consoles, the console owners receive a cut of every game license sold.

In a world where both the platform and the app developer generate revenue primarily from advertising directed at the same set of users, the issue is less clear. Yes, virtual goods and commerce type models are growing, but these still remain the minority within the broader Facebook ecosystem. Surely the pot of potential revenue is vast enough to split somehow — assuming Facebook emerges as a huge platform (that they are a huge social network and destination website isn’t in question). Microsoft developed the biggest game franchise for Xbox internally, including the Halo series, after acquiring Bungee in 2000. Yet, there’s still ample room for EA, Take 2, Activision, and others to still exist. In my opinion, Facebook should at least permit larger scale developers to flourish if not actively encouraging them given the symbiotic relationship. It’s the trade-off we as venture capitalists often discuss with entrepreneurs around the tough issue of dilution… the choice between owning virtually all of a small pie versus owning a smaller (though still very substantial) piece of an immense pie.

Speaking of venture capitalist, a handful have launched specific pools of capital focused on Facebook developers (parallels can be drawn to Kleiner’s iPhone fund). Only time will tell obviously how these investments play out, but certainly Accel, Founders Fund, and Bay Partners are all successful firms and should be applauded for forwarding the thinking here. But if Facebook truly succeeds as a platform in five or ten years we won’t be talking about specialized funds allocated to Facebook app developers. Instead, this will cease to be a novel investing approach and virtually every venture capitalist who focuses on the internet and software sectors will simply have companies in their portfolio developing in part or maybe even exclusively for Facebook. That will be one of many true benchmarks of the evolution beyond cottage industry.

While I may have raised some critical questions in this series of posts, let’s take a step back for a second and acknowledge the success Facebook has achieved with its platform. Within three months of launching it, nearly 2/3rds of users on the site interacted with at least one app, and the usage has only scaled dramatically since then. You can’t draw a direct analogy of course, but Microsoft spent years and billions on the Xbox platform to achieve comparable user adoption. Yet extraordinarily few startups in our lifetime will have the opportunity to become a dominant platform in the way Facebook might. I truly hope they seize that unique opportunity and play a important part in evolving their developer ecosystem beyond a cottage industry.