Innovators who build platforms face a difficult set of trade-offs. To begin with, the very definition of a platform requires buy-in from others — and not simply from the consumers you hope will purchase your product. The consumers and suppliers of complementary products need to make real investments too, investments that will enhance the overall value of the platform. So the platform owner not only has to build a base of consumers who will use its platform and buy products, it also has to manage the relationship with complementary suppliers and their consumers. Google, Twitter, and Apple have each taken a different path in handling this balancing act.
Perhaps the best example of such a complementary supplier is the app developer that provides software programs that you might download on an iPhone or iPad, just like Coke and Huggies supply products that make supermarkets valuable.
But if a supermarket chain fails, their suppliers have other options (other supermarkets, for example). This is not so for many of our current technological platforms. Consider the failure of Google Wave, Google’s collaboration tool, launched in 2009 and killed in the summer of 2010.
In some circles, Google was applauded for experimenting and not being afraid to admit failure. But Google, as a platform innovator, suffered reputational damage as a result of the failure. The company had opened that platform to developers to supply modules that allowed greater levels of interaction among users. Developers had invested in Google Wave, building on the back of Google’s seeming commitment only to find themselves at sea without a platform. The next time Google touts a new platform, it will find it harder to entice developers.
Twitter’s and Apple’s platforms also need developers to improve them, but they are walking a different knife’s edge. Twitter encouraged developers to make clients so that users could use Twitter in different ways, giving rise to a healthy ecosystem of different ways of accessing Twitter, each of which suited a different kind of user, leaving the fundamental structure of Twitter unchanged.
But then Twitter acquired Tweetie — the most successful iPhone client — renamed it the Official Twitter App and re-launched it free. Now the platform owner had integrated and siphoned off the potential earnings of the other client developers. To be sure, those developers might react by stepping up their game, which would have given Twitter’s strategy some sense. But equally, developers might stop developing altogether, harming consumer variety and innovation. When a few months later Twitter acquired the popular desktop client Tweet Deck and enacted other new rules limiting development access, it became pretty clear that Twitter had reversed its open platform course.
With over 400,000 apps in the iTunes App Store, it would be hard to accuse Apple of a similar move to foreclose on app development. But time and time again, Apple has shown itself willing to take successful independent apps and integrate them into their operating system. It did so with iBooks and its Voice memo app. More recently, it embedded HDR functionality straight into its own camera (something that independent camera app makers had already developed) and iOS 5 will include a “read it later” function for its Safari browser. This prompted Instapaper founder Marco Arment totweet an expletive. (Arment thought better of it all a little while later and wrote: “Glad I’ve invested in social and editorial features,” he said. “Not dead yet!”)
We should think a little more about why Apple and Twitter have made such moves. Here is my theory: successful platform owners will defend their turf against platform strategies in their own backyards. They are looking to avoid a fate like IBM’s loss to Microsoft. IBM let Microsoft own the operating system and build its own platform tied to IBM’s new one. We know how that ended up. Apple and Twitter likely fear the same outcome. They won’t worry about Angry Birds having lots of users. The game is short-lived and it is unlikely a game developer will control the platform. But popular clients, especially ones with social features, are another matter. If these kinds of clients become too popular, platform ownership could transfer to them.
Perhaps the best case in point is what happened when independent game developers got social and set up Open Feint and the +Plus networks on iOS. These were networks that signed up game developers and allowed users to share scores and leader-boards and even challenge one another to games. It was a clever strategy to leverage popular individual games into a broader network.
It turns out it was too clever — at least too clever for Apple. Last year, Apple integrated its own Game Center right into the operating system and into the Software Developer Kit. Game Center did what those other two networks already did, and came with Apple’s seal of approval. While the other game networks limp on, Game Center appears to rule the iOS world. Apple has effectively defended itself against a platform ownership challenge.
The defensive strategies of Apple and Twitter do walk a fine line. Just what these platforms should and do regard as a threat isn’t transparent. In such an environment, there is every chance they might chill valuable innovation by independent developers as they try to control their platform’s evolution.
Digitopoly 
Hi Josh, I just discovered this blog, two months after the fact, now that you announced it on Google+.
I like this post and totally agree with your concerns. I’ve been following these events too and had a similar reaction. My thinking on this goes back to the “divided technical leadership” theory of Bresnahan & Greenstein (1999) (i.e., your co-blogger Greenstein). The part of their theory that jumped out at me here is the idea that complementors, rather than new entrants, may pose the greatest threat to the platform leader. This threat leads to the kind of defensive behavior you describe.
Of course, as you mention, platform leaders also want to encourage complements. The possibility that the “chilling of innovation” could turn out to be harmful to the platform leader is something I modeled in Miller (2008). In their book, Gawer & Cusumano (2002) discuss, but don’t really model, the possibility that the platform leader can try to develop a reputation for not stepping on the toes of its complementors. Finally, the idea that a monopolist in one market would want to extend its monopoly to related markets as a defensive strategy also shows up in Carlton & Waldmann (2002), although as I recall they focus on moving into markets for substitutes rather than complements.
Yes, it is rather cheap of me to plug my own paper, and lazily list a few other papers from its bibliography. It is so closely related, though, I just couldn’t resist.
Uh oh, looks like I forgot a tag in my prior post. That link to Bresnahan & Greenstein (1999) is huge!
Uh oh, looks like I forgot to close an “a href” tag in my prior post. That link to Bresnahan & Greenstein (1999) is huge!