What Makes a Platform Open or Closed?

#bizarch @nickvitalari sees signs of resurgence at #Microsoft. Everyone Beware: Microsoft is Alive Again and May Become an Elastic Enterprise (April 2012)

Vitalari sees Microsoft's deal with Barnes and Noble as a promising signal, although he admits that it is open to conflicting interpretations. On the one hand, the idea that you can compete with Amazon by getting into bed with Barnes and Noble looks suspiciously like the idea that you can compete with Google by getting into bed with Yahoo. On the other hand, the deal might have the potential of getting Microsoft membership of the elite club that Vitalari and Shaunessy call "The Elastic Enterprise". (See my post on Business as a Platform at Amazon, discussing Shaunessy's recent blogpost.)

Among other things, I was intrigued by Vitalari's comparison between Apple and Microsoft.

"Microsoft was the undisputed leader in the formation of 20^th Century ecosystems, along with Cisco and WalMart. Everyone envied Microsoft’s partnership with Intel and the extensive VAR (value-added reseller) system they created.  But the ecosystem was closed.  Anyone who participated in the old Microsoft ecosystem had to be approved and most importantly it did not have a powerful business platform as seen today with Apple, Amazon or Google.  But that is changing."

Most software historians perceive the difference between Apple and Microsoft the other way around.

"The more restrictive a platform, the less attractive it becomes. Had Microsoft imposed T and Cs as restrictive as Apple’s on Windows software developers, Windows would not have achieved the dominance it did during the 1990s." (Jack Gavigan, Apple's Platform Strategy, June 2011)

"What Apple is trying to do is prevent companies from building phones with multi-touch user interfaces. They want to freeze innovation in the space so that they are the only ones to have the features people want." (John Carroll, Why I (now) hate Apple, ZDNet, 4 March 2010)

"From the beginning, Jobs resolutely resisted efforts to unbundle the Macintosh operating system from the hardware. This contributed significantly to the long-term erosion of both Apple’s hardware and software market share in the computer business relative to more focused players like Microsoft and Dell.  From his perspective, perfection requires tight integration. In working with Disney, Jobs is likely to push for larger, more tightly integrated content – exactly the opposite direction from the one required to maximize returns on content development in a world of attention scarcity." (John Hagel, Disney, Pixar and Jobs, February 2006)   

From this point of view, Microsoft beat Apple because its platform was more open. Some people may have thought that Microsoft didn't care enough about controlling the quality of third party developers, and Apple cared too much - but for many years it was the Microsoft platform strategy that was more commercially successful. Apple's commercial breakthrough didn't come as a result of changing its platform strategy, but extending it - in other words, radical adjacency. Above all, iTunes as the killer app.

But then this is evidence that a proprietary platform, like iTunes or Twitter, will sometimes produce the strongest network effects. Here's Jack Gavigan again.

"In principle, there’s no barrier to building an open version of Twitter. ... But Twitter don’t want that because, in that scenario, the network effects that currently drive everyone to Twitter would be neutralised." (Jack Gavigan, Open vs Proprietary platforms, March 2011)

People talk about different platform strategies, using terms like "open" and "closed", but these terms seem to have worryingly unstable meanings. Platform strategy is clearly a critical question for business architecture, but business architects don't always seem to have consistent ways of describing platform strategy, let alone predicting its consequences.