How the West was Won (or, another round of “open” vs. “closed”)

John Gruber posted a critique of a Tim Wu piece in The New Yorker. The ideas in the Gruber post are nothing new, but it’s interesting to see this topic come up time and time again.

Wu’s basic argument is “open beats closed,” with the modifier that “closed can beat open, but you have to be a genius.” Unsurprisingly, he enlists Wintel and Google as supporting evidence for why open beats closed.

Gruber specifically disagrees with Wu’s logic for why Windows defeated Mac in the original PC platform wars of the 80s-90s. In Gruber’s view, Windows won not because it was more open, but because Mac innovation had stalled, allowing Windows to catch up. He uses Mac’s brief period of allowing 3rd party licensing as evidence that being more open did not help Apple grow the Mac business; quite the contrary, after Jobs came back and closed off Mac licensing, Apple begun its resurgence.

Over at Techcrunch, Michael Arrington chimes in by stating that the Internet was the unmentioned factor that leveled the playing field for Macs – because the Internet became the core application, it mattered less that Macs had far fewer compatible software.

To me, the question to ask is (and has always been) why did Wintel win in the 80s/90s, and why was the Mac able to stage a come-back in the 2000s. Open vs. closed is simply a popular variation on this core question, because it has been twisted by folks such as Wu to be the critical success factor. It is not.

Wintel was immensely successful due to its leverage of network effects – i.e. the utility of the product grew as more people used it. Microsoft Word is powerful because everyone uses it, and the .doc format is near ubiquitous; if only one person used Microsoft Word, it wouldn’t be that useful outside of creating documents to print.

Wintel was also a two-sided network made up of both hardware/software vendors on one side and consumers on the other. This further reinforced the network effects on the consumer side.

The fact that Wintel chose to be “open” at the hardware layer (can be installed on any IBM-compatible PC) certainly helped drive adoption, but does not itself create network effects. The simple counter-argument is iOS – iOS is certainly “closed” at the hardware layer (exclusive to Apple’s products), but that does not prevent iOS users enjoy the network effects of iOS-exclusive apps (such as Instagram, which for a long time was iOS-only; another example is iMessage, which will probably always be exclusive to iOS).

Pre-Internet, the core application of computers were productivity applications such as Office, and Microsoft Office was (and Lotus 123 etc., before Office) exclusive to Wintel. In a sense, it is a bit of a chicken and egg problem – Wintel’s “openness” to hardware vendors drove OS market share, which in turn amplified network effects of the most popular applications on this platform, which in turn lead to more OS market share. It was a great, virtuous cycle.

It’s hard to say what choices Jobs would have made had he stayed at Apple in the late 80s; it’s a convenient side argument that Apple lost the 80s/90s platform wars due to poor business leadership, however I find this side argument to be often distracting.

Moving on to the late 90s, Arrington is correct in stating that the Internet leveled the playing field. Specifically, as the Internet became the core application, it removed the network effects exclusive to Wintel thanks to Office and other Windows-exclusive software. (Jobs’ successful negotiation to get Microsoft to develop Office for Mac is also Apple’s attempt at leveling the playing field.) And ever since then, network effects have had diminishing influence on PC platform wars – this is the underlying reason why Macs could stage a come-back from low single-digit market share; the beautiful execution (consistently excellent hardware/software iterations) also certainly helped.

As a corollary of this observation (diminishing network effects due to the Internet being the ultimate cross-platform application), we can predict that in the mobile platform wars, despite the seemingly dominant positions of Android and iOS, it is certainly possible for a late-comer (such as the new Firefox OS, and/or other new entrants) to enter the market and capture significant value. However, the success or not of those mobile OSes will not be determined by whether they are “open” or “closed” – by that measure, we can certainly already declare Firefox OS as the winner. The next few years in mobile will be very interesting.