Over the summer I had a really good intellectual workout on technology strategy in my MBA class called Strategy for Technology Media and Telecommunications firms (shortened in class to TMT). One area where I deepened my understanding was in the area of network effects – hard facts and a natural instinctive intuition about it both penetrated the old noggin. On the instinctive side, I came up with an axiom based on what I noticed really quickly: once a network reaches entrenched status, it can never really be unseated unless there is a really major paradigm shift. This shift must be a relatively big and even then, most entrenched networks remain more than vestigial for a long time. Some examples from history include government intervention (e.g. AT&T and Uncle Sam), a dramatic drift in technology (the ongoing saga of the internet vs. the desktop), a major change in standards (html 5 vs. html 1-3) etc. Some of the unexplored inflection points include a change in morality (watch out fossil fuel companies) and other changes in humanity (development of mass telepathy is bound to have an impact if it ever happens). Even with all of this, change is slow and entrenched networks are hard to purge in the real world. Witness the longevity of radio in the age of the internet; the staying power of land lines (75% of homes in the US still have it) in the age of cell phones and the emerging hegemony of Facebook.
I pondered these things as I was thinking about the emergence of Windows 8 (sometime next year) in the age of the iPad. Will this new platform succeed in leveraging the traditional strengths of the Windows ecosystem (massive supplier/oem support, global foot print, embrace of disparate hardware and accessories, etc.)? It’s too easy to look at the Mac vs. Windows 95 and predict that the Windows strategy will rule the day. However in 1994/95 the Mac did not have network effect. Moreover, a massive value of Windows was putting an abstraction layer on really raw hardware in the days before the maturity of hardware APIs (PCI, PCX, ATAPI, BIOS, etc.). In contrast to today, the iPad is well on its way to network effect and hardware modularization (in the age of the SOC) is such that an OS is no longer the same monumental endeavor it once was; which should allow many more OSes to thrive (i.e. keep up with hardware improvements, even if they are not commercial successes).
All these factor point to much lower penetration rates for Windows 8 as a tablet platform as compared to Windows in PCs. It already has formidable competition that is fast gaining network effect. And that is hard to unseat. The best time to attack a network is before it is formed. Once it’s there, the laws of nature take over and all bets are off.
Stages of a network: Formation, Growth, Entrenchment
