Core investment theses (IV) - De-verticalization

The fourth and last thesis is the least developed, and one where we are probably less predisposed to play in, because it tends to attract a lot of VC interest with greater relevant expertise and because it can be very capital intensive.

The key argument is that as industries emerge they tend to do so in a vertically integrated fashion and as they mature they tend to de-verticalize. For example, when the beer industry was going through its growth phase, many brewers owned glass-makers because they needed to guarantee supply of glass bottles, and they were also one of the major buyers of glass; however, once the industry matured, they spun out the glass companies since they were no longer core and they depressed their returns on capital.

Closer to the mobile tectonics world, the computer industry grew up highly vertically integrated - notably IBM controlled a vast proportion of that value chain. In some measure, the past 30 years of computer industry innovation can be seen as a gradual de-verticalization of IBM, leading to the emergence of many sub-industries, and emergence of powerful new corporations (Microsoft, Dell, Intel, etc).

Similarly, the mobile industry was highly vertically integrated for many years. The big OEMs did a lot of their work in-house, and the operators also did everything (network, customer care, IT, stores, etc). During periods of high growth and high margins, such as the industry enjoyed for many years, this was acceptable. People like their empires and don't like to let go, and without external pressure they seldom do.

However, as growth in the West stalls or even heads into reverse, and margins also decline, the pressure for de-verticalization will become ever more intense. In addition, the entry of players from the computing world who have already been down this path, and at a time when telephony and computing are rapidly converging, will cause this trend to accelerate.

Examples of de-verticalization abound: from the recent decision of Sprint to spin-off its network operations to Ericsson (a forward looking move in the US, though already old hat in India); through to Motorola's decision to junk its traditional handet OS in favor of 3rd party OS's such as Android (again following from earlier such moves at HTC and others); the emergence of MVNOs (the "virtual" should give us a hint) is another example.

This phenomenon is by no means unidirectional nor inevitable. iPhone is actually driving the process of de-verticalization into reverse, since in that ecosystem, the vendor controls much more of the end-to-end experience than a traditional telecom operator would. Similarly, some operators who had outsourced some activities are realizing that they lost critical control and are bringing them closer to home. However, any industrial theorist who knows their Porter will tell you that the net trend towards de-verticalization is inevitable.

The emergence of 3rd party software platforms - be they OSs such as Windows Mobile or Android, or Appstores/run-time environments such as iTunes, BREW or some emerging efforts to create NAAS (Network as a service) by exposing core applications into a standardized format, will become powerful enablers of the de-verticalization process.

• By creating mass-market standards they will encourage hardware manufacturers to focus their innovation spending on the machine side now that they no longer need to each build their OS'. This could be particularly powerful for many of the Asian OEMs, who have traditionally been much better at hardware than at software. So we can see that the Android/WinMo set-up could set the stage for Samsung, LG, HTC, ZTE, Huawei to further gain share from other OEMs.
  
• The emergence of standards has also famously re-energized the developer community. BREW did so for the CDMA world, through as CDMA is becoming marginalized it lacks the platform for growth. iPhone's development platform has clearly set the app development world on fire. Outside of iPhone, one critical issue is whether RIM, Android and WinMo are truly standards - if they are adapted to each operator or OEM's needs, they may create so many variants that they bring us back to the bad old days of multiple Java builds
  
• Software-defined radio, such as Vanu Inc, could also be interesting, by enabling the replacement of dedicated network hardware by a combination of software platforms and standardized hardware.

As mentioned, we have not even begun to articulate how we will play in this space. There are many big boys with limitless pockets playing in these games. However, we think that as this trend advances it will enable the emergence of some very exciting niches that can lead to VC level returns.