I've always thought it a bit weird that Microsoft (MSFT) has gone into the business of selling devices. At first, this seems like a wonderful idea - why deal with those pesky OEMs getting a cut when you can have the entire pie to yourself? This is the business model that Apple (AAPL) has very successfully capitalized on, but it is not a business model that is appropriate for all companies. It is my view that Microsoft's dominant position as an ecosystem vendor whose top and bottom lines are driven by the multitude of partners that it enables makes it inappropriately positioned to play the Apple game. Microsoft selling Surfaces is an idea as preposterous as Apple trying to license Mac OS X and iOS to other device vendors; it needs to focus on enabling - not competing with - its partners so that Microsoft Windows continues to be a ubiquitous, household name that consumers and enterprise users alike can't live without.
Surface Is A Bad Idea
The PC business is not a particularly high margin one. Apple has done well in this space primarily because it caters to the high end markets. However, while Apple's Mac business is profitable, the real bread-and-butter for the firm is its much higher volume iPhone and iPad devices, as well as iTunes and the app store. Apple's problem is that it now faces the real margin squeeze from the hordes of tablet and smartphone vendors, armed with Google's (GOOG) Android platform, producing similar products and selling them for much lower cost. The vendors doing this to Apple are used to the low margins in the PC space, so any incremental business is a win for them, but devastating to Apple.
Microsoft's Surface RT and Pro devices are certainly solid pieces of hardware. However, as I watched Intel's (INTC) showcase at Computex Taipei, I noticed something interesting; does this look familiar to you?
It's a Surface clone! Now, instead of Windows RT, this is a fanless, high battery life Windows 8 machine with full app compatibility and good processing power. What did it take to make such a device? Not much, it seems. Buy the display, buy the RAM, buy the SoC and the board, slap it together in a chasis, and call it a day. There is very little barrier to entry into this market, and the high margin secret sauce for Microsoft is the license of Windows. Why take on the Herculean task of trying to market a tablet amidst a sea of competitors all armed with the same components, when you can sit back and get paid for each Windows license sold?
Further, it seems that Microsoft's attempt to "compete" with its partners is ultimately self-destructive as this will just drive the tablet vendors to use the Android operating system. While I suspect that Intel has a vested interest in seeing Windows 8 succeed in this space and could probably persuade Microsoft's partners to keep the faith, this is very unhealthy for Microsoft's business model in general.
Taking On Unnecessary Risk
The wonderful thing about software such as Windows is that it costs literally nothing to distribute. There is no risk of "building up inventory" of copies of Windows since there is nothing physical there, but a bunch of unsold Surface RTs end up getting sold at or below cost to students in a last ditch attempt to clear unwanted inventory. By licensing Windows, Microsoft's only job is to make sure there's enough demand so that each time an OEM puts a copy of Windows on its devices, Microsoft essentially prints money. Given Microsoft's dominant position in PCs and notebooks, and the real potential to be a dominant force in tablets, I do not see why Microsoft would want to upset this beautiful business model by trying to compete in selling commodity hardware and competing with its partners.
It seems that in this new world of computing, everybody is trying to vertically integrate, not remembering that there was a reason that horizontal ecosystems dominated for many years (and will again).
I suppose that Microsoft's Surface needed to really kick-start the PC and tablet designs going forward with its Surface products, but I think the rest of the industry gets the message. The winners are the ones with the best designs that sell through to fickle customers, and the losers are the ones with the mundane and tough to use products. Microsoft's job should be to make sure the Windows OS is as user friendly and as great as it can be, Intel's job should be to make sure Windows products have the best silicon, and the OEMs' jobs are to figure out what devices based on these solid building blocks that people like and to build them to scale.
Microsoft's Surface doesn't work, and the sooner Microsoft cuts the cord on both RT and Pro devices, the better it is for Microsoft and its partners.