Microsoft (MSFT) has a surprise hit on its hands: the Microsoft Surface Pro -- A 10.6", 2lb laptop/tablet hybrid retailing from $899. Unusually for the relatively staid laptop marketplace, it quickly sold out upon launch on February 9th, with new orders now forecast to arrive from March 1st onwards. This is an impressive reversal of the Surface RT launch in November. Despite being available on launch day to most developed markets, and despite being around half the price of the Surface Pro ($499 upwards), the Surface RT struggled to make an impact, with estimates suggesting less than a million sold so far. To put this in context, Asus, a relative lightweight in the tablet world, sold three times as many in Q4 2012.
The contrasting successes of the Surface line present an unexpected problem to Microsoft, which will necessitate a strategic shift by the company. If Microsoft fails to recognize what has occurred within the next few months, it will lose the tablet sales war for another software generation. To help the reader gauge for themselves the true scope of this problem, a quick recap of how this came about is in order.
Microsoft's Strategic Choice
In November 2012, Microsoft launched two operating systems: "Windows 8" and "Windows RT". The former is designed to take advantage of the x86 instruction set architecture. This market, traditionally dominated by Intel (INTC), has supplied most of the world's PCs for the last 30 years. Historically, the x86 family's chief strength could be characterized as delivering the raw power required for complex applications. Almost unnoticed at first, however, the ARM (ARMH) instruction set architecture has grown in prominence. Chances are, if you're reading this from a tablet or a smartphone, you're using it right now. Characterized by energy efficiency, this simple instruction set has become the platform of choice for vendors such as Apple (AAPL), Broadcom (BRCM), Nvidia (NVDA) and Qualcomm (QCOM), creating products where space is at a premium, such as the archetypal tablet.
As tablet sales exploded, Microsoft had been left out in the cold. With Windows 7 unable to operate on ARM and not being designed for touch, over 97% of tablets sold in 2012 relied on non-Microsoft solutions. To establish a presence in this market, Microsoft created Windows RT, specifically designed to work in energy efficient situations. This supposedly stripped-down version of Windows 8 nevertheless requires a massive 8 GB, taking up over twice the space of iOS or Android. To show the capabilities of these two sister operating systems, Microsoft created the Surface Pro -- running Windows 8 and the light-weight Surface RT -- running Windows RT.
What Microsoft Envisaged Happening
Microsoft envisaged Windows 8 and Windows RT complementing one another through classic market segmentation. The full-blown W8 OS would run on advanced PCs and laptops. RT would run on tablets and small hybrid machines. This would enable Microsoft to establish a presence on tablets and create a viable 3rd ecosystem alongside Apple's iOS and Google's (GOOG) Android. Then, through having a shared marketplace (known as the Store) for W8 and RT, they would achieve economies of scale and make it a profitable venture for 3rd party app developers. This marketplace would then slowly make the existing "desktop" applications obsolete, with most new apps bought and downloaded from the "Store." Most pertinently, this Store is curated by Microsoft, which for its trouble, takes a rather generous 20-30% cut (see 5b) of any sale made.
The good news for Microsoft is that W8 sales are strongly on track, with 60 million licenses sold to date. Just like its predecessors, one can predict that this will become the world's biggest operating system (though it may be under the different moniker of Windows Blue).
RT sales, on the other hand, have singularly failed to take off. To give an insight into the popularity of the Surface RT, the device supposed to showcase RT's capabilities, the following Google Trends graph is illuminating:
I believe this graph is a direct inverse of what Microsoft had predicted. This mistake is probably based on the very simple supply/demand principle that the lower the price, the higher the demand. One can see how the Surface RT may sell twice as much as the Surface Pro, being half the price, for example. However, this is not happening. Microsoft's poor demand forecasting system made around 1.25 million Surface RTs available at launch, but extrapolating from the small number of Surface Pros distributed to retail outlets throughout the U.S., I estimate only 50,000 units for the Surface Pros. The result was predictable. An overabundance of Surface RTs with a launch possible in many countries simultaneously (expect many of these to be discounted for sale shortly); and not enough Surface Pros for sale, despite being available only in North America. I do believe, however, this number will greatly increase over time as Microsoft's supply chain shifts production to the Surface Pro in preference to the Surface RT to tap the unmet demand.
What Microsoft Needs To Do Next
I believe that Windows RT is an engineering solution to a problem that didn't need solving. Microsoft strategists, in a classic piece of MBA-thinking, evaluated Intel and its low-power capabilities and compared them with the ARM chipset makers and their capabilities. One can very clearly imagine a boardroom wracked with the political intrigue Microsoft is famous for, where strategists far from Microsoft's centers of innovation agreed that this market position was static. They did not foresee that Intel is making giant strides in low-power chips, as it has fully recognized that ARM presents an existential threat and needs to be responded to. Despite hiccups along the way, one can fully expect Intel, which spends $2.5 billion on R&D per quarter, to eventually catch up with ARMH, which spends 50 times less on R&D. Effectively, this means, even with the most pessimistic forecasts, within two years, Intel will be able to fully compete in the tablet marketplace, negating the entire raison d'être for Windows RT.
Just as importantly, Microsoft's strategists did not foresee that a Windows RT tablet, with the core capabilities of Windows removed (i.e., the capability to run desktop programs), is not a very attractive proposition at all. This unattractiveness is increasingly apparent with Samsung (SSNLF.PK) giving up on a global rollout of the Samsung Ativ RT, and Nokia apparently cancelling the launch of its Windows RT tablet entirely. I did voice my doubts in an earlier article about Nokia's (NOK) RT tablet, and I do believe it has made the right choice if it has switched to developing a tablet running the full Windows 8. With other industry leviathans voicing doubts about RT, I believe ultimately this platform must be regarded as stillborn.
Microsoft's next steps are thus crucial for investors and interested observers in the stock. Windows RT was an interesting experiment, which can be considered a failure. If Microsoft persists in trying to bludgeon its way into an incredibly competitive tablet market with Windows RT, it will not succeed. It will lose money, and it will take away valuable developer time from ensuring the success of Windows 8, where Microsoft actually has demonstrated a competitive advantage through the Surface Pro. It needs to continue with spreading the usage of Windows 8 on tablets through optimizations. If Microsoft instead persists with RT development, this will call into question whether Microsoft will ever be able to out-innovate Google and Apple and stop its decade-long oscillation around the $27 stock price mark. Google especially has the brilliant ability to drop products that are not working in its annual "Spring Cleaning". This is perhaps a key reason why, as of today, Google's market capitalization has overtaken Microsoft's, and is likely to stay that way for the medium term. The ideal reaction for Microsoft then would be to quietly shift emphasis away from Windows RT. If this does occur, perhaps we really are seeing a new more determined Microsoft, which seizes opportunities when it sees them and divests itself of failures quickly. If it doesn't, expect continued share-price stagnation.
Disclosure: I am long NOK.