Operating system diversity: no standard emerges on the smartphone or tablet
Smartphones and the new generation of tablets have three things in common: (1) they use similar, low-powered 1GHz processors; (2) they are being used like personal computers, although they are not personal computers; and (3) Deloitte predicts that by the end of 2011 no operating system on these devices will have a dominant market share. Some will have more than a 5 percent share, but no single player will have yet become the de facto standard, as has been seen in other computing ecosystems in the past.
The emergence of a defacto standard is very important to everyone involved in the smartphone and tablet markets (i.e., “non-PC computing devices”). Technology industries with a single dominant provider of hardware or software tend to have very different economics than those with multiple providers.
Moreover, the non-PC computing market is becoming increasingly important: Deloitte predicts that in 2011 more than half of all computers sold – over 400 million units – will be devices that are not PCs1 . In 2012 and beyond, non-PCs’ will likely extend this lead. Being the dominant operating system (OS) provider for non-PCs would be a tremendous prize; however, a clearly dominant OS seems unlikely to emerge in 2011.
The obvious value of being the standard OS for non-PCs is one reason why a single player is unlikely to dominate. The company that popularized the PC was very adept at making both operating systems and central processing unit (CPU) chips. But when it decided to launch its new computing device, it did not source its OS or CPU internally; neither did it select other large OS or CPU manufacturers. Instead it turned to companies that were relatively small at the time2. No one expected that PCs would sell in their billions generating trillions of dollars over the next three decades3. In contrast, the potential value of the non-PC market is already obvious.
In the history of computing devices, no company has ever become the standard OS after it was clear that the value of the technology involved was likely to be tens of billions or more. The smartphone and tablet markets are already generating hundreds of billions in revenues, and are expected to keep growing rapidly: everyone knows there is “gold in them thar hills”.
Equally important, existing competition in the non-PC market seems unlikely to fade any time soon. In both the smartphone and tablet markets, the top five OS companies based on market share each enjoy annual revenues in the tens of billions; have billions in cash; are growing their top and bottom lines; and are gaining ground with consumers. Even companies that are losing market share are still growing in absolute terms. None appears ready or willing to abandon the market, there is no shareholder pressure to exit, and all should have more than enough resources to develop new products and market them to mass audiences.
Mobile network providers seem as unlikely to allow a dominant OS as the OS competitors themselves. Providers appear to actively encourage OS diversity, even if they seldom state this explicitly. Whenever any operating system or device maker, in any geography or segment, approaches 50 percent market share, the carriers seem to deploy their marketing muscle, retail presence and subsidies to push one device or OS over another. There is even a name for this strategy: carriers pick “hero” devices and push them hard. This approach can cause major swings in market share within a single quarter4.
Device manufacturers also seem to believe that a diverse OS ecosystem is in their best interest. Perhaps seeing the commoditization of some technology industries where de facto standards do exist has shown them that having different operating systems to choose from allows them to differentiate their products more effectively, strike better deals with network providers and generally achieve higher margins.
With all of these major players fighting to prevent an industry-wide standard OS from emerging, it seems that only a nearly irresistible force could make it happen.
One possibility for such a force is the “network effect.5 ” As an emerging standard becomes more and more prevalent, it can rapidly reach a tipping point where the number of users on its platform relative to others gives it overwhelming momentum. Software developers embrace the platform because it has the most users, which leads to new apps that attract even more users, and so on – creating a virtuous circle that eventually locks the platform in as the de facto standard.
However, unlike other markets where network effects have created industry-wide standards, both smartphone and tablet devices are highly fragmented at the hardware level, and are likely to remain so. The variety of screen sizes, CPUs, text inputs, form factors and other configurations means that no single OS version addresses a sufficiently large percentage of devices to create a powerful network effect.
A market that revolves around a standard OS offers the advantage of simplicity: if an application works on one device, it can be counted on to work almost everywhere and with nearly all devices. The current non-PC market is more complex and fragmented, but looks like it may be the new reality. With fewer than half of all new computing devices being PCs, and if no non-PC OS becomes dominant, it is mathematically impossible for any OS to have more than a 75 percent share of all new computing devices.
Many stakeholders would be affected if a standard tablet or smartphone OS does not emerge in the near term. As noted earlier, mobile network providers, device makers and software companies are all intimately involved and all have a keen interest in the battle. However, there are three other groups with a significant stake as well.
Applications developers might need to get comfortable in a world where they must pick and choose their platforms. If a clear standard emerges, then their choice is made for them. But in the more fragmented OS world that seems likely, no app developed for a single platform can address the entire market. Developing customized apps or versions for each OS requires significant time and money (typically $5,000-500,000, depending on complexity6 ), so it is likely that many smaller developers will not be able to address all markets.
Media companies face a similar challenge. In general, ad-supported media is looking for the largest possible audience. In the PC market, the platform with the largest audience was clear. However, in a diverse, non-PC computing market, publishers will probably need to prioritize some audiences over others, or even exclude some entirely.
Finally, IT departments are likely to face significantly higher costs to support this new, more diverse technology environment. Administering a traditional PC computing ecosystem centered around a single OS can cost thousands of dollars per employee per year 7. Supporting five times as many operating systems is unlikely to require fewer people and less money. Yet telling employees to go back to the days when they had to standardize on one OS or device seems impossible: that particular genie left the bottle long ago. Given these trends, the cost of IT support seems set to rise.
1For more discussion on this theme, please see the prediction “End of an era: more than half of all computers aren’t computers anymore”
2How the PC Really Got Started, Fast Company, 19 December 2007:http://www.fastcompany.com/articles/archive/pc_bday.html
3"There is no reason anyone would want a computer in their home." -- Ken Olson, president, chairman and founder of Digital Equipment Corp., 1977 to the World Future Society
4Smartphones shifting power from carriers to OEMs: Yankee Group report dated 8 June 2009http://www.yankeegroup.com/ResearchDocument.do?id=51584
5 Network effect Definition from PC Magazine Encyclopedia: http://www.pcmag.com/encyclopedia_term/0,2542,t=network+effect&i=47879,00.asp
6How much does it cost to develop a mobile app on iPhone, Android and other platforms?, Mobile Marketing Universe website, 7 September 2010: http://goldengekko.blogspot.com/2010/09/what-are-costs-of-developing-mobile-app.html
7Gartner Says Effective Management Can Cut Total Cost of Ownership for Desktop PCs by 42 Per cent, Gartner Press release, 10 March, 2008: http://www.gartner.com/it/page.jsp?id=636308