"Sometime around the middle of this decade -- no one is sure exactly when -- executives on the go will begin carrying pocket-sized digital communicating devices. And although nobody is exactly sure what features these personal information gizmos will have, what they will cost, what they will look like or what they will be called, hundreds of computer industry officials and investors at the Mobile '92 conference here last week agreed that the devices could become the foundation of the next great fortunes to be made in the personal computer business."
The promise of the mobile Internet has been long in coming. In 1992, then Apple CEO John Sculley was promising this "pocket-sized digital communicating devices" market would be "the mother of all markets", while Intel CEO Andrew Grove called it "a pipe dream driven by greed." Since then, the mobile phone business has exploded, and personal digital assistants (PDAs) like the Palm Pilot (1997) have burst onto the scene. The launch of the RIM Blackberry (1999) brought a real email interface to the PDA world (2002). The World Wide Web itself continues to grow enormously, with Netcraft's December 2008 survey receiving responses from 186,727,854 websites.
We are just now arriving at a convergence in the market that is 16 years in the making. Handset and PDA manufacturers, mobile network operators, chip manufacturers, and computer platform hardware and software vendors all collide with the economics of the Web, collaborative development, and open source software (OSS). Indeed, we are seeing a point in history in which the mobile handset manufacturers and their partners are using OSS and collaborative development to ensure they do not get trapped in the narrow margin price war that caught the personal computer (PC) original equipment manufacturers (OEMs) in the previous technology wave.
A Little History
The historic business of the mobile network operators is based on phone calling time (minutes) and data transfer (bytes). Browsing the web has remained uninteresting for some time, due to low bandwidth and poor interfaces. A few square centimetres makes a poor window onto the Web, especially as web sites have become more content rich and application complex over time.
A significant revenue stream also developed in the mobile communications space around text messaging. The global short message service (SMS) revenue for 2007 was US$100 Billion. This is greater than the combined revenues of the movie box office, DVD sales and rentals, the music industry, and video games. The SMS revenue stream also has an amazing margin, as most of the traffic happens with small limited messages in the control channel reserved for network management. The phone has become a texting device for a whole new generation of users.
In July 2007, the iPhone launched with the wifi-enabled iPod Touch coming shortly on its heels. Within a short few months, these devices dominated traffic on the mobile Internet, claiming 65% of the traffic by January 2008. The communications device world finally had the innovation example it needed for a mobile Internet. Bandwidth was also catching up as 3G networks came on stream. In the past 18 months, all the main device manufacturers have been rushing products to market that present significantly better viewing surface area for media and the Web.
Apple's innovation example went beyond the device. The Apple iTunes Store enables a network ready service from either Wifi or the mobile carrier network. Customers demonstrated they were more than happy to pay for the convenience and immediacy of instant music downloads. This was rapidly extended to the App[lications] Store, also hosted through iTunes. A class of iPhone/iTouch mobile applications appeared on the App Store as Apple encouraged a developer community to form around their mobile platform.
Nokia entered the services market very shortly thereafter with their US$8.1 billion acquisition of digital map maker Navteq in October 2007. Navteq maps are used by almost all major Internet mapping tools, including MSN, Yahoo, and Google. Garmin GPS devices are a primary customer of Navteq. As the world's largest handset maker, Nokia appears to be staking enormous bets that location dependent services will be a booming business in the mobile device world.
Google has long dominated the Internet advertising business. But with approximately 3.3 billion mobile subscriptions in the world versus 900 million personal computers, one can see that Google would want to extend its reach into the mobile Internet space quickly. In November 2007, Google led the formation of the Open Handset Alliance to collaborate on developing a Linux-based royalty-free development platform (operating system and middleware) for mobile handsets.
As Google was bringing Android to market, Nokia announced in June 2008 its intent to acquire Symbian Ltd., the makers of the Symbian handset operating system, for US$410 million and its intentions to form the Symbian Foundation in early 2009, making the operating system available royalty-free and as open source through the foundation through 2010.
Christensen Economics and Innovation
The trend in mobile operating systems to open source licensing and collaborative development can best be described in terms of Clayton Christensen's observations about economics and innovation. Christensen's model provides a great lens through which to view the disparate mobile efforts from the iPhone to the various Linux-centric efforts, and why they differ so widely. Three ideas captured in the model are relevant to our discussion:
- In an early market, the product that "wins" is tightly integrated because the manufacturer has complete control of all aspects of the device and can deliver best on the customer experience. Over time, as the technology space matures, new players can enter the market at the bottom with "good enough" products that satisfy the low end of the market.
- As a technology market matures, standards emerge, enabling more players to enter a market where each attempts to differentiate their offerings.
- Products live in a network of connected technologies, services, and experience. As standards cause value to fall in one of the "nodes" in the network, the value moves to adjacent complementary technology spaces in the network. For example, as inexpensive standards-conforming PCs eroded the value of minicomputers, the profits moved into operating system revenues.
The iPhone and the Blackberry
We will now discuss how open source impacts on the various players within the mobile market. While the iPhone runs an operating system that is related to the Apple-led Darwin OSS project, Darwin has never been a major project and isn't a primary engagement mechanism for Apple with its developer community. The iPhone, true to Apple's history of profitability over market share and its cult of design, is first and foremost an amazing consumer experience. The complete Apple experience depends upon controlling the entire technology stack (hardware, software, application development layer, and service provision) in a tightly integrated fashion.
In the Christensen economic model, the iPhone is a new class of product. It will deliver more value to its target customer through tight control and integration than can be delivered through standardized interfaces and components, and Apple will reap the margin benefits. That focus on integrated design means that unlike Symbian, Apple will never own 65% of the global market. Apple won't be interested in producing "cheap" iPhones for developing economies as Apple is about "life style" computing. The iPhone has demonstrated to the mobile industry what the mobile web experience can be from both a device and a services perspective.
What works so well from a tightly integrated design perspective turns out less well when we look at other single-source devices like RIM's Blackberry. Initially, RIM had their own new class of device. What started as two-way paging in 1999, rapidly became the niche push email service that was key to business users. The Blackberry Enterprise Server connects to the corporate email system, pushing email in real-time down to Blackberry devices. By tightly controlling the entire stack from the hardware and operating system through to the email service provision, RIM was able to present the best business consumer email experience, as well as the associated profit margins.
However, technology has converged, the iPhone has licensed the Microsoft Exchange ActiveSync technology for the push email used with Microsoft Exchange servers, and RIM is left trying to evolve the Blackberry against a device with two years of advanced features and services well beyond its own push email niche. While Apple continues to expand the connections within its technology network to best effect, RIM is left somewhat in a one-trick pony position.
Linux Variations on a Theme
Symbian and Microsoft were able to deliver powerful smartphone operating systems that ran on a wide variety of handsets across all the mobile networks within their regulatory environments. They were rewarded through their per handset royalty. Handset companies, however, are exploring Linux as a mobile operating system platform.
Symbian and Microsoft looked to China, India, and other developing economies and assumed a proportionate claim of market share as those markets evolved. The problem from the handset manufacturer perspective changes when you start considering billions of dollars in royalty payments from those potential markets. At that level of cost, Linux experimentation and investment makes good business sense.
Motorola delivered the Ming phone into China using a Linux base. This was almost a year before the iPhone was announced and, for the Chinese market, was arguably a more useful phone. The Ming supported full stylus input for Chinese characters, an enormous consideration for texting. The Ming also had the first 2 megapixel camera, a media player, and came in a sleek package.
At the same time, Nokia was delivering the N770 and N800 Series Internet tablets, devices not much bigger than the first iPhones, with stylus input and a viewing area similar to the iPhone. These were Linux-based and Nokia was working directly in the open source community on several fronts to maintain concurrency with the Debian Linux distribution rather than living on a fork. From these efforts, Nokia developed the Maemo open source developer community.
This was a significant step. As the handset industry changed and the operating system royalty became a problem, each handset manufacturer wrestled with Linux. Each wanted a royalty free operating system, but trying to integrate into the Linux community has been a source of frustration. Things that are critically important to handset manufacturers, such as battery life, aren't necessarily interesting to the mainstream Linux community. Each handset manufacturer was forced to fork with the attendant costs. Nokia, working within an existing community, was a huge step forward.
Out of the desire to avoid forking Linux came LiMo. LiMo is a not-for-profit organization formed by the key mobile players in January 2007 to collaborate on a Linux kernel for mobile. This level of collaboration is not foreign as the mobile handset manufacturers shared technology through Symbian Inc. for ten years. Symbian Ltd. was a for-profit corporation, but its primary shareholders and board members were the handset manufacturers. The handset manufacturers do not want to fall into the trap that the PC manufacturers fell into with Microsoft, where the manufacturers took the biggest hit in the narrow margin price war, while Microsoft made money on the operating system. So, for the handset manufacturers to work collectively to manage mobile operating system costs while continuing to deliver their own differentiation on the handset makes good economic sense.
Before Google announced Android and before Nokia announced that Symbian would become open source, LiMo was likely the best opportunity for a shared royalty-free Linux mobile platform. LiMo's downfall will probably be Nokia's proven ability to work directly with the open source community. With the formation of the Symbian Foundation and the Open Handset Alliance, it remains to be seen if LiMo will deliver the Linux distribution for mobile, or whether the handset manufacturers will choose to instead focus investment in the other two efforts.
There are other Linux-based efforts underway in the mobile Internet space. Intel created Moblin to work on a similar mobile focused Linux distribution for the growing "netbook" and in-vehicle devices. Canonical Ltd., the primary developer of the Ubuntu Linux distribution, is working on the Ubuntu Mobile Internet Device Edition (MID) which is focused on the growing netbook market. Each of these are carefully not targeting the mobile phone space. It will be interesting to see how this space evolves as the mobile phone and laptop/netbook technology spaces converge.
Google led the formation of the Open Handset Alliance to collaborate on developing a Linux-based royalty-free development platform (operating system and middleware) for mobile handsets. This culminated in the release of the Android mobile software platform as OSS in October 2008, and was shortly followed by the Android Market as a distribution mechanism for Android-based applications. Like Apple, Google is forming a developer community around Android and wants to set up the Android Market as the marketplace for developers and consumers to meet.
Google wants to drive Android application development that uses Google services in order to grow their ad revenue as the mobile web comes into its own. They are providing the necessary focus around which the handset manufacturers, mobile network operators, and related OEMs can collaborate. Based on their own open source experiences, they are also providing the necessary architecture for participation required for a successful open source community to develop.
Android is likely to face two key challenges. First, Google is not an operating system company. They don't necessarily have the experience and resources to meet the calendar demands the handset manufacturers have for regular releases, something that Symbian has traditionally done well. Second, while it makes sense from Google's perspective to drive mobile application development around their services, the rest of the world may not feel that way and have their own, possibly competing, perspectives.
This should be an opportunity for the newly forming Symbian Foundation. Google application services running on a Symbian base would seem to be a win for Google, application developers trying to settle on a model, and for Symbian doing what it does best around the mobile operating system. Since the not-for-profit Symbian Foundation will not be under market competitive revenue pressure, profit-centric competitive decisions are removed that might have historically put cooperating with Google at risk. This would seem to be a good time for Symbian and Google to explore their complementary spaces in the technology network.
The Symbian Foundation
In December 2008, regulatory approval was given and Nokia completed the acquisition of Symbian Ltd.. Nokia will launch the Symbian Foundation early in 2009. Fujitsu, Motorola, Nokia, NTT DOCOMO, and Sony Ericsson will contribute related software and documentation assets to the newly formed foundation. All the assets will be available to foundation members under a royalty-free license.
A new platform will be developed from SymbianOS and other donated components. The platform will offer the means to build a complete mobile device while providing the tools to differentiate devices through tailoring of the user experience, applications and services. Platform assets will be made available as open source gradually over the next 2 years. The intent is to use the Eclipse Public License (EPL), making the platform code available to all for free.
The Symbian Foundation's use of the EPL will likely make handset manufacturers more comfortable than the GPLv2 under which all Linux work must be licensed. The IBM-lineage of the license ensures that a hardware patent clause is still intact to protect unambiguously all hardware patents.
There will certainly be challenges. Nokia will need to manage the challenges that come with any acquisition. Organization and governance of the new foundation will be key, but the Eclipse and Mozilla Foundations provide excellent blueprints for the Symbian Foundation. Quickly putting into place an architecture of participation for everyone to understand how to contribute and participate will be essential for the Symbian Foundation to succeed.
Microsoft faces interesting challenges for the Windows Mobile platform. First, there's the royalty problem. With collaborative development, open source intellectual property (IP) management, and a lack of royalties defining the investments of the industry, it is difficult for Microsoft to justify the value of their royalty-laden offerings. Windows Mobile only had a quarter of the deployment of SymbianOS in a pre-acquisition, pre-foundation world. Now SymbianOS will be royalty free, as is Android, anything delivered from LiMo, and the Intel and Ubuntu work. Offering a royalty free operating system, however, only makes sense for Microsoft if it drives significant revenue growth and profitability in related products and services in the adjacent nodes of their solution network.
Second, there's the open source culture versus IP protection problem, both internally and from an external partner perspective. It is unlikely that Microsoft will risk participating in any of the existing open source foundation or alliance efforts.
There is also a more subtle business problem. The handset manufacturers do not want to be caught in the same trap as the PC OEMs. Mobile industry partners have banded together through various foundations and alliances to force the economic devaluation of the operating system. This will further isolate Microsoft from the industry.
The mobile Internet has been a long time coming. Even as mobile phones and PDAs became prominent and the World Wide Web itself exploded, poor bandwidth and awkward viewing interfaces hampered the mobile Internet from taking off. We are finally at a point of industry convergence where these problems are being solved rapidly.
Christensen's economic model for innovation allows us to see why the different device manufacturers behave as they do. OSS and collaborative development are redefining how the mobile Internet will be built. The various handset manufacturers and their mobile industry partners are sharing the costs of mobile operating system development through multiple different alliances. This allows them to drive their own differentiation on the handsets themselves to maximize margins.
At the same time, the Apple iPhone continues to provide a great innovation example for all to follow. It remains to be seen how RIM and Microsoft will respond, if the situation allows them to respond at all.