A year ago I wrote an article reasoning that Symbian’s outlook is no longer promising. In the article I argued that Symbian’s strategic value is deteriorating, as Nokia (both an ally and competitor) stifles Symbian’s efforts towards developing an independent, self-sustaining and competitive operating system for mobile handsets. My thesis was that Symbian’s only strategic option was to seek political and financial support from a network operator forum, as it is only the operators whose interests are aligned with those of Symbian.
Fast forward a year later and so much has changed in the handset OS chess game: Palm and DoCoMo are using Windows Mobile, open operating systems are heading towards mass-market and the OMTP operator-centric forum looks set to achieve very little.
However, Symbian’s future is looking worryingly same, destined to be reduced to a software house serving the needs of Nokia. Other manufacturers are becoming increasingly disillusioned with Symbian and Nokia’s S60 licensing strategy appears a costly but doomed experiment. Why invest in an independent software house, with the most complex software development and integration process in the industry when Nokia could have developed its core OS in house, faster, better and cheaper ?
I would argue that what Nokia needs to do is revise its strategy to form an ally of Symbian instead of a co-opetitor, to help the ailing OS to become an autonomous, turnkey but flexible operating system. But first, it’s worth reflecting on why Nokia’s strategy towards Symbian has failed so far.
Behind Symbian’s glossy numbers Nokia has been the main force behind Symbian, and its over 100% year-on-year rise in sales for four consecutive years. At the same time it has been depriving Symbian of platform value, by supplying its own middleware components (PIM applications and their engines, security, browser, messaging components, sync engines, Java VM, DRM engine and UI customisation engine). This has broken Symbian platform story, which coupled with Symbian’s idiosynchratic C++ language and lack of a reliable IDE has made for the most complex, arduous development of applications on mobile handsets.
Furthermore, as several essential software components have to be provided by parties outside Symbian and integrated closely with the hardware reference platforms the OS integration and testing process is the most complex and time-consuming in the handset industry.
Essentially, the operating system has not been as easy to customised and integrate on different handsets as other manufacturers had wished. Siemens’, Samsung’s and Panasonic’s delayed (and often failed) attempts at producing Symbian-based handsets is a testament to this. At the same time, DoCoMo’s investment of several tens of millions to co-develop a UI and middleware layer that sits on top of Symbian is an indication of how inadequate the Symbian OS is on its own.
The complexity of OS integration has impacted not only time-to-market, but also commercial viability (how many Symbian OS-based projects have been abandoned?) and scalability (how much longer can Symbian keep doubling its handset projects year on year ?).
One may argue that Symbian has become the least desirable open operating system out there. OK, so there’s Windows Mobile which comes with the negative reputation Microsoft carries, but which has been garnering commercial support by the likes of Samsung, Siemens, Palm, Motorola and the endless array of Chinese ODMs.
But it’s not just Symbian that’s suffering because of Nokia’s strategy. It’s Nokia itself.
Why Nokia shot itself in the foot Manufacturers (perhaps with exception of SonyEricsson) have been disillusioned with Symbian’s value and potential as an operating system for their handsets and have been slowly pulling out in different directions. Even Palm’s Ed Colligan noted that ‘Nokia owns Symbian’, suggesting that Palm would not support the operating owned by a competitor in the smartphone space.
Nokia has indeed developed the organisational structure, expertise and processes to keep up with its smartphone expansion strategy. However, Nokia’s S60 licensing operations haven’t fared well. Here’s why.
One might reason that Nokia’s S60 licensing strategy has been based on two pillars: firstly, extending the S60 platform to handsets beyond Nokia’s own and secondly, influencing the roadmap of competing manufacturers. Therefore, as manufacturers lose faith in Symbian, the S60 strategy suffers. As a result, Nokia’s restrain on Symbian has restricted the potential for S60 beyond Nokia handsets.
In addition, Nokia has had to pay dearly for the Symbian OS through shares and licensee fees (over $100M a year in license fees alone according to ARCchart). In this sense is Nokia’s investment in Symbian sound, when it could have spent less resources and time to develop an OS internally?
No one wants to play with Symbian So where does Symbian look for a helping hand ? SonyEricsson seems to be a persistent supporter, most clearly displayed by its announcements for the next generation Walkman and M600 handsets. However, I expect Nokia to continue keeping SonyEricsson at a safe distance from having a greater say on Symbian strategy. How about Fujitsu and DoCoMo ? DoCoMo simply wants to have a healthy choice of OS suppliers (Symbian, Linux and recently Microsoft), rather than taking an influential stake at Symbian. In parallel, DoCoMo doesn’t want to upset Nokia, as it needs the Finnish giant to supply handsets for its ailing i-mode operations in Europe (where poor handset selection is a key reason for poor i-mode penetration). European operators such as France Telecom/Orange and Vodafone have joined forces in the OMTP, but which is bound to fail amidst the cacophony of self-centred and divergent opinions echoed by its members. And where does an innovative operator like Vodafone go when it decides to invest in a software platform? To Nokia for a S60-based UI customisation layer and next-generation Java APIs in the form of the MSA initiative.
Who’s there to help Symbian? No one.
Nokia has taken some steps to restore the shine on Symbian’s platform story. It has revamped its Codewarrior code development suite (purchased in 2004 from Metrowerks) to provide a unified set of tools for developing across the whole range of Symbian variants, from S60 to UIQ and for entry level to professional programmers.
But this doesn’t address the ailing integration process that harms Symbian’s time-to-market, its scalability, the commercial viability of its projects and has ultimately hurt Nokia’s S60 expansion strategy.
Nokia should u-turn on its strategy Clearly Nokia has to rethink its strategy of depriving Symbian of platform value. For the benefit of its financial investment and S60 expansion strategies, it needs to support Symbian in becoming an autonomous, turnkey, flexible operating system.
Here’s my thesis. In terms of technology, Nokia should allow Symbian to have ownership and control of key middleware components such as platform security, messaging apps, synchronisation engine, Java VM, DRM engine and UI customisation engine. Nokia can retain control of high-value components such as the browser and S60 UI, but ensure that they are clearly abstracted and componentised, so that they can be developed, validated and integrated with as much independence from the core OS as possible. Manufacturers should be able to drop the UI, core OS and customised components, and integrate with the easiness touted by Open Plug’s FlexibleWare architecture. If Open Plug can accomplish this for Linux that has only lately been optimised for mobile handsets, then why can’t Symbian develop a similar componentised architecture for itself ?
Furthermore, in terms of process, Symbian needs to be able to take in direct requirements from operators (not wishlists that may or may not make it to the final handset) and take sole ownership of the integration process for hardware reference design vendors like TI and Intel.
In other words, Nokia should realise that it’s only option is to help Symbian grow its own market and build on it, rather than grabbing as much share of the pie as it can.
And what about Symbian’s shareholder equity ? As Michael Mace, formed Chief Competitive Officer for Palm, put it, “how long will Benq and Panasonic want to remain part owners in an OS they longer use and that damaged their phone businesses? That means 18.9% of Symbian is likely to be for sale in the near future (if it isn’t already)”. Buying this 18.9% would put Nokia above the psychological barrier of the 50% of ownership, but below 70% needed to approve major initiatives, under Symbian governance rules. But again, if Nokia wants to own Symbian, then it’s much cheaper building an in-house OS development team rather than maintaining investment in an independent entity like Symbian. And following the previous thesis, if Symbian is to become an autonomous, turnkey OS then it needs to have a balance of shareholder power across its board.
I only hope Nokia is listening.