This essay was written by David Wood and distributed inside Symbian in August 2001.

Towards the 100 millionth Symbian OS phone

It’s hard to forecast the future with any degree of confidence.  Investments can go down as well as up.  Past performance is not a guarantee of future performance.  The beating wings of a single butterfly can give rise to raging tornadoes.  You know the kind of thing.

But despite these uncertainties, I have a prediction to share.  Some time around the year 2006, the 100 millionth phone will be sold that contains Symbian OS.

With a good prevailing wind, that day might arrive sooner – perhaps even as soon as 2005.  With a less co-operative set of market conditions, that day might be delayed, perhaps into 2007.  As I said, it’s hard to predict the future.  So don’t just take my word for it.  Look over the evidence for yourself, test fly a few thought experiments, weigh up the imponderables, and reach your own decision.

Who needs an OS anyway?

In the year 2000, more than 400 million mobile phones were sold.  I’m not giving away any state secrets if I tell you that considerably less than 1% of these contained Symbian OS.  On the face of things, that portends badly for any idea that there will be 100 million Symbian OS phones within just six (or five, or seven) years.

But bear with me.

Symbian OS wasn’t the only “big name” phone OS to fare badly in 2000.  Microsoft CE, Palm OS, Linux, you name it – all these operating systems equally failed to come near breaking into single figure percentage points for the mobile phone market.  As a rule of thumb, if a member of the general public could give a name to an OS, that OS won’t feature in anything more than a footnote to the sales figures for mobile phones in 2000.  Instead of “big name” operating systems, the mobile phone market has been dominated by nameless operating systems.  After all, how many people can give the name of the operating system that is at the heart of the Motorola V50, the Siemens C35, the Ericsson T20, the Panasonic GD35, or the Nokia 3310?

Of course, these operating systems do have names, and each of them has had countless developer-years of hard work poured into them.  That huge effort has paid off, through large market shares, albeit not by becoming a household name.  And so the mobile phone industry is currently dominated by operating systems that the general public neither knows about, reads about, nor cares about.  You probably don’t know what operating system is running your car, or your TV, or your music centre – and likewise you don’t know what operating system is running your mobile phone.  It’s not something you’ve had cause to worry about.

But that is about to change.  The change will be due to the problem of ever-increasing new functionality, subject to unique constraints of compactness and efficiency.  The change will also be due to the fact that Symbian OS provides an excellent solution to that problem (after all, not every problem gives rise to an OS that sells upwards of 100 million units).

The problem of new functionality

What is the reason why more than 400 million mobile phones were sold in 2000?  That averages to about one mobile phone purchased for each ten people in the world above the age of thirteen.  Given the expenses of owning and using a mobile phone, these devices have got to be fulfilling an important set of needs.

And they are.  In short, these devices

• allow users to communicate,
• inspire a sense of safety,
• make it easier for users to fill in “slack time” (through easy communications to friends and, increasingly, games to play),
• and even add to fashion,

all without being too expensive, and without being so heavy or bulky as to deter portability.

To summarise the summary, mobile phones have been wildly successful, because they successfully address the seven-fold user needs of (1) communications, (2) safety, (3) entertainment, (4) fashion, (5) affordability, (6) mobility, and (7) reliability.

Mobile phones didn’t start off this way.  They reached their present state of mass-market usefulness only after many generations of hard-won incremental improvement.  Each generation added some extra functionality, enhanced efficiency, increased portability, boosted fashion sense, improved reliability, refined usability, and/or reduced cost.  The end result is one of the wonders of the modern world.  (And I’ve only told a fraction of the story.)

But that’s the past.  What’s next?

The simplest prediction is to expect yet further efficiency and portability (increased battery power and more convenient pocketability).  But I predict there’s going to be lots more than that.  The biggest changes will come from new functionality.

And that new functionality is going to break the existing operating systems.

To be more exact: That new functionality is going to see the existing operating systems dramatically surpassed by a new operating system – Symbian OS – which has significantly greater ability to supply this new range of functionality.

Without Symbian OS, most of this new functionality would remain just a dream – something to be discussed by journalists, academics, and over-keen marketeers. With Symbian OS, this new functionality is going to be discussed by upwards of 100 million consumers.

Three kinds of new functionality

There are three overlapping waves of new functionality in forthcoming mobile phones.  Each wave will be more compelling than the last.

The first wave is that of auxiliary functionality.  The second wave is integrated functionality.  The third wave is community functionality.

Auxiliary functionality is when some extra functionality happens to be bundled with the phone, such as a calculator, a notepad, a game, a music player, a watch, or an alarm clock.  This new functionality, when it works well, allows consumers to travel light, leaving other items at home (and in the shop).  Why carry seven devices, when one will do just as well?

This kind of extra functionality is “nice to have” – ranging from “quite nice to have” to “very nice to have”.  But it’s not as compelling as integrated functionality.  Integrated functionality doesn’t just give you “a phone plus something else”.  It gives you “a better phone”.

Community functionality is even more compelling.  It will propel the wide take-up of new generations of phones, with all the energy of a thriving network effect.  But more of that later.  First, let’s take some time to appreciate the attractiveness of new integrated mobile phone functionality.

Phones with ever greater integrated functionality

Integrated functionality enhances features that are core in existing phones.  For example,

• Phones allow you to store phone numbers in them, and to assign names to these numbers for easier recognition. This is the start of a contacts database.  Better phones have better contacts databases.  Key features include synchronisation to contacts databases held elsewhere (such as on PCs), multiple phone numbers per contact, multiple ways of searching the database, and ease of adding new entries or editing existing ones.  The end result is that it becomes easier and easier for you to use your phone to keep in touch with all the people who are important to you.
• Let’s not forget that phones are, at heart, about communications. Better phones support richer kinds of communications.  Direct voice communication was the start.  Voice mail was one development.  Text messaging was another.  Beyond text messaging comes concatenated text messaging, picture messaging, and “multi-media messaging”.  In turn, “normal” voice communications will be enriched by simultaneous data communications – sometimes called “voice plus”.
• Communication doesn’t need to be restricted to two parties at a time. Better phones make it easy to have “conference calls”, with multiple parties joining or leaving the discussion at will – just like a real conversation!
• A particularly powerful form of communication is “information publishing”. The World Wide Web is a striking example of this, in the wired world of connected PCs.  In the wireless world of mobile phones, WAP was initially launched as “the Wireless World Wide Web”.  That initial positioning was a mistake.  But much of the original vision remains true: WAP can bring all kinds of information to suitably equipped mobile phones.  The better the phone, the easier it is for users to selectively subscribe to WAP services they like.  So rather than ringing a series of people to ask them a question, you can “ring” a suitable information bulletin board, and can find the answer more quickly yourself.
• We communicate for all kinds of reason. A lot of these reasons are to conduct commerce – to buy or sell.  Future mobile phones will integrate many of the functions of your wallet and your credit cards.  Travel will become even lighter!
• For most of us, communication with our business co-workers is a vital part of our overall communications. Our office email systems contain large amounts of fast-moving conversations, meetings minutes, action lists, escalated problems, creative solutions, heartfelt pleas, tactful reminders, and tasty office gossip.  Love it or loathe it, office email is here to stay.  Keeping in touch with important email communications, while mobile, is just as important as keeping in touch via mobile voice communications.  Better phones make it easy to access office email selectively and painlessly.  They even help us to love our office email, rather than to loathe it.
• Phones are about helping people to be productive, but also about helping people to be frivolous. They help people to use scarce time effectively, but also help people to use slack time entertainingly.  So, mobile games are a serious business.  Better phones will provide multi-player games where the other players are literally a phone call away, where the game can be stopped and re-started at will, and where communications play a key role in the fun.
• Mention productivity and effectiveness, and you’re bound to think of scheduling and calendars. Better phones will make it easy for users to consult and update their calendars while speaking over the phone.  Calendar information will be readily available on the device, along with contacts information, finance information, and location information.
• Finally, phones are about individuality, personalisation, and enchantment. Better phones don’t just come with lots of functionality crow-barred into them.  They will come with adaptable personalities and true “ease of use”.  They will thrive because users will almost think of them as indispensable close friends.  Although they are jam-packed full of complexity, they will ooze simplicity.

Community functionality

If there is already strong pressure on phone users to buy new models with new features, to keep up with the latest new “must have” items, one effect will shortly increase this pressure by a factor of around ten.  This is the “community effect”, whereby the new features on phones will create new communities of users, who can swap new kinds of messages and other information with each other – so long as everyone in the community has phones that support these features.

If a consumer has a phone without these features, they won’t just be feeling envious of their friends who have phones with these new features.  Worse than that, they will be feeling excluded from the latest buzz and chatter – excluded, because their phones are not powerful enough to cope with the latest kind of picture messaging, rich text messaging, forum messaging, mobile chat rooms, or whatever.

This community effect has already been seen with the first generation of text messaging.  You can only send a text message to someone with a phone that can receive such a message.  But as soon as group interchange of messages becomes easier, this effect will become stronger and stronger.  People without the latest messaging capabilities on their phones will risk missing out on fast-moving changes to plans for which pub to meet at, which disco to drop into, or whose place to end up at.

The community effect is going to be particularly strong when Bluetooth gains in popularity.  This will be a whole new way for phones to talk to each other (and to lots of other devices besides). You can already see people occasionally beaming contacts information to each other, using Infrared.  Bluetooth is a much more sophisticated beast, and will encourage huge amounts of new kinds of communications.  Communications, that is, among users who have the latest sophisticated Bluetooth features in their phones.

All this leads to the 1 billion-dollar question.  What kind of software is going to be able to power increasing numbers of mid-range phones with the latest high-end features?

Solutions to the problem of new functionality

There are three kinds of possible answer to the 1 billion-dollar question:

• The nameless operating systems, which have brought the mobile phone industry to its present heights, through multiple generations of incremental improvements, will go on incrementally improving, and will enable phones to keep on getting better and better.
• Alas, the nameless operating systems are already at their limits, or very near to them. So, any dreams of significant new functionality are just that, namely dreams.  The above descriptions of better mobile phones are science fiction.
• Yes, the nameless operating systems are approaching their limits, but there is a new kind of operating system available, that is ready to step into the breach. This will be an example of a disruptive technological innovation.  Disruptive technologies tend to take a long time in preparation, but then suddenly zoom into a new position of market prevalence, once a suitable “tipping point” has been reached.

Spotting successful disruptive technologies in advance is one of the hardest tasks that falls to business leaders.  Candidate disruptive technologies appear from all sides, all the time.  Such-and-such a technology might have the capability to radically transform an industry.  The world is full of bright ideas and wacky business plans.  And when you look at them, these candidate disruptive technologies are always inferior, in many ways, to the incumbent technologies.  It’s no wonder that they nearly always come to nothing.  So it’s very tempting for pragmatic business leaders to ignore them all, and to just plan on evolution, not revolution.  Luckily, there are some clues to seeing when disruptive technologies will come to rather more than nothing!

Parallels between industries

The idea of disruptive innovations has received considerable attention in the wake of the watershed book “The Innovator’s Dilemma” by Clayton Christensen.  The book has an alarming subtitle: “When new technologies cause great firms to fail”.  The great firms fail because they use their well-honed critical skills to see all the problems with specific disruptive technologies.  Because of the problems, the great companies decide not to throw in their lots with the new technologies.  The problems are real.  As a result, the great firms prosper, 99 times out of 100.  But on the hundredth occasion, they miss the boat.  They fail to spot the rare counter-example, of the disruptive technology that transforms their industry, despite its real problems.  So they get left behind (unless they are truly great, and can turn their business on a sixpence).

There is another book that covers the same basic territory.  It’s “Mastering the dynamics of innovation”, by James Utterback.  You may find Utterback’s book the easier of the two to read.  Don’t be put off by the gee-whiz endorsement from Tom Peters on the front cover of the book, “This is the most valuable book I’ve read in years”.

Utterback draws examples from industries most of us know at least something about: typewriters, electric lighting, photography, refrigeration, motor cars, semiconductors, and so on.  Maybe one day there will be a new edition of this book, with a new chapter devoted to the phone industry.

There’s a common theme running across all these industries.  The histories of these industries contain lengthy periods of fairly predictable evolutionary development.  But then something much more seismic interrupts things.  The seismic upheavals are made possible by a combination of new product designs and new process innovations.

There can be hints of an imminent seismic upheaval ahead of times.  One thing to notice is the appearance of significant new product features in high-end models.  High-end models are more expensive, and only have a limited market.  But provided the new product features show themselves to be genuinely desirable, there will be an increasing pressure for the high-end features of one product generation to migrate into the mid-tier products in the next generation.  Electrically operated windows, central locking, air conditioning, and automatic wipers – all started life in the more expensive motor car models.  But consumer appeal provides the impetus for them to become increasingly available in lower cost cars.  That’s all right if the features can be added, through clever engineering, without significantly raising the overall cost of these cars.  But when there is a whole gamut of desirable new high-end features on the scene, the scope for cleverness in engineering is liable to expire – especially if these new features interact in wicked ways with each other (and with the core original features of the product).  Evolutionary product development no longer suffices.

Symbian OS as disruptive technology

This relentless increase in functionality requirements on phone handsets is causing the operating systems used in current mobile phones to creak at the seams.  The small, embedded operating systems adopted by handset manufacturers ten years ago were never designed for such complex applications.  They were designed to support telephony protocol software, run a rudimentary user interface, and not much more.

The complex integration involved in making a modern phone on such a simple operating system causes ever lengthening development schedules and produces ever less reliable products.

But the leaders of the mobile phone industry saw this coming.  During the early months of 1998, senior managers and technicians from Ericsson, Motorola, and Nokia, in a remarkable display of foresight, held a series of discussions on the need for a brand new operating system for future generations of mobile phones.  Senior managers from Panasonic joined them later.  They reached agreement on four basic principles:

• Their existing phone operating systems had limited life left in them
• A next generation phone operating system would be a very major investment, that would take years of effort to develop
• The different companies were all in the same situation, and had a great deal to benefit by co-operating in the identification, support, and development of a next generation phone operating system, as an open standard that exists for the sake of the whole industry
• The best starting point for this new operating system was the system then called “EPOC” that had been developed by software engineers of Britain’s Psion.

“EPOC” has since been transformed into “Symbian OS”, Psion Software became Symbian, and the rest is history.  Except that a lot of that history is still in the future.

The innovator’s dilemma with Symbian OS

Symbian OS is a long-term play.  It wasn’t designed to support phones that are fairly straightforward improvements on existing models.  It was designed to support phones going as far as ten years into the future – phones that will be seriously different from those of today.

As with all disruptive technologies at an early stage in their adoption, there are clear drawbacks in a company switching to it, away from their existing nameless operating systems.  Here are some:

• The in-house development teams are extremely familiar with the existing operating system. That OS may have warts, but they’re warts that are well known (in house).  In contrast, Symbian OS has a significant learning curve.  It makes lots of use of advanced features of C++ and object-oriented programming, has an unusual approach to memory management and asynchronous events, and has a very large set of APIs (Application Programming Interfaces).
• The many years of deployment of the existing OS in generations of different mass-market products have led to that OS being highly optimised. In contrast, while the sound design principles of Symbian OS cannot be denied, it has less battle history behind it.  Fewer optimisations have been applied to it for the sake of specific phones in the market place.
• For this reason, a Symbian OS re-implementation of a specific existing phone is liable to require more hardware – more ROM and RAM – than the implementation in the existing OS. More hardware means more cost to the end-user.
• The timing of the onset of phones with significant new functionality is hard to predict. There’s a huge amount of marketing hype around, but everyone in the industry knows that dates are liable to slip, for all kinds of reasons, perhaps by as much as years from earlier estimates.  And so it’s tempting to seek to squeeze in yet more rounds of incremental product development based on the existing OS.  Such development is bound to be quicker than a first product development on a brand new OS.  Better the devil you know than the devil you don’t know?

That neatly illustrates the innovator’s dilemma.  The disruptive technology is more costly than the existing technology, in several different ways.

Adopting Symbian OS

But the truly wise would-be innovator takes a longer-term view of things.  In this view, there are increasingly compelling reasons to adopt Symbian OS:

• There’s more to creating a new mobile phone than can be handled by the in-house development team alone. New modules can and should be bought in from external suppliers, rather than being developed in-house.  This avoids the in-house team having to be experts in an extensive array of different technology areas (connectivity, Bluetooth, SyncML, browsing, Java, multi-media, security, etc, etc).  But what operating system will these external suppliers write for?  They won’t be keen to learn the details of the nameless operating systems of each different phone manufacturer.  Far better if there is a standard operating system for next generation phones.  After a phone manufacturer switches to Symbian OS, a large community of technology suppliers suddenly becomes available to the phone manufacturer.  This community is much larger than would form around one of the existing nameless operating systems.
• Adoption of a standard operating system suddenly adds lots more value to potential industry partnerships, alliances, and joint ventures. For example, the tie-up of the advanced mobile divisions of Sony and Ericsson allows a very fruitful pooling of expertise in Symbian OS (as both companies are Symbian OS licensees and have undertaken significant Symbian OS development projects).  The tie-up would be considerably poorer without this shared operating system expertise.
• A first generation product development based on Symbian OS may indeed be more expensive than a similar product development based on the existing OS. But that first project has more outputs than the initial product itself.  Critically, it buys the company a whole lot of experience with Symbian OS.  Yes, the initial learning curve for Symbian OS is considerable.  But, once that curve has been climbed (which requires practical experience as well as theoretical study, needless to say), the full power of Symbian OS is available to the company.  And that’s a very major asset.
• The existing operating systems may still have some vitality in them, which can be squeezed out by clever engineers, resulting in products that are minor evolutions of the previous generation. However, Symbian OS brings with it support for complex new requirements, enabling revolutionary product development.  Having a much more sophisticated design, it has a much greater element of being “future-proof”.  It will prepare companies to withstand unpredictable changes in technologies, customer expectations, market requirements, and so forth.  Whereas the existing operating systems are stretched to their limits and cannot cope with significant new requirements, Symbian OS is only just getting into its stride, and can gobble up these change requests with relish.
• Symbian OS is designed for platforms, not just for products. A platform is something that enables a whole series of products.  In fact, Symbian OS is designed for multiple  Phone manufacturers can use it to create several different lines of phones, that different significantly from each other.  The very first Symbian OS product may take longer to create, as compared to a similar (low-spec) model based on the existing OS.  But subsequent products will take less and less time to create.  That is the tipping effect.  And that is the power of a successful disruptive technology.

The start of the revolution

As I write this, there are two phones in the market running versions of Symbian OS: the Ericsson R380 smartphone and the Nokia 9210 communicator.  These are just the beginning.  But they give an important foretaste of things to come:

• You would expect a Symbian OS phone to contain software from two companies: from Symbian, and from the phone manufacturer. However, in reality many more companies get into the act.  The early Symbian OS phones contain software from a large number of other contributors.  This should not be seen as a weakness – as any indication of a failure of self-reliance on the part of the phone manufacturer or Symbian.  It should instead be seen as the start of the fulfilment of the vision of an enlarged community of developers, who are collectively creating a huge array of software from which phone manufacturers can pick and choose.  One of the key attributes of Symbian, as a company, is its profound preference for collaborative solutions.  Symbian OS provides the framework for the alignment of the many collaborating partners, but the contributions of all these partners is extremely material.
• The first two Symbian OS phones are very different from each other. They are very far from being simple clones, or copies, of each other.  This heralds the power of Symbian OS to support a range of different devices, suited to different niches within the overall mobile phone market.  This also illustrates the flexibility of Symbian OS, to accommodate major ideas from phone manufacturers about how the software should operate in their devices.
• Not only do the first two Symbian OS phones have very different form factors. They also have very different UIs (User Interfaces).  These UIs are each different again from the UI on the PDA devices created by Psion, such as the Series 5mx and the Revo, from earlier versions of Symbian OS.  With Symbian OS at its core, there is no risk of the mobile phone market ending up in the same situation as the PC market.  There is no risk of the device manufacturers having a smaller and smaller role to play, with the main design of the device being dictated by the operating system provider.  Flexibility of UI system was one of the very earliest of the design goals of the Symbian OS, and the first devices built with it show how successfully that design goal has been fulfilled.
• One thing that the early devices have in common is that they provide a glimpse of what can be accomplished with a large graphics screen. The large screens make the devices bulkier than others on the market, but bring lots of compensation.  You just need to see them in action, to begin to see their possibilities.  With clearer resolution, some of the same effects will shortly be available on slightly smaller screens.  Of course, these more advanced screens require more advanced software to take proper advantage of them.  Thankfully, Symbian OS is more than fit for this task.  The resulting rich UIs will stimulate strong market demand for Symbian OS phones.
• Another thing that the early devices have in common is that they both demonstrate a strong integration of communications and messaging software with the original phone software. As reviewers have remarked, functions such as sending text messages are given very central positions within these devices, and are not in any way afterthoughts or oddities (as in some other would-be smartphones).  The three core functions of (1) contacts database, (2) data messaging, and (3) voice phone calls, all interact very smoothly with each other.  That will be a defining characteristic of a Symbian OS phone.
• Finally, the Ericsson R380 smartphone and the Nokia 9210 communicator are each examples of “one box solutions”. They are capable of replacing two devices: the user’s previous phone, and the user’s previous PDA (or handheld computer).  Before these devices came to the market, it was hard to predict whether one box solutions or two box solutions would prove to be more popular with end users.  Two box solutions (in which a phone communicates with a PDA by means of Infrared and/or Bluetooth) have the apparent advantage of not needing to compromise on the form factor of either the phone or the PDA.  However, the early market experience of Symbian OS phones confirms that there is a huge advantage to the convenience of only having to carry one device.  This bears out the fact that the Symbian OS enables an essentially new class of devices.  The more than word of mouth travels around consumers about the dramatic potential for this new class of device, the more the market will demand these devices.

The next three years

Some time within the next three years, the 10 millionth Symbian OS phone will be sold.

Compared to the eventual size of the market for Symbian OS phones, this is still the very early market.  Symbian OS phones will still be at the high end of the overall range of mobile phones on sale.  However, considerable consolidation will have taken place, strengthening the position of Symbian OS as the standard operating system for next generation mobile phones.  This consolidation will be marked as follows:

• A number of phone manufacturers will have released at least three different Symbian OS phones each. These manufacturers will have reached the next stage of the vision of Symbian OS being a platform enabler, reducing the time-to-market and development costs of follow-on products.  Quite how many such manufacturers will be in this state is hard to predict.  Next generation mobile phones are hard to create!  We are already seeing companies exiting this market, and new alliances among those left behind.
• In keeping with the emergence of Symbian OS as the new industry standard, there will be a vibrant development community. Upwards of a thousand companies will be producing high-quality commercial add-on software for Symbian OS phones.  The rich support for Java inside Symbian OS phones will play a critical role here.  Advanced Java solutions will run with higher performance on Symbian OS phones than on any other phones.
• The first Symbian OS phones are targeted at the GSM/GPRS wireless standards. Starting shortly, Symbian OS phones will also be available for CDMA and 3G wireless standards.  Most applications will be unaware of any differences between these different standards, and will run equally well in the Japanese, Korean, American, and European market areas.  Symbian OS will gain full exposure in all these areas.
• As key new technologies and standards break through into the IS and telecomms marketplace, new revisions of Symbian OS will be uniquely placed to interact with them and take advantage of them. This is on account of Symbian’s unparalleled position as the highly trusted supplier of software to the leaders of the telecomms world.  Symbian benefits from peerless industry foresight.  Symbian learns, from its close contacts, which technologies are just vapourware, and which are for real.  And so Symbian OS is ready, for the onset of the technologies that matter.
• The early Symbian OS phones lay the foundations for the solution of the so-called “chicken and egg problem”. Which comes first, next generation phone devices, or the next generation phone services that are tailored to these devices?  The existence of advanced data services adds to the appeal and the value of next generation mobile phones.  But these services are difficult and expensive to set up, so no one is going to create them, until there’s a large enough population of next generation phones to justify this expense.  In turn, there has been a worry, that no one will buy expensive next generation phones, unless there are compelling advanced data services already in place.  Chicken and egg?  Well, if you’re still reading, you’ll already know the answer to this minor dilemma.
• The answer is that the next generation devices are compelling enough in their own right, for enough users, even without the existence of new data services. The auxiliary functionality and the integrated functionality see to that.  This bootstraps the whole process, and triggers the network effects of the third wave, community functionality.  With ever-greater numbers of Symbian OS phones in circulation, able to perform sophisticated on-board processing of information, there will be added incentives for the creation and popularisation of the advanced data services that will shine on such devices.  Instead of a “chicken and egg problem”, there will be a strong virtuous cycle.  The popularity of Symbian OS phones will lead to increasing prevalence of advanced data services, which in turn will lead to increasing popularity of the Symbian OS phones on which these services run particularly well.  You can guess the rest.

The three years after that

Some time within the next six years, the 100 millionth Symbian OS phone will be sold.  In addition to all the trends noted above, only one more trend needs to take place, to allow the sales to escalate up from 10 million units to 100 million units.

This trend is one of decreasing hardware cost.  Just because 10 million people in the world are prepared to pay a given price for a wonderful new mobile phone, it does not follow that there will be 100 million people in the world willing to pay the same price.  Even although the features in that phone are just as appealing to them, this does not mean they will pay the same price.  This is basic economics.

The problem here is that a phone running Symbian OS needs, other things being equal, more ROM, RAM, and processor power, than a similar phone running one of the old-style (“nameless”) phone operating systems.  So, it costs more.  Of course, this argument only holds for the low-spec phones that the nameless operating systems can actually run in.  But these are the phones that have the lower cost, and which are the gateway to a true mass market.

Happily, four separate factors will bring the price of Symbian OS phones within the reach of the mass market:

• The general effect of Moore’s Law means that, other things being equal, you can have roughly twice as much ROM, RAM, and processor power in your phone, for the same price, if you wait around 18-24 months.
• Specialised semiconductor solutions are already under development for Symbian OS. These sophisticated integrated chips are tailored to Symbian OS, and will substantially reduce the costs of Symbian OS phones.
• Careful optimisations borne from the heat of many product development projects are steadily reducing the hardware requirements of Symbian OS. The trick here, ironically, is not to rely on Moore’s Law.  Instead, keep on looking for software solutions.  That’s a very important mindset to adopt.  It’s the mindset of the original software team that created Symbian OS.  It’s a mindset that runs counter to most of the software development in the world today, which results in bloated desktop offerings.  It’s a mindset that’s worth its weight in gold.
• As well as the general optimisation measures, new configurations are under preparation for the entire Symbian OS, which are specially prepared for low-spec hardware devices. These devices will omit some of the fancier abilities of the standard configuration, but will still provide more than enough oomph for the users to know that they have purchased a very special new phone.

Risks

Any prudent prognosticator is obliged to consider risks that might undermine a forecast.  What could go wrong with my prediction of sales of 100 million plus for Symbian OS?

I can’t speak for acts of God, for meteor impacts, plagues, super-volcanoes, area 51, or accelerated global warming.  Forgive me, I’ll confine my final remarks to the set of doubts that I most often hear – from journalists, conference attendees, job interviewees, and devil’s advocates:

• Do operating systems really matter? Don’t end users want greater simplicity, rather than greater complexity?  Well, yes and no.  Yes, they do want greater simplicity.  But, the route to greater simplicity, paradoxically, is via a greater underlying complexity.  It takes great skill to make things look simple.  It takes longer to write a short letter than a long letter.  A UI that makes the user relax and feel comfortable, is actually working a great deal harder than a UI that makes things look complicated.  You won’t get a top-class UI unless you anchor it in a top-class OS.  And in any case, although end-users want simplicity, they also want utility.  They want phones that allow them to contact the important people in their lives, easily and reliably.  And that utility requires an OS to make it happen.  End users may neither know this nor care about this, but the phone manufacturers are certainly interested!
• What about Palm OS? Aren’t there more Palm devices than Psion devices?  But, this is the wrong market to be looking at.  The mobile phone market dwarfs that for PDAs and handheld computers.  And in any case, as the whole industry knows, Palm OS is facing the same problem as the nameless operating systems that are in mobile phones today: it has reached the limit of its architecture, and cannot add significant new functionality without very major endeavours.  It is far from being future proof.
• What about Microsoft CE? Isn’t Microsoft the world’s most formidable competitor?  And doesn’t it have the deepest pockets, to fund an arbitrary catch-up effort?  Well, if required, the pockets of Symbian’s owners are pretty deep too.  And of course, you can’t buy trust.  You can’t overnight turn round a history of seeking domination in all markets, to become a trustworthy partner for telecomms companies.  If you are well known for approaching standards with an attitude of “embrace, extend, extinguish”, you can’t complain if your offers to help define a standard operating system are regarded as hollow.  Nor can money compress development schedules more than a certain amount.  That’s the core premise of the “mythical man month”.  To build an OS suited to next generation mobile phones, you have to start small, and then steadily build up the team on it.  That’s how Symbian OS was developed.  As reported by David Banks in “Breaking Windows – how Bill Gates fumbled the future of Microsoft”, the senior manager of the Microsoft CE team came to the realisation that CE was not suited for mobile phones.  He recognised that Microsoft needed to “start from scratch” with a new OS.  However, Gates rejected the plan, saying there was no time to do the job properly.  How true.
• Finally, what about Linux? Hasn’t the open source development model at last found the counter-example to the mythical man month maxim?  Yes, the open source community has produced some remarkable software.  But it’s not a miracle worker.  And in any case, the shared source model practised within the core Symbian community, with the source code to Symbian OS, achieves many of the same benefits.

Like Microsoft CE, Linux comes from a heritage that is far removed from mobile phones.  In mobile phones, efficient power management is a necessity.  That didn’t feature in the requirements for Windows, nor for Linux.  Likewise for a close focus on efficient use of memory, for avoiding memory leaks even in low memory situations, and so on.

As it happens, Microsoft has failed in numerous projects in the past, including MSN, Bob, and Interactive TV.  It’s totally wrong to regard them as an invincible opponent.  Intimidating, yes, but invincible, no.  The real competitors to Symbian OS aren’t Microsoft CE or Linux, despite what some elements the press may want to portray as an exciting “clash of the titans”.  That analysis misses the point.  The real competitors are the software systems that are selling in 100s of millions of phones already.  These share with Symbian OS the vital feature of being designed for mobile phones.  However, unlike Symbian OS, they have exhausted most of their architectural vision, and cannot succeed in driving the phone market to the next generation of functionality.