The 100 million club: some surprising facts about mobile software

[Research Director Andreas Constantinou, discusses the latest update to VisionMobile’s 100 million club, and some surprising facts about the companies that dominate mobile software]

100 million club logoWe ‘ve just released the latest version of our 100 million club: the watchlist of software companies whose products have been embedded on more than 100 million mobile handsets.

In this H2 2008 update we ‘ve identified 26 software products from 21 companies which have shipped on more than 100 million handsets cumulatively as of the end of 2008. We ‘ve had a few important changes in the who’s who of the 100 million club; the introduction of HI Corp’s MascotCapsule 3D, a graphics acceleration software that has shipped in 490 million mobile devices as of the end of 2008, and embedded in Japanese, but also European handsets. Other noteworthy changes are due to the consolidation that is underway in the mobile industry; Nokia completed the acquisition of Symbian in November 2009. Esmertec merged with Purple Labs to form the Myriad Group (with 2 products in the 100 million club; Esmertec’s JBed and the ex-Openwave browser). Nuance acquired Zi Corp (as part of its string of 15+ acquisitions in the last 4 years), making Nuance the only company with 3 products in the 100 million club.


(click to go to the download page)

Traditionally we have looked at the cumulative shipments of mobile software products (the orange-red bars on the chart) – and the sea of challenges that keep them constrained to a small portion of the one-billion-a-year handset market. In this update we have also compared the 26 products in terms of penetration in the mobile market as part of the devices sold (the blue bars on the chart).

What are some of the most popular software products in mobile? Looking at the headlines one might suggest the Symbian operating system, or the Opera Mobile browser. In reality Opera and Symbian/S60 are in only 2% and 6% respectively of the devices sold in 2H08. There’s far more successful companies in terms of penetration of the sales base:

– ENEA’s OSE: Founded in 1968, ENEA is a Swedish software & services company which offers network management software, development tools and real time operator systems. The OSE RTOS forms the basis for both radio stacks and application stacks for many handset models from Sony Ericsson, Samsung, Nokia and others. All in all, we estimate that OSE has been embedded in 32% of all handsets shipped in 2H08.

– Mentor Graphics’ Nucleus: Founded in 1981, Mentor Graphics is a US-based  hardware and software design solutions. Its Nucleus real-time operating system has powered both radio stacks and applications in billions of mobile handsets – we estimate that Nucleus is embedded in 34% of handsets that shipped in 2H08. The secret behind Nucleus’ success is its revenue model which is based on per project or site fees, rather than per-unit royalties.

Adobe’ Flash Lite is another success story. Flash Lite has been embedded on over 950 million mobile devices as of the end 2008, hitting the 1 billion installed base in 2009. Unfortunately, a large percentage of Flash Lite installations is closed to third party developers, which Adobe is now trying to fix with the Open Screen Project. It’s interesting to note that under OSP, the Flash 10 runtime will be available for zero royalties for product implementations which meet 3 criteria: a) the Flash runtime has to be certified for compliance with Adobe’s test suite, b) the runtime is open to developers and third party content and c) the runtime is updateable over the air, so that the installed base can be continually brought up to the latest version.

There are many more notable software products with high penetration within devices sold which often shy the headlines: Myriad Group’s (ex-Openwave) browser (still at 24% of the sales base due to feature phone embeds), Beatnik’s MobileBAE audio codec (21%), BitFlash’s SVG engine (18%), NXP Software LifeVibes audio/video middleware (23% – which recently also broke into the ‘500 million club’), Red Bend’s vRapid Mobile firmware update technology (18%) and Nokia’s Series 40 operating system (19%). Last but certainly not least, Nuance’s (ex-Tegic) T9 predictive text engine is embedded within an impressive 56% of all devices sold.

We ‘ve analysed other noteworthy aspects and insights of the 100 million club in previous articles:

– Only 26 products have made the 100 million club, a tiny figure compared to the 250-300 companies that license embedded mobile software products – not to mention the circa 30,000 mobile software developers (see analysis in the earlier article mobile software is dead.. long live mobile software).

– The emergence of de facto software standards like Flash Lite and WebKit (see our analysis in the earlier article The 100 million club: the bigger picture of mobile software), compared to closed-door standards like the LiMo Foundation (see our critical analysis on Why the LiMo Foundation needs to go back to the drawing board)

– The challenges of pre-load mobile software vendors; long sales cycles, deteriorating per-unit royalties and costly product adaptation (as highlighted by Morten Grauballe’s original The inner secrets of the 100 million unit club which provided the inspiration for the launch of VisionMobile’s 100 million club)

Comments welcome as always.

– Andreas
twitter: @andreascon

Hardware is the future of mobile software, and vice versa. What Intel’s acquisition of WindRiver really means.

[Intel buys embedded software company Wind River with a 40% premium. What is the hidden asset that silicon vendors value beyond standard metrics? Guest blogger Andy V. O’ Lay believes there is no more point in differentiating hardware and software in the mobile arena. The race to SHW has started…]

road-mergeIt’s not the first time in the mobile industry that a silicon vendor acquires a software company. Let’s not forget that behind the success of Qualcomm uiOne, there was Trigenix. Some also see beyond the failure of TI the shadow of the missed acquisition of TTPcom. These moves often prove decisive a posteriori. Here is why Intel and Wind River have set an important milestone in our industry.

A previous post (Mobile software is dead… long live, mobile software) highlighted the shift of mobile software development to post-load applications, enabled by the sudden proliferation of App Stores. Among the roadblocks in the embedded, pre-load business, two major issues were stressed:

– For software vendors, it is difficult to fight commoditisation and the related price-erosion while scaling up shipments. Only those vendors who have some unique IP managed to go down this route.

– Beyond the first design-win, deploying software across a variety of platforms and customers is a logistical nightmare. Less than 10% of the 250 mobile software companies we know have made their way to VisionMobile 100 million club charts.

Since then tables have turned and the mobile embedded software landscape has been considerably modified. Intel acquired Wind River, one of the champions of mobile software deployment, pioneer alternative business models favouring professional services fees for software integration, customisation, certification and indemnification.

The Intel-Wind River transaction is worth considering, not only because Intel agreed to pay a 40% premium over the stock price, neither because of the $884 million price tag (quite a lucky number even outside of China). It is a meaningful acquisition because it values mobile software companies on a totally unusual scale: what if IP was not a software company’s main asset?

Hardware & Software: Just Married

What Intel has bought and why
The press release claims that “the acquisition will deliver to Intel robust software capabilities in embedded systems and mobile devices”. Beyond such obvious statements, we must consider the fact that Wind River is all about delivery. The company brings enviable competence to ‘irrigate’ equipment manufacturers with new technology and more importantly, help with the painful integration process that will transform a collection of best-of-breed modules into a coherent product.

This type of competence is fundamental for an ambitious silicon vendor like Intel with PC-centric views on the mobile market. Indeed, it can take years to transform a design-win into a running line at the wafer fab (silicon speak for manufacturing line) and the key to hardware success is software.

We are not talking of specific software IP modules like a protocol stack, a midi engine, a browser or a codec. We are talking about system skills, and the capability to support large manufacturers on their own premises with outstanding engineers. We are talking about blurring the line between the system provider own workforce and the one of its customers. We are talking about permeable boundaries between supply and demand. And like Intel, I would argue that mobile software companies are instrumental in making silicon solutions pervasive, because they tick two major check boxes: reference design and support.

The hidden asset of mobile software companies
Mobile embedded software companies (e.g. Myriad, Access, Aplix, NXP software, Azingo) have a unique understanding of products as a hardware/software system. They understand how silicon can be leveraged, encapsulated and productised to fast-track device delivery. Beyond specific IP, these companies have the capability to build the very first product that will feature a new chipset. This is the reference design check box.

A reference design is not a half-baked breadboard; it is a scale 1:1 device, ready to ship.

The absence of such capability has drained Texas Instrument, once the mobile silicon leader. Alternatively, true reference design has been instrumental in the success of Qualcomm and Mediatek.

Mobile software companies have a structural asset: in order to promote the adoption of their modules, they have deployed local workforces to support each of the big handset OEMs. Their engineers are often working on the same premises as their customers, they eat at the same table, drink the same coffee (or tea) and solve issues together. On a P&L sheet, they call it Professional Services. On the Balance sheet, it should be called customer relationship. This is the support check box.

Support is not a way to charge more than just royalties, it is about making sure that the product will actually ship. Support is the last mile, the distribution channel that delivers a facility right into the customer premises. Support is what made Ericsson or ST successful; it is what was valued when NXP was acquired… and this is the rationale for Wind River’s acquisition.

What Intel has incidentally bought
The power to harm is seldom the main reasons for investing, rather the cherry on the cake. Whether used as dissuasion or tactical weapon, it should not be disregarded. There are a few nasty decisions that the “new” Wind River could be taking, which would seriously impact competing developments. Yet, it would be interesting to understand what is Intel’s competition: Qualcomm? Google? LiMo?

Along the same lines, it is interesting to note that although Wind River does not pretend to be an IP company, it has acquired almost accidentally an interesting patent from FSM Labs regarding virtualisation. I wouldn’t think this patent has a wide scope, but in the hands of Intel lawyers, it could scare some people off… and more about virtualisation in a future post.

Where does the industry stand today?
On the silicon side, we have an interesting situation with 3 big vendors.
– Intel has just acquired support capabilities, yet they miss a fully baked reference design.
– Qualcomm has a complete reference design, but they lack support capabilities, especially to address European markets.
– ST-Ericsson is difficult to read, they should have capabilities both in terms of reference design (from Ericsson) and support capabilities (from ST and NXP). Yet it is unclear whether such capabilities will survive the painful merger exercise.

On the software side, with Nokia acquiring Trolltech last year, and now Intel buying Wind River, the number of mobile software companies with a system-wide scope (capability to write/integrate modules at all levels of the software architecture) and Linux capabilities is reducing quickly: Access, Aplix, Myriad, NXP Software… who will be next?

The key point here is the hidden assets. Mobile software companies are bought for what they enable, not what they earn. Trolltech allowed the creation of a single application environment on top of Nokia S40, S60, and Maemo where Nokia can base both its Ovi *and* its ‘signature’ apps. Wind River equips Intel with a heavy-duty support channel. Access and Myriad both have an important hidden asset: operator-compliance. Two words that Apple and Google are slowly learning to spell.

The future is reunification
So Intel, who was already lining up hundreds of software engineers solely dedicated to Moblin (not counting the 5,000+ Moblin community), is now acquiring 1,600 software engineers who will “more tightly align their software expertise to Intel’s platforms”. This is a visionary move. Hardware (HW) and software (SW) guys realising that they need each other to grow.

So what does this mean?
– HW needs SW to sell.
– SW needs HW to scale.
– there is no longer HW and SW; there is SHW (read Systemware, pronounce chew, and remember you read it here first).
On the mobile racetracks, there is no room for two-seaters. Wintel has lost its initial, Qualcomm never got one… We all are Berliners.

– Andy

[Flashback. Three years ago. An influential venture capital firm gathers its telecom think tank off-site. During the wine-tasting session (wine-tasting is the venture-capitalist equivalent of engineers coffee machine) somebody asks: what is the only billion dollars mobile software company? Silence. More wine-tasting. No spit. Then comes the answer: Qualcomm. Hardware is also the future of mobile software.]

Bright thinker looking for bright readers? Join us at the VisionMobile blog, the stage for mobile industry thinkers.

The Amazon Kindle: More revolutionary for the mobile telecoms industry than the iPhone ever was

[The iPhone has ushered in a new era of user experience on mobile hardware. But the business model Amazon negotiated with Sprint set a precedent that could radically reshape the future of the industry, writes guest blogger Stefan Constantinescu]

upside-downWhen Apple CEO Steve Jobs got on stage in January 2007 and announced the iPhone, the world collectively paused, took a deep breathe, and then yelled at the top of their lungs with joy at a device that not only changed their perception of what can be done with something that fits in your pocket, but how one interacts with small screen gadgets in general.

In Europe people smirked; EDGE only, 2 megapixel camera, no MMS, is this a joke? Contrast that to America, which at the time was still known as the “Land of the Motorola RAZR,” the radical idea of having the full internet in your pocket was new and exciting. Nearly 10 months later, in November 2007, it would be Amazon’s CEO Jeff Bezos that would climb on stage to show off his device: the Amazon Kindle; expensive, single purpose, limited content, by some accounts it was quite difficult to look at as well.

Why is it then that the Kindle is more important to the mobile telecommunications industry than the iPhone?


The Kindle is the first device to be sold with lifetime cellular connectivity included in the purchase price and therefore it is the first device to carve out a path towards a new business model for operators.

Apple had the opportunity to change things. They could have sold the iPhone unlocked from day one and educated the American public about SIM cards and why buying a device from an operator on a two year contract is unwise. They could have launched the iPhone internationally, unlocked, without having to negotiate with operators due to the fact that many people in Europe and Asia are used to paying full price for their device and buying a SIM card + service separately. They could have prevented the large exodus of iPhones that were meant for the American market, but ended up on the international grey market, from ever happening, but they didn’t.

Apple launched a revolutionary device and to buy it you had to go through the traditional channels.

Amazon’s Kindle however, once you purchase the device and turn it on, it connects to Sprint’s network automatically without any user configuration. Unlike buying a subsidized netbook from an operator today, where you still have to pay a monthly fee, the Kindle is connected for life after you make that initial purchase.

Intel has already admitted that the speed-wars are over and now their future will be focused on high volume shipments of their Atom processor, they even licensed their Atom intellectualy property to TSMC; the goal being to connect more and more devices to the internet. With all of these new devices connecting to the network, be it our cars, our refrigerators, our power meters, our televisions, anything and everything, how exactly does one enter their WPA2 security key on a toilet which has a single button, flush?

This little convenience, connectivity out of the box, has huge ramifications for the mobile industry if operators choose to play their cards right. We’re entering an era defined by people’s expectations of being able to browse the internet and access their favorite services on most, if not all, of the new devices they purchase.

Today operators cry foul when people demand that they turn into dumb pipes. Operators today still believe that innovation occurs at the core of the network, versus the edge. Operators come up with poor reasons, even poorer attempts at new businesses, and some are beginning to adopt defensive tactics such as limiting what can be done on their network in order to protect the business models that have been allowing them to expand for nearly 3 decades.

The Kindle was the first step in a new direction. Sprint effectively became a pipe for Amazon’s customers to purchase books and read Wikipedia on an electronic ink display. These new devices that will soon connect to the network, the cars, the televisions, the toilets, can either depend on users being knowledgeable and willing to configure the correct settings for access, or the device manufactures themselves can negotiate with operators beforehand to allow said devices to have connectivity out of the box.

According to an interview with Glenn Lurie, President of Emerging Devices for AT&T Mobility, a unit that opened in December 2008, his goal is to “develop relationships in the ecosystems around […] devices and launch those devices wirelessly enabled.” Later he added “you can imagine we’re talking to every OEM on the planet, there are a lot of people that build devices.” I’m going to speculate here and say that this unit, Emerging Devices for AT&T Mobility, was launched as a direct result of the Amazon Kindle.

The revenues operators can expect to receive from device manufactures will start small. Roger Entner, SVP, Nielsen’s Head of Research and Insights for Telecom, estimates that Sprint is receiving only $2 per Kindle subscriber per month, but just as data traffic, and in turn data revenue, leapt passed voice on landlines, the same will happen sooner rather than later with mobile operators.

The question is: are operators ready to experiment with new business models, billing methods and dealing with new customers that are device makers versus the individual?

The modus operandi we’re used to today is operators buying hardware, attempting to create a unique software experience, and then selling the final product to the consumer. In a brave new world why can’t it be the device makers who go to the operator, buy network access in advance, and then sell their devices directly to the consumer?

People would not have to buy network access and therefore churn, meaning customers leaving your network to join a competitor’s network, would be reduced. People would not have to care about paying a monthly bill, since it is the device manufacture covering the expense. People would no longer be tied to 1 or 2 year contracts and be stuck with a device they dislike; they would simply use a gadget until they no longer fancy it and buy another.

The benefits for the operators are clear. One customer, paying one sum of money, for one month of access, for one device, one customer you have to compete for every 1 to 2 years due to their contract expiring, is a profitable business to be in, but it isn’t forward thinking since the size of that market is limited to the population of a city, state or country. Having device manufactures purchase network access, with the amount of devices a consumer has today and will probably own tomorrow, has the potential to push penetration numbers past 200%, even 300%. Less money will be spent on advertising the operator brand since it becomes irrelevant. Less money will be spent on hiring software engineers to create those unique software experiences on devices. Less risk of ending up with excess stock somewhere in a warehouse because the devices an operator purchased for the Christmas season were not as popular as predicted.

The benefits for device manufactures and the consumer are even more clear. A greater number of devices connecting to the network, more services being used, zero headache configuration, unlimited access. Additional revenue can be extracted by charging more for a device to maintain a small margin on the network access or by partnering with service providers to make their service the default option.

The Amazon Kindle carved out a new business model, time will tell whether or not it becomes the de facto revenue generator for operators. They are pipes after all, but why is that such a bad thing again?

[Stefan Constantinescu is a guest blogger, currently job hunting, a former Services Strategist part of Nokia’s Corporate Strategy Team, and a former blogger with IntoMobile.]

Bright thinker looking for bright readers? Join us at the VisionMobile blog, the stage for mobile industry thinkers.