[Purple who? Research Director Andreas Constantinou analyses the recent acquisition of Openwave’s client business by Purple Labs and how a relatively unknown French software vendor is becoming a key player in the new world of open source software  for mobile phones.] acquired the client business of Openwave, bringing together a Linux-based software stack with a browser and messaging suite. The acquisition brings Purple Labs in head-to-head competition with the Access Linux Platform (ALP) and Azingo, a US-based vendor of Linux-based operating systems for smartphones. It also makes mobile open source one of the main battlegrounds of the mobile industry today, following close on the heals of the Symbian Foundation announcement (see full analysis here). Based in France, Purple Labs has transitioned from a little known design house arm of Vitelcom (a Spanish ODM) to a well-funded 100-strong OS vendor boasting the only commercial single-core 3G Linux software stack and three Linux mid-tier phones shipped in Europe since 2006. The transformation has been led since late 2007 by Sofinnova Partners, Earlybird Venture Capital, and Partners Group who have invested $15 million in addition to supporting the $30 million acquisition of the Openwave client business. This puts Purple Labs on a level playing field with Access Linux Platform (backed by ACCESS, a billion-dollar valuation company) and Azingo (funded with $30 million to date) and should help Purple Labs build its professional services team which is a critical element of a Linux-based OS business. It’s also worth noting that Purple Labs is so far the only vendor with a commercial 3G stack, and one which has been integrated on a single core handset, both of which are major engineering achievements. It is also the only vendor of licensable software stacks (alongside Mizi who had shipped its OS on 3 Samsung phones) and the only Linux-based software stack which has shipped on European phones which come with demanding requirements for GCF certification and operator (incl. i-mode) customisation. Too good to be true ? The next six months will tell, as ALP, Azingo and Purple Labs are preparing to launch their next handset projects and OEMs are being called to choose between S60/Symbian, Android and Linux stacks. Purple Labs is led by Simon Wilkinson and the ex-Magic4 management team who led Openwave’s client business. The same team has now effectively led the acquisition of Openwave’s browser and client messaging products including all code, patents, customer contracts and the engineering team. Bringing Openwave back from the ashes Opewave has suffered a major decline in the last year. In May 2007 it was announced the business was up for sale, it’s looking for a new CEO and its share price (OPWV) has dropped to less than a fourth of what it was one year ago. This is the same company who had shipped software on more 1.4 billion handsets with 50 device manufacturers (as it still mentions on the ‘about’ page). In its heydays in 2005 Openwave was celebrating 1 billion handsets shipped with its browser and launching MIDAS, a then-visionary service delivery framework, the same strategy that underlines today’s widget adoption by OEMs and operators. The Openwave browser business has been perhaps one of the first victims of the open source phenomenon (alongside Obigo); it was open source development practices which allowed the likes of Apple, Nokia and Google to build next-generation browsers by pooling development costs through the WebKit project – see our seminal article Bye Bye Browser for a full analysis of the story. Since MWC 2008, Openwave has refocused its business on network service management products – including content optimisation, mobile advertising and personalisation. The $30 million that Purple Labs paid for the acquisition is peanuts compared to the valuation of Openwave’s client business circa 2005, considering that browsers typically command $1 per-unit royalties (volume dependent) and Openwave had around 50% market share of browser shipments. [update: For 1Q08, Openwave had reported $11.1 million in revenues from its client business, including software license fees and related engineering services]. Purple Labs also inherits the WAP browser business where WebKit competition is not yet relevant, as well as the MIDAS scripting framework, which can be assumedly be repurposed into an on-device service delivery platform. If this doesn’t sound familiar, on-device platforms for service delivery are what Qt, Silverlight, Java FX Mobile, Qualcomm’s Widgets, Flash Lite and Tamarin are all about. I would assume Purple Labs plans to reuse Openwave technologies to help address the requirements of LiMo Foundation members such as Vodafone and Orange. The open source battleground is consolidating and intensifying. Watch this space! – Andreas [want to know more about open source use and best practices in the mobile industry? Check out our 360 degree workshop on mobile open source.]

[Nokia celebrates Symbian’s 10 year anniversary with an acquisition and a royalty-free, open source roadmap for S60. Research Director Andreas Constantinou distills the ramifications of this major industry announcement] Symbian Foundation in celebration of the company’s anniversary since its foundation in 1998. Exactly ten years later, having shipped 200 million devices across 235 models from the top-5 OEMs and built an ecosystem of 4 million developers, Symbian now undergoing a major transformation process. The facts – Nokia is to acquire Symbian Limited with closure expected by the end of 2008, subject to regulatory approval. On closure, all Symbian employees will become Nokia employees. – Nokia will be spending  EUR 264 million to buy the remaining 52% of Symbian shares from Sony Ericsson, Ericsson, Panasonic, Siemens and most likely Samsung. This is about 2.5x of what Nokia paid to acquire Trolltech and about 20x less than what it paid to acquire Navteq. Symbian makes about GBP177 million a year (calculated as 7% increase over 2006 revenues, in line with 1H07 y-o-y growth), so in Nokia’s share offer, Symbian was only valued at just over two times its annual revenues (!). [Updated: ARCchart has a good analysis of how Symbian’s valuation has remained pretty much constant of the last 5 years despite have shipped its OS on 200 million devices.] – A new, complete operating system is to be formed by merging Symbian OS and S60 in 1H09. Sony Ericsson and Motorola intend to contribute technology from UIQ, while DoCoMo intends to contribute MOAP-S assets (the Japanese flavour of a Symbian-based application platform). – The platform will be managed by a new entity, the Symbian Foundation. The foundation will be formed in 1H09 from Nokia as well as several partner OEMs (Fujitsu, LG, motorola, Samsung, Sony Ericsson), network operators (AT&T, DoCoMo, Orange, T-Mobile, Vodafone), hardware platform and chipset vendors (Broadcom, Ericsson, Freescale,  ST, Texas Instruments), system integrators (Digia, Teleca, Wipro), content providers (EA Mobile) and software vendors (PlusMo). – The Symbian Foundation platform will be available a royalty-free license to all foundation members – starting in 2H09 and to be completed in 1H10 – Platform assets will be available  under an open source license gradually over the next 2 years, with the intent to use the Eclipse Public License (EPL) 1.0. EPL is a weak copyleft license which means that source code modifications/derivatives created must be published if distributed but can be combined with proprietary software. There is also an explicit patent grant in the EPL which implies that IPR-encumbered code from third parties is unlikely to find its way into the Symbian Foundation platform. – The foundation will operate as a meritocracy. Device manufacturers will be eligible for seats based on number of Symbian Foundation platform-based devices shipped, with the other board members selected by election and contribution. – The foundation will provide a single point of access for developer support, offering SDKs, documentation, samples, knowledge base, application signing and tech support – The Symbian Foundation platform will be backwards compatible with Symbian OS 9, S60 3rd Edition. Note that this backwards compatibility warranty isn’t offered by either Android (which may suffer from fragmentation by design) or by LiMo (which effectively standardises middleware and kernel, less so the application environment) – The foundation will be open to all on a low $1,500 yearly fee from 1H09. Being a foundation member gives you rights to access, modify and contribute foundation source code, access meeting minutes and documents, participate in working groups and annual member meetings. What this means for the industry So what does this mean for the mobile industry ? The repercussions of this announcement are rather disruptive: – Nokia will increase its control over Symbian OS, not through ownership, but through the governance model of the Symbian Foundation the sheer weight of its contributions (an estimated 1,000+ Nokia engineers working on the S60+Symbian OS). Previously Nokia was shipping around 70% of Symbian-based devices, but only had just under 50% of ownership. Share of shipments and control (board seats) are now aligned. [update: Nokia gets only one board seat out of five manufacturer seats initially – board seat assignment is volume based and this allocation may change long term. In effect control comes not through board seats, but through the weight of Nokia’s engineering efforts and share of code contributions relative what other OEMs can dedicate to the project.] – This is a logical move for Symbian, which was crippled without control of the UI, application stack and the core OS under the same entity – The current Symbian OS license will be extremely simplified into EPL, a popular weak copyleft license that’s been tried and tested across 10s of commercial projects by the Eclipse Foundation. – Sony Ericsson and Motorola (whatever its fate is) should be adopting S60 and dropping UIQ. That’s a major change, especially for Sony Ericsson who had built 6+ current phone models on UIQ. It’s also a sign for lay offs at UIQ who had been building a team of nearly 400 people in Ronneby and Budapest [update: UIQ will be laying off 200 out of 375 staff] – DoCoMo should be replacing MOAP with S60, again a major undertaking for Fujitsy, Sharp and Mitsubishi who had shipped over 60 models to date on MOAP. – Unlike LiMo, the Symbian Foundation will be controlled by a single player, Nokia, based on the weight of its code contributions its ownership of Symbian and the shipment-based assignment of board seats. This governance model is effectively similar to the Open Handset Alliance which is controlled by Google and the WebKit browser core development, which is owned by Apple (although the underlying mechanisms are quite different) – Windows Mobile is the only licensable OS for mid/high-end phones which doesn’t have a consortium-based contribution model and an open-source-like license (apart from selected parts of the Windows CE kernel source code which are under varying Shared Source licenses). I would expect Microsoft to react in the next quarter by open sourcing more of Windows Mobile – If Android signalled the commoditisation of the mobile operating system business, the Symbian Foundation platform is the nail in the coffin. This will make things rather challenging for many Linux stack vendors (especially those targeting the smartphone market) who charge on a royalty basis. I expect to see revenue models move towards professional services fees from certification, customisation and productisation (see earlier post on open source business models). What triggered the disruption The announcement surprised almost everyone in the mobile industry. It was hard to predict, but is not too hard to rationalise. It may be argued that there are two reasons for Nokia’s acquisition and open source roadmap: Firstly, as a royalty-free, open source licensed OS, Android was too hard to resist for any OEM. Nokia’s acquisition of Symbian is essentially the answer to Android, resonating the same core principles which are a) royalty-free b) open source licensing for pooling costs of maintaining a commoditising OS and c) majority ownership by a single player [Update: Nokia (via Michael Mace’s blog) clarified that they only own one board seat out of 5 OEM seats in the Symbian Foundation. Which implies that Nokia would have comparatively little control over the Symbian Foundation – however in reality Nokia will be dedicating an estimated 1,000+ engineers to the S60+Symbian project, far more than any other OEM can, in effect biasing the roadmap of the OS towards its own agenda. This works in much in the same way that Apple control open source browser project WebKit.] Secondly, Motorola was bleeding heavily in a financial sense and so it would have been keen to sell out its UIQ shares, which Sony Ericsson could not assume the burden of as UIQ was a long way from being profitable. With UIQ out, S60 was the only Symbian-based alternative for Sony Ericsson and a strategic choice in keeping Google from becoming the Android-inside of the mobile industry. First reactions The Nokia+Symbian disruption is now creating three centres of gravity around licensable mobile software: LiMo, Android and the Symbian Foundation (one could also add Qualcomm BREW for specific markets). Besides layoffs at UIQ, the repercussions to the industry will be far-reaching, but the full story will take time to unravel – after all the first Symbian Foundation devices are not expected until 2010. Interesting times indeed! – Andreas [want to know more about open source use and best practices in the mobile industry? Check out our 360 degree workshops.]

[True, there are no textbooks on making money from open source, so how do you go about it? Research Director Andreas Constantinou explains the key revenue models used in mobile open source] Bill Weinberg says. We ‘re delivering our mobile open source workshop on June 12th in Berlin, as part of Informa’s Handsets World conference. There’s still places left, so if you can still sign up if you hurry (check the product datasheet for the agenda). Building the workshop has been a huge learning experience over the past two years – from interviewing tens of insiders involved in mobile open source to developing our understanding of the intricacies of open source with each workshop that we deliver – intricacies such as business models, community governance models, licensing best practices and the hard realities of building a Linux OS. Here I ‘ll explore the revenue models that make mobile open source tick; in otherwords how can a new product strategy make or save money by open source as a key element of the business model. The following slide is taken from our 360 OSS workshop: Revenue models for creating a product from FOSS: 1. Per-unit royalties. Who said open source was free? While the Linux kernel may be accessible to anyone with a web browser (subject to GPL terms), there is a huge leap between a kernel and a fully integrated, optimised, customised, certified and stable operating system. That’s why vendors like Azingo, ALP, Purple Labs and Mizi Research do charge royalties for the Linux-based software stacks. 2. NREs (non-recurring engineering fees) for integration & productisation. Most open source projects are designed to be 90% complete.. but the remaining 10% of pushing a project to ‘shrink-wrap’ product status requires an entity with commercial interests to the deliver the project to the finishing line. As such, system integrators and software vendors such as MontaVista and WindRiver will happily engage in integration and productisation project for Linux-based OSes, in exchange for professional services or NRE fees. 3. Subscriptions for product updates & support. This revenue model is common with dual-licensed open-source products, where the product is branched into a version that’s available under GPL non-commercial terms and one that’s available under commercial non-copyleft terms. Companies like Funambol, Volantis, and Trolltech offer paid-for subscriptions to product updates as a service to customers of the commercial product branch and an incentive to move from trying the GPL branch to to buying/licensing the commercial branch. This revenue model presents a growing opportunity for any system integrator involved in the mobile industry, as both device-side and network-side software products based on open source are becoming increasingly used, while at the same time lacking support contracts and service level agreements (SLAs) that customers have come to rely on. 4. Certification and compliance testing fees. In the case where open-source-based products need to be certified or pass a compliance test – as is the case with Java JSRs – an additional fee may be leveraged for undergoing these tests – as is the case with the TCKs for Sun-owned JSRs, specifically the phoneME MIDP2 implementation. 5. Hardware sales. A more subtle revenue model is that of making the software available for free, but charging for the hardware. Taiwanese manufacturer FIC practices this model for OpenMoko, the distribution which is almost 100% open source. Here customers have a reason to go to FIC to build OpenMoko-based devices for them, so as to leverage from the product know-how and hardware integration expertise that the manufacturer has on OpenMoko. 6. Insurance for product liability and indemnification. This is a straightforward insurance service that software vendors often provide as a premium, which indemnifies or insures the customer of an open-source software product against liabilities. 7. Sharing development costs. Last and certainly not least, open source licensing can be used as a modern approach to shaving costs off software development, by pooling that development effort across multiple industry participants. Companies participating for example in Eclipse, Webkit, Maemo and Android projects seek to share their development costs of a commoditising software base with other peers (even competitors), while leveraging on that base to build essential value add. For anyone attending Handsets World next week in Berlin, you can catch us at the event (Timo, George or myself) and sign up to our mobile open source workshop on Thursday 12 June. – Andreas

We ‘ve just launched a new survey for mobile application developers. If you ‘ve programmed for BREW, S60, UIQ, Android, Windows Mobile, Java, Palm, Linux or Flash Lite, we want to hear what you have to say. The survey spans across a wide range of topics, to gauge how mobile developers feel about IDE features, ease of debugging, emulator glitches, support forums, documentation & sample code available, application portability, pains of going to market and desirability of new features like scripting and POSIX support. We have announced a prize draw for each developer who completes the survey – a $1,000 Amazon voucher, more than enough to get hold of a snazzy new phone to test your applications on! The survey will run for four weeks (closing on Friday 27 June) and the $1,000 winner will be announced on Friday 4 July 2008. Are you are a mobile application developer ? Love or hate your mobile OS ? Have your SAY and a chance to win $1,000. – Andreas

The iPhone has created a revolution and has changed the mobile landscape significantly more than any other handset. Surprisingly, it is not its amazingly user-friendly UI, ultra cool design or the ease of integration with iTunes. It is the revenue model that redefined the business dynamics between handset manufacturers and mobile operators. Other manufacturers are now trying to follow the same path. Apple has been offering the iPhone exclusively through four selected mobile operators in Europe and US, after having negotiated revenue share agreements. Mobile operators have agreed to give much more to a handset manufacturer than they usually do: AT&T has been rumoured to be giving $150-200 back to Apple for each iPhone plus a percentage of monthly revenues from the subscribers that have signed up for an iPhone contract. Orange, O2 and T-Mobile are expected to have signed similar agreements. A back-of-the-envelope calculation yields the following:Addressable market (1Q 2008):153 million usersMarket penetration:3-5%Shipped iPhones (1H 2008):4.02 millionRevenue per iPhone sold:$153iPhone Published revenues (1H 2008):$619m Apple is the only manufacturer that could demand this from mobile operators and get recurring revenues. By doing so, mobile operators have access to high spending customers (in essence, this is a form of negative Channel ARPU – read previous post). So can Apple do the same with the new version of the iPhone? First of all, lets look at how you could, can and will be able to get your hands on an iPhone. How you could get an iPhone From its launch until April 2008 and if you lived in the US, France, UK or Germany, you could walk in an operator retail outlet and sign up for an agreement and get an iPhone. Or you could go to an Apple store, buy an iPhone (the 16GB version was sold in both operator and retail shops for the same price) and activate it through iTunes, where you had to enter a contract to activate your iPhone. How you can get an iPhone From April 2008 until today (May 2008), Apple stores list all iPhones as unavailable (in all countries where it has been previously sold). If you currently want an iPhone, you can only get one from an AT&T store (but you need to be an AT&T customer or sign up). How you will be able to get the iPhone 3G Instead of having exclusive agreements with few operators, Apple has signed blanket deals with operator heavyweights: Vodafone, Orange and several others. In parallel, Apple is expected to be selling the iPhone 3G in retail shops without getting mobile operators involved (yes, this has happened in Germany but the iPhone has been selling there at €1000 – just to conform with legislation). What has changed? Apple may have realised that exclusive deals with mobile operators may provide sustainable high-margin revenues but at the same time this limits the scale of volume shipments due to the smaller number of exclusive agreements possible. The original iPhone revenue model has been disrupted by unlocking and shipping to grey markets (1 million iPhones are reported to have bypassed the activation process) which has deprived Apple of its recurring revenues. Nevertheless, the iPhone ecosystem includes iTunes and even unlocked iPhones can generate revenues for Apple when a user buys content through the online shop. Why has Apple changed its strategy? Apple may have realised that singular, exclusive agreements cannot provide significant revenues. In order to bring in more revenues, higher volumes are necessary and possibly smaller revenue shares in order to lure mobile operators. By doing so, Apple can get access to a much larger addressable market, have a larger market share even if the profit margins are not as high as those of the original iPhone. Apple may have learned its lesson: restricting user choice is not a good thing and users are most likely to find a way to get their hands on an iPhone. A sign of the new strategy is that the iPhone is not available in Apple Stores (where a user can buy it and unlock it) but only in AT&T where you need an agreement with the operator to buy one. Apple may have ran out of stock but I find it quite unlikely that the manufacturer goes out of stock before the mobile operator. Apple may be creating a new market for its new iPhone 3G – the retail environment where anyone can just walk in and buy an iPhone with no strings attached. I suspect that the iPhone 3G will be selling in many markets without a contract – or with a contract and subsidized. Lets assume that Apple will make a modest $50 on each iPhone 3G sold and that iPhone penetration will remain at 3-5% (both are likely to be higher). Total revenue from iPhone sales for 2H 2008-1H 2009 can be estimated as:Addressable market (2Q 2008):575 million usersMarket penetration:3-5%Expected shipments (2H 2008-1H 2009):23 millionRevenue per iPhone 3G sold:$50Estimated iPhone 3G revenues (2H 2008 – 1H 2009):$1.15bn I would argue that the iPhone 3G will cause a bigger revolution in the mobile market: it will expose its design, UI and software to a much larger audience and increase pressure on other OEMs to develop more advanced software and UIs (see the S60 Touch UI, Sony Ericsson XPERIA, Nokia Tube, Philips X800, HTC Diamond). I guess we will all have to wait for WWDC for Steve to pull one of his usual product releases. The iPhone 3G is suspected to be made available shortly after, so I will hold my breath until then. – Dimitris

[how are the revenue models changing in the mobile industry? Research Director, Andreas Constantinou introduces Value Quadrants, a tool that deciphers the multitude of revenue models and maps how value creation is changing in mobile]. Take for example content; the sale value of content is constantly decreasing as sideloading and PC-loading are becoming the norm route to get the latest ringtones and wallpapers. In 4Q07 m:metrics reported that the percentage of subs buying a ringtone fell consistently over 12 months in UK, France, Germany, Italy and Spain. The value is in content remixing and communication (see our Mobile Megatrends 2008 presentation). After the years of investments in content megadeals and next-gen services, it’s pure bits that count. Some mobile operators are making more money from data traffic than from content and premium services – for example in 4Q07 Telecom Italia’s mobile service revenues were made up from mobile browsing (6.1%), content and premium services (5.3%) and messaging (12%) according to data from Informa’s Mobile Communications Europe. Mobile software is also changing fast as a business; Per-unit royalties are rapidly disappearing in favour of per-activation, per-user and ad-funded revenue models. At VisionMobile we are often involved in strategy advisory projects; ‘where should I go next’ is a typical question asked by clients and as a result we ‘ve developed a model that demostrates how value creation is changing in the mobile industry. Introducing Value Quadrants Value Quadrants maps value creation and revenue model changes by asking WHEN and WHERE is value being created. The WHEN of value creation is function of the handset lifecycle, from design, development and production (pre-sales phase) to point-of-sale and in-life use (post-sales phase). There WHERE of value creation is either on the device (hardware, software, patents and industrial design) or on the services ‘cloud’ (for designing and delivering content and services). Each quadrant that is formed corresponds to a distinct type of value creation, as shown below. Starting from bottom left and moving clockwise: – On the bottom left quadrant, value is created from intellectual property (hardware, software, patents and industrial design) during handset creation, development and production – examples include Nokia, Android, Flash, IDEO, Qualcomm and Cibenix. – On the top left quadrant value is created through tools licencing, i.e. tools for building and managing software, content, UIs and services – examples include Adobe, Teleca and Trolltech. – On the top right quadrant value is created from infrastructure (hardware and software) used to deliver services – i.e. to view, edit, share, buy and deliver software, content, UIs and services in general – examples of companies in this space are Vodafone, Nokia (Ovi), Google, Admob, Bango and Logica CMG. – On the bottom right quadrant value is created from monitoring and activation of on-device assets- in other words from monitoring usage of devices, software, content, UIs and the network. Examples are Carrier IQ, m:metrics and Bluestreak. Mapping revenue models with Value Quadrants Value Quadrants become more interesting when you observe how revenue models change across quadrants. Starting from bottom left and moving clockwise: – Pre-sales, on-device revenue models are typicaly per-unit royalties, per year maintenance fees and per-project fees (NREs, platform porting fees, etc) – Pre-sales, service-related revenue models are tools licensing, i.e. per-developer-seat or per-CPU, as well as per-year support contracts. – Post-sales, service-related revenues come from usage fees (per-user, per-active user, per-use, per-level), subscriptions or advertising fees (CPC or CPM). – Post-sales, on-device revenues come from activation of on-device software and from monitoring and analytics for device/network performance and usage monitoring. Understanding revenue model changes in the mobile industry Last but not least, Value Quadrants can be used to map revenue model changes in the mobile industry. The sale value of on-device intellectual property is constantly decreasing – this applies equally to handset average selling prices, hardware platforms, and software. It’s no secret that with operating margins of the order of 8-12% handset OEMs are finding it tough to survive and invest in R&D. Moreover, in the case of embedded software, the sale value is dropping, driven by 5 market forces; – Linux, which is commoditising the OS kernel – WebKit, which is commoditising browsers – Android which is commoditising Java and the whole OS space, – Mediatek, which is commoditising the business of hardware & software reference designs and – Flash, which is commoditising rich content platforms (see Adobe’s April 08 announcement for zero royalties for Flash Lite). Furthermore, industrial design firms have also been under pressure in the last 3 years, as OEMs and even ODMs (e.g. Flextronics, Lawton & Yeo) have been developing in-house ID divisions. The only type of value that is sustainable is essential patents; which is where both Nokia and Qualcomm have a stronghold. The LTE cross-licensing pact between Alcatel-Lucent, Ericsson, NEC, NextWave Wireless, Nokia, Nokia Siemens Networks and Sony Ericsson is another testament to the importance of essential patents, which can easily command 5% or more of the handset wholesale price. I would argue that essential patents are the most sustainable source of revenue for the handset industry in the foreseeable future. So where is value creation migrating to ? The answer is simple: to the remaining 3 quadrants. For example: – Adobe has been aggressively subsidising Flash Lite (2% of Adobe’s revenues) in order to drive sales of its industry-leading tools. Microsoft has been following a similar strategy for its mobile division, investing cash to its hemorrhaging Windows Mobile platform for years in order to drive Visual Studio and Office sales; same for BREW and Qualcomm’s QCT (chipset) and QTL (licensing) businesses. – Google has invested in acquiring Android and developing the OS in order to multiply the advertising inventory that will boost its post-sales ad business. Nokia’s Ovi will also rely on Trolltech’s Qt as a service substrate, in which Nokia invested over 100 million euros. Nokia’s whole mobile organisation is in fact banking on Ovi as the vehicle to transition Nokia from a manufacturing and software development business into an internet services business. – Mobile software vendors are flocking away from per-unit royalties into per-activation fees, as network operators are more willing to invest in revenue share opportunities rather than up-front licensing fees; Bluestreak is a such an example of software vendor in what we see as the rule rather than the exception in mobile software revenue models. Moreover, a new revenue stream is emerging in the form of monitoring device, network and service usage (and the subject of a VisionMobile report on Mobile Service Analytics that is due to launch soon). Value Quadrants are discussed extensively in a forthcoming VisionMobile report titled ‘Mobile Business 2.0: Opportunities for business model innovation in the mobile market’. For additional insights into the migration from data ARPU into Channel ARPU see also the article Making money on the last mile: introducing Channel ARPU. Comments welcome as always. – Andreas

[How do open source communities work? Research Director Andreas Constantinou discusses how community dynamics can be mapped and better understood]. Handset OEMs are using Linux and browser open source components – and those that don’t will unavoidably do soon. Mobile operators are considering use of open source for next-generation service delivery via WebKit. Software vendors are realising they need to figure out how to exploit the many open source projects out there and even use open source as a key component of their strategy (see Funambol, Volantis, Sun, Trolltech, Mozilla, …). However, open source is a very complex subject matter, particularly relating to the undocumented business models, the diversity of open source cultures, the complexity of open source licenses, current best practices and case studies on how are mobile industry players exploiting open source to make or save money. Mapping open source community dynamics Understanding open source communities is one of the most fascinating topics. Communities are defined by two major attributes: – the type of license: a license is primarily characterised by the strength of the copyleft obligations (how use and modification rights are passed on along with the code). – how contributions are managed and accepted: code contributions into open source project can be managed loosely, by a moderator, by members-only or even by a single commercial entity. To compare the dynamics of popular mobile open source communities, the next chart maps 11 such communities based on these two attributes – listed are the Linux kernel, GTK, Maemo, Mozilla, Eclipse, Funambol, LiMo, Qt, Java phoneME, WebKit and Android. The chart is from VisionMobile’s 360 degree workshop on mobile open source. For an updated version of this chart see the article Mapping open source into mobile: who, where and how. Two patterns emerge by observing this chart. Firstly, weak copyleft licenses are most popular in open source projects used within the mobile industry – this is in contrast to non-mobile open source projects, where around 50% of total projects use a strong copyleft (i.e. GPL) license. Secondly, open source communities are anything but the ad-hoc formed, loosely-coordinated, grassroots movements formed around altruist developers working for free. Most successful open source communities are sponsored by commercial entities, and several are controlled by single companies. WebKit is perhaps the most striking example; while WebKit browser core components are used by Nokia, Adobe, Google and Motorola under a LGPL license, Apple is effectively in control of the main source code branch, based on the gravitas of Apple’s contributions and the cost of maintaining a fork away from the tip of the tree. Perhaps an even more interesting observation is that there is no ‘secret recipe’ on what makes a successful open source community. It’s all a question of not just the license type and governance model, but the overall mix of attributes forming the open source business model, including how the bridge between commercial and community developers is structured and the relevance of the component to ‘scratching a popular ‘itch’. For clarification, the next table lists that main licenses used in open source projects, their background, their popularity, and IPR category. The table also shows how each license treats modifications and derivatives, and whether indemnities, warranties and patent grants are explicitly provided. The table is again sourced from VisionMobile’s 360 degree workshop on mobile open source. Comments welcome as always. – Andreas

We ‘ve just launched the 100 million club: the watchlist of software companies whose products have been embedded on more than 100 million mobile handsets. Despite the apparent opportunity in the one-billion-a-year handset market, there is a multitude of challenges for software vendors who want to make it big: 2+ years technology development, 6+ month OEM sales cycles, 1+ year operator sales cycles, 2+ years for time-to-royalties, diverse OEM and operator requirements across tiers, contrasting regional mobile economics and lack of a common technology substrate. [update: the latest edition of the 100 million club is here] As a result very few software companies have managed to overcome the commercial and technical challenges inherent in the mobile industry and reach significant market penetration. (click for download access) Just over 20 products have shipped on more that 100 million handsets (incl. feature phones) as of end 2007 : Adobe Flash Lite, Aplix Jblend, Esmertec Jbed, ACCESS Netfront, Openwave Browser, Opera Mobile, Picsel Browser, Ikivo SVG Player, Scalado CAPS, TAT Kastor, Beatnik MobileBAE, Nuance Vsuite, PacketVideo CORE, Red Bend vCurrent, Symbian OS, Nokia S60, Qualcomm BREW, Mentor Graphics Nucleus, ENEA OSE, Nuance T9 and Zi eZiText. These vendors are listed in the chart segmented into five categories, based on where their products sit alongside the software stack: – Application execution environments such as Flash Lite and Java – Browsers, such as Openwave, Opera and Picsel. – Middleware such as audio and video codecs (Beatnik and PacketVideo), speech recognition (Nuance), graphics engines (Ikivo, Scalado and TAT) and firmware update agents (Red Bend). – Operating systems such as Symbian, S60, BREW, Nucleus and OSE. – Text-input engines such as Nuance and Zi. We have excluded the KVM virtual machine for which Sun did not provide any shipment numbers and Teleca’s Obigo browser which has been discontinued since May 2007. Based on Morten Grauballe’s original article on which this watchlist has been based, there are several common traits for members of the 100 million unit club. 1. All software solutions have been embedded on the handset pre-load, i.e. before the entire software stack is loaded onto the handset ROM. With the exception of Opera Mini (a downloadable Java application), we are not aware of any other downloadable software that would cross the 100 million mark in terms of downloads. In addition, downloaded applications are much harder to locate and access once on the handset due to long click-distance, which is why we have focused our 100 million watchlist on pre-loaded applications. 2. The vast majority of software vendors have relied on feature phones and the 8-12 leading proprietary operating systems that exist, rather than just smart phones and open OSes. The exceptions are Symbian OS and Nokia’s S60 which have crossed the 100 million mark. 3. The 100 million club members have mastered the complexities of software distribution. Most vendors have a direct relationship with the tier-1 handset manufacturers. Fewer vendors establish relationships with mobile operators for distribution as part of operator handset variants. In general 100 million club members understand the intricacies of the global mobile market, how to do business in operator-led vs manufacturer-led markets and the importance of leadership in standards bodies. It is worth noting that based on device volume and model data reported by vendors, we estimate that there are 2 million devices shipping per device model. This ratio depends on the accounting model for channel (incl. operator) variants – the number reduces to a 500K if one accounts for operator and regional variants. Moreover, the 2 million figure reduces as the extent of channel customisation increases, which occurs higher up the software stack due to channel customisation requirements. A note of caution: huge volumes does not mean huge revenues, as is attested by Openwave’s share price tumble, the discontinuation of Obigo browser, Zi’s negative balance sheet, the tiny royalties for RTOSes and Tegic’s valuation at the Nuance acquisition. Yet there exists a sweet spot of market penetration in-between the 100 and 500 million marks where several vendors boast both profitability and rosy revenue prospects. Warm congratulations to the vendors who have succeeded in crossing the 100 million handset mark and watch this space for the next expanded and extended revision in 6 months time! – Andreas

[Has the industry been expecting too much from data ARPU? Research Director Andreas Constantinou revisits Channel ARPU as a new way of capturing not just the user’s wallet, but also his attention and heart] It seems everyone wants to own the user these days. Or to be precise, everyone wants to own the last mile to the user, the ability to grab the user’s attention, heart and wallet. Mobile operators have been concentrating on the wallet relationship and winning an increasing share of the user’s wallet – which has led to building tens of billions’ worth of network infrastructure and a drive to increase data ARPU. Handset manufacturers have been concentrating on winning the user’s heart with the latest features, the sexiest designs and the simplest user interface. Retailers have tried to grab the user’s attention and guide the purchase decision through the physical shopping experience. Media brands have exploited the user’s brand affinity to establish further inroads into the user’s wallet. Advertisers, the newcomers to the mobile industry have focused on reaching out and attracting the user’s attention through SMS, MMS, in-game, in-video and idle screen ads. The flaws in data APRU Traditionally, the industry has been measuring these efforts in terms of data ARPU and voice ARPU; in other words the average user spend on calls, text messages and more sophisticated data services. Particularly, data ARPU has most often dominated discussions about network operator performance. As voice ARPU has been declining in both mature and emerging markets, data ARPU is seen to this date as the main revenue source for network operators. Yet it seems that the focus on data ARPU is overrated for several reasons. Firstly, data ARPU does not account for how users have multiple relationships with multiple brands besides the operator. These relationships are global, given the worldwide reach of internet brands versus the local reach of the operator. In other words data ARPU accounts for degree of wallet ownership but not wallet share, which is how user-brand relationships manifest. Secondly, the data ARPU mentality puts the emphasis on technology (e.g. advanced messaging, mobile TV and faster pipes), but not use cases (e.g. communicating to your closest friends, or living the brand experience of your favourite pop star on your phone). Data ARPU is a superficial measure that looks at the result but not the root cause of operator’s revenue and hence likely profitability. Thirdly, data ARPU describes how good the operator is at grabbing the user’s wallet, but not his attention or his heart. Yet the latter two elements are particularly important in a world of information overload, multiplicity of choices and impulse buying. Win the user’s attention, then his heart and his wallet will follow. Perhaps most importantly, data ARPU does not address monetisation through the last mile to the user. In other words, the data ARPU notion focuses on revenues derived directly from the user but sidelines a whole range of revenue sources; sources which stem from exploiting access to the user’s attention, heart and wallet. Introducing Channel ARPU Both voice and data revenues are derived by billing the end user for anything from calls and texts to mobile TV subscriptions and m-payments. Increasingly however, revenues may be derived from third parties as a fee towards the party facilitating the delivery of mobile services through the last mile to the end user. Channel revenues are not a new concept; Rental, revenue share or other fees for channelling services, products and promotions to the end user are commonplace in fast-moving consumer goods (FMCG) are well as mainstream media industries. Increasingly, mobile services can be modelled after the business models prevalent in the FMCG industries, as the margins in manufacturing decrease and the value moves towards defining the customer proposition and controlling the delivery of services through the last mile to the end user. We ‘ve coined the term Channel ARPU to refer to this new source of revenue, i.e. third parties wishing to channel services to the user. This is for a number of reasons; – firstly, Channel ARPU is derived from a new part of the pie, i.e. service providers wanting to tap into a share of the consumer’s attention, heart and wallet. Such service providers are really a new source of revenue. – secondly channel revenues are derived from attracting consumer eyeballs, gaining brand affinity and understanding consumer behaviour, activities which all take place in the last mile to the consumer (or last few inches if you consider the phone screen). – thirdly, because revenues from service providers are most interesting if seen from a per-user context, particularly as per-user metrics relate best to existing network operator performance metrics. The many forms of Channel ARPU Channel ARPU can arise from a number of sources; we have categorised these into a number of distinct types, based on the medium through which the user is reached. UI inventory leasing: the ability to channel promotions and services throughout the handset user interface (aka real-estate), across the user journey. This is about promoting content, infomercials and ads not only in-SMS, in-video, in-game and on-idle-screen, but throughout every application on the handset, from the start-up screen through to dialler, the inbox and the main menu. As an example consider Android – a means for Google to massively increase its ad inventory and derive Channel ARPU from advertisers and service providers utilising Google services on top of the operating system. The Blyk MVNO is another good example here, where Channel ARPU is used to subsidise free minute and text bundles and offset customer acquisition costs. In Blyk’s case, the MVNO leases UI inventory in the form of in-MMS ad messages. Retail sub-leasing: leasing of shelf space to OEMs and service providers through the physical retail shelf space. Today this is often seen as a handset promotion on a stand; however this model hasn’t really been exploited as much as it has been in other FMCG businesses such as supermarket shelf leasing. Yet there are plenty of opportunities for leasing shelf space for promoting services using product-like visual clues (i.e. physical boxes with the branding and conditions of use of the service) – much like what you have in a video club. Service delivery leasing. Network operators have traditionally made a cut from premium SMS service providers and handset application developers – as much as 70% in many mature markets. These revenues can be accounted for as Channel ARPU, i.e. per-user revenues derived from provided service providers with access to the last mile to the user. More recently, operators have revenue share deals with handset OEMs, by allowing OEMs to deliver services to the end-user through their network. The example here is the deals that Vodafone, TIM and Telefonica have established with Nokia’s Ovi. These deals involve bundling of Ovi services on a Nokia handset in return for a share of Ovi revenues with the operator. Channel ARPU here is the revenues operators derive from Ovi. User analytics leasing. Increasingly, efforts like Nokia’s 360 and vendors CarrierIQ, Agilent, m:metrics, mFormation and Nielsen Mobile use a handset software agent to derive a broad range of information on device usage, service usage, user profile, network usage, user social graphs and more. This is an underhyped area of mobile solutions which deliver unprecedented insight into mobile consumer behaviour; an opportunity which may ultimately result in creating a metrics aggregator that is to the mobile industry what Nielsen is to TV. As in every industry, metrics on sales, performance and analytics on consumer behaviour can be a very lucrative business. Channel ARPU therefore includes revenues which operators can source by leasing metrics and analytics to third parties. Customer access leasing. Apple’s iPhone can command a revenue share for providing network operators with access to high-spending customers. For the operator, this can be accounted for as a negative Channel ARPU, as it flows outbound towards the OEM. In a sense, this is a similar business model to the revenue sharing agreements that retailers practice with mobile operators, for each subscriber that a retailer signs up on behalf of the operator. Adopting the Channel ARPU mentality The notion of Channel ARPU focuses on growing wallet share based on collaboration with third parties. This is contrary to the traditional view of voice and data ARPU which is about growing wallet size based on customer ownership and exclusive provision of services to the customer in the all-too-famous one-stop-shop model. Channel ARPU becomes more relevant as the mobile handset business shifts to resemble the fast moving consumer goods (FMCG) business, where manufacturing costs are low and value comes from using multiple partnerships and channels to access the right consumer at the right moment. Moreover, Channel ARPU better accounts for monetising access not only to the user’s wallet, but also to the user’s attention and brand affinities through the last mile. Channel ARPU is a refreshing and meaningful approach to viewing per-user revenues derived by leasing access to the consumer through that last mile. However, much like data ARPU, network operators have to formalise and standardise the metrics by which Channel ARPU is calculated, if this new form of measuring ARPU is to be adopted. Perhaps a task for standards bodies such as the OMTP ? – Andreas

[UI development flow and actors: Graphical Designer, Interaction Designer, Software Engineer, classical technologies: GTK, Qt, next generation: Flex, Silverlight, WPF, TAT, XUL, SVG… guest blogger Thomas Menguy describes what are the main concepts behind all the UI technologies, what the new generation ones have in common, what those modern approaches are bringing to the product development flow…and what is missing for the mobile space]. A good UI is nothing without talented graphical designers and interaction designers: How the plethora of new UI technologies are helping unleashing their creativity? What are the main concepts behind those technologies?Let’s try to find out! UI is trendy… thank you MacOS X, Vista and iPhone! UIQ S60 iPhone Put the designers in the application development driver seat! Here is a little slide about the actors involved in UI design UI flow actors and their expertize What does it mean? Different actors, different knowledge …. So different technologies and different tools! Those three roles can be clearly separated only if the UI technology allows it. This is clearly not the case in today mainstream UI technologies where the software engineer is in charge of implementing the UI and the service part, most of the time in C/C++ , based on specifications (word document, Photoshop images, sometime adobe flash prototypes), that are subject to interpretation. The technologies used by the designers have nothing in common with the one used to do the actual UI. The technologies that allow UI implementation…require an heavy engineering knowledge. Big consequence: the software engineer decides at the end! The picture is different for web technologies where it has been crucial and mandatory to keep strongly uncorrelated the service backend from its representation : Web browsers have different API and behavior, backend have to be accessed by many other way than web representation…and above all data is remote and presentation is “half local/half remote”. Separating representation, interaction and data has been the holly grail of applications and services development for years. It has been formalized through a well known pattern (or even paradigm in that case) : MVC (Model View Controller) MVC pattern / source: wikipediaFrom wikipedia: http://en.wikipedia.org/wiki/Model-view-controllerModel The domain-specific representation of the information on which the application operates. Domain logic adds meaning to raw data (e.g., calculating if today is the user’s birthday, or the totals, taxes, and shipping charges for shopping cart items).Many applications use a persistent storage mechanism (such as a database) to store data. MVC does not specifically mention the data access layer because it is understood to be underneath or encapsulated by the Model.View Renders the model into a form suitable for interaction, typically a user interface element. Multiple views can exist for a single model for different purposes.ControllerProcesses and responds to events, typically user actions, and may invoke changes on the model. All the UI technologies are offering a way to handle those 3 aspects and, as a consequence, are providing a programming model defining how information and events flow is handled through the MVC. See below a simple schema I’ve made describing a GTK application: when you look at an application screen, it is made of graphical elements like buttons, lists, images, text labels, called widgets (or controls) . Rmk: the term “widget” is used with its literal meaning : “window-gadget”, this term is now used a lot in web 2.0 marketing terminology and by Yahoo/Google/MS to represent a “mini application” that can be put on a web page or run through an engine on a desktop PC or a mobile phone, to avoid confusion I prefer the term of “control” over widget for the UI technologies, but will continue using “widget” in the rest of the GTK example as it is the term used by GTK itself. Widgets are organized hierarchically in a tree, meaning that a widget can contain other widgets, for example a list can contain images or text labels. In the example below the “root” widget is called a “Window”, it contains a kind of canvas which itself contains a status bar, a title bar, a list and a softbutton bar. Then the list contains items, the title bar has a Label, the softbutton bar contains some buttons and so on. A widget is responsible for Its own drawing using a low level rendering engine, called GDK in the GTK case (GDK offers API like draw_image, draw_text, etc). Computing its size according to its own nature (like the size of the text that will be displayed for example) and the size of its sons. Reacting to some events and emiting some specific ones: the button will emit a “press event” when it is pressed with the touchscreen or when its associated keypad key is pressed. The widget tree will propagate system events (keypad/touchscreen, etc) and internal events (redraw, size change, etc) through the widgets. The developer will register callbacks (in fact functions, piece of code implementing a functionality) that will be called when widgets will fire events (like the “press event”) . GTK Widget tree structure: a phone screen example The major GTK/gLib formalism is how those events/callback are handled: through what is called a “gloop” where all events are posted in the loop queue, dequeued one by one and “executed” in this loop, meaning their associated user callbacks will be called. This loop is running in one thread. This is what we call a programming model. In nearly all the UI technologies such a loop exists with various formalisms for the queue handling, event representation, etc. To finish with the above schema the user callback will then access to the middleware services, the various databases and so on. There is no clear MVC formalism in that case, the controller is mixed with the view …and even the model that is mixed … with the widgets! (so with the view) Qt Model is really identical to the this one. One last point very relevant for application development and design: the notion of states. Each application is in fact a state machine displaying screens linked by transitions, like in the example below where in the state 1 the user decides to write an SMS, it will open an SMS editor screen and clicking send will go to a selection of phone numbers. Application State Machine: write sms example Here is an attempt to formalize a modern UI framework with Data binding (for Model abstraction). UI engines formalizationControl:equivalent to a widget but where the MVC model is fully split. A Data Model as to be associated alongside with a Renderer to make it usable.Control Tree:equivalent to the widget tree: aggregation of Controls, association of the controls with a Renderer and a Data Model. Possibly specification of Event Handlers.Data Model:Object defining (and containing when instantiated) a set of strongly defined and typed data that can be associated with a Control instance.Data Binding:Service used to populate a Data Model.Control Renderer:Object that is able to graphically represent a Control associated with a Data Model, using services from a Rendering Library.Rendering Library:Set of graphical primitives, animations, etc.Event Handling (and Event Handler):code (any language) reacting to events and modifying the current state machine, the Control Tree, etc.Standardized Services:Interfaces defined to access middleware directly from the event handling code.Server Abstraction:Possibility to transparently use Data Binding or any service call locally or remotely. Ok if you are still there, and your brain is still functional, here is what’s happening today in this area…. In traditional UI frameworks like GTK, Qt, win32, etc the control tree description is done with a C/C++ description … a little niche technology have paved another way: it is called HTML: after all an HTML web page description is defining a tree of controls, W3C use a pedantic term for it : the DOM tree. JavaScript callbacks are then attached to those widget to allow user interaction. It is why all the new UI technologies are based on an XML description for this tree, it is muuuuuch more easier to use, and allow a quicker description of the controls, and above all it allows nice design tools to manipulate the UI….Apart from this XML representation the majority of the UI technologies are coming with: An animation model, allowing smooth transitions, popularized by the iphone UI, but it was already there in MXML (Adobe Flex Format), XAML (MS format), SVG, TAT offer…. Modern rendering engines (Flash for Flex, MS has one, TAT Kastor). Nice UI tools for quick implementation: Adobe Flex Builder, MS Expression line, TAT Cascades, Digital Airways Kide, Ikivo SVG… In many case : a runtime, to be able to run a scripting language. Here are some quick tables, really not complete, of some of the most relevant UI technologies in the PC and mobile phone space. Just to explain the columns: RIA : Rich Internet Application, delivered through a browser plugin RDA : Rich Desktop Application: delivered through a desktop runtime Runtime: ok, galvoded name here, just a name to represent the piece of technology that allows the UI to run UI: Technology to describe the control tree (you know what it means now!) Event Handling: the dynamic UI part, and how to code it (which languages) Tools: UI tools RIA&RDA Chart Embedded rich UI technologies So its time to answer the main point of this post: How those technologies are helping unleashing designers creativity? By defining a new development flow, allowing each actors to have a different role. Here is an Adobe Flex “standard development flow: Adobe Flex&Air tools flow In the next schema I try to depict a more complete Adobe Flex flow, adapted to the mobile world, where, for me, a central piece is missing today: it is not possible now to expand “natively” the adobe air engine, this is mandatory for mobile platform with very specific hardware, middleware, form factors. So I take the adobe flow more as an example to demonstrate how it should work, than as the paradigm of the best UI flow because this is not the case today (same remarks for MS flow, less true for TAT for example) An Adobe UI design flow for embedded This shows clearly that the “creativity” phases are clearly uncorrelated: different tools are used between the designers, they can do a lot of iterations together, without any need of the software engineer. This one can focus on implementing the services needed by the UI, optimizing its platform, adding middleware features. Interaction Designer defines the application high level views and rough flow Graphical Designer “draws those first screens” Interaction Designer import it through Thermo Graphical Designer designs all the Application graphical Assets Interaction Designer rationalizes and formalize what kind of data, events and high level services the application needs Interaction Designer & Software Engineer are working together on the above aspects HINT: A FORMALIZM IS MISSING HERE once done: Software Engineer prepares all the event, data services, test it unitarily, in brief: prepare the native platform Interaction Designer continues working on the application flows and events, trying new stuffs, experimenting with the Graphic Designer based on the formalism agreed with the Software Engineer. Once done … application is delivered to the Software Engineer that will perform target integration, optimization…and perhaps (hum certainly) some round trip with the other actors 🙂 So this is it! this flows really focus on giving power to the designers…taking it from the engineer hands. Some technologies are also missing to really offer a full Mobile Phone solution: All the PC technologies are about building ONE application and not a whole system with strong interaction between application. With today technologies, the designers are missing this part….leaving it to the engineer: How to cleanly do animation between applications? Strong theming and customization needed for: product variant management: operator variants, product variants (with different screen sizes and button layouts for example), language variant (in many phones 60 languages has to be supported, but in separated language packs). A not well known one: Factory line fast flashing of those variants. It is very long to flash the whole software of a mobile phone while on the factory line … so if you are able to have a big common part and a “customization” part as little as possible but with the full UI…you gain productivity…and big money 🙂 Adapted preset of widget or controls (try to do a phone with WPF or Flex…all the mandatory widgets are missing) Anyway an UI technology is only the way to interact with a user…to offer him a service. Most of the technologies presented above are about service delivery and not only UI…My next post will be about this notion of service delivery platforms. [Update] Replaced the term “ergonomics specialist” by “Interaction Designer”, thanks Barbara, see comments below. Thomas Menguy #mxml #svg #gtk #xaml #adobeflex #mobilephone #userinterface #tat #daw #ui #thermo #phone

[All consumer electronics, mobile and software companies have to in-source third party software – but the perils and complexities of managing that software are several, as the software moves through your organisation and out to the customer. Guest blogger Ã…se Stiller distills years of experience in software licensing in simple guidelines for managing 3rd party software] The particular challenge I would like to draw your attention to in this article is the risk for unauthorized distribution or reuse of the third software. It is easy to see the need to assess a potential supplier, but it is equally important to turn around and take good look at your own company’s ability to safely manage the responsibility, End-To-End Whether you are developing a 2D game or a complete application framework putting together software products is not always as “Lego-like” an activity as we like to draw in block diagrams to management. In many cases bringing in 3rd party source code is the only viable alternative, but you must be sure that everyone dealing with the code knows what they are doing.- and you can´t very well ask all engineers to study the contract. Pardon my French but it’s often hard to ask engineers to “RTFM” or to be precise “RTFC” (C for contract). It´s is hard enough to get the right code, at the right price, in the right shape delivered at the right time but you must also get it with the right terms to fit well into your own development process, not to have to burden developers with unnecessary restrictions and contract details. Sourcing of software is and must be a team activity. Only a true Hercules can deal with all the Hydra-heads single handed. Staying above the line of software commoditisation Developing an application, game, application framework or any piece of software for mobile phones means adding unique and sexy features to a whole lot of commodity, and checking that nobody else gets there first. What was cutting edge technology two years ago is now all taken for granted – users don’t want to pay for Bluetooth or FM radio these days and 5 mega pixel cameras is slowly becoming the norm. The value line of software is increasing up the stack with each month going by. For software companies, as new technology ceases to be a marketing differentiator and turns into standardized commodities, it becomes appropriate to share the costs for maintenance and further enhancements with competitors and partners in the same business. The easiest way to find this economy of scales is of course through sourcing software from Independent Software Vendors (ISVs) or by using software which comes under an open source license (e.g. Eclipse IDE and WebKit browser core). Another good reason to source software is of course to get access to specific technology that you cannot, or cannot afford, to develop in-house. Some things are simply best done by experts, like preparing Japanese puffer fish and developing e.g. telephony modules for mobile phones. In mobile phones as well as all other complex consumer electronics you will find a lot of common functionality developed and maintained by external ISVs. The type of components vary from highly visible functions like Web-browsers to completely anonymous drivers, and only very few suppliers will get their logo in a prominent place. A mobile phone would have to be the size of a car to allow for “NN-inside” stickers for all externally developed software components. There may be many links in the chain between the original developer of a software component and end-user. A lot of the sourced components themselves come with software developed by others than the supplier, like e.g. open source parsers and specific security solutions, and the further away from the original owner of the IP the harder to keep track. A good practice is to make sure that such components are declared well in advance before you sign an agreement to source software, specifically if there are inherited restrictions or obligations pushed on to you and your customers. Make sure that the sourced software doesn´t come with an undetected Hydra-head embedded in the license. If there is Open Source with obligations to disclose code to end-users this will affect you or your customer, or your customer’s customer. The perilous journey of software through an organisation Regardless of the reason for sourcing 3rd party software it is important to truly understand how the sourced component will be used in the internal development process, and how the component will be integrated with existing product. For example, there may be a need to distribute code to sub-contractors or pilot users, customers, additional development sites etc, during the development process . Your license must meet these needs and allow for that – or you must plan for an alternative way to work. Deviations from standard development process may add to the overall cost for the sourced component, and must be taken into the cost/benefit analysis. Naturally you must not panic but maintain a healthy balance between the added cost to eliminate a risk and the weighted cost for an actual breach. If the sourced component is self contained and easy to replace your worries are less, even with a un-permissive license. However with a less modular and more realistic dependency you need to be more cautious about handling the code. A good way to reach an understanding of all your needs for re-distribution of a sourced component is by identifying the End-To-End journey for the software through your company. You will then be able to anticipate what distribution rights must be catered for in the contract, and to prepare for a change in the process if you cannot win these rights in the license. Another benefit of such practice is that you will gain a better understanding of how the third party product is to be integrated with you product. You can describe the level of integration and understand how that will impact your negotiation, or choice of license for Open Source software. The natural place for an end-to-end flowchart and for the description of the integration is in a Risk Analysis for the sourced component. Such Risk analysis should also include the standard commercial risks, market risks, legal risks etc – but that is a different subject. Plan for the software’s journey There is no single solution to how to safely manage third party code through your development process; any solution must be case specific. Naturally it depends on the size of your organization and of how many will have access to the code, but it also depends on the level of integration with the rest of the product. Based on the End-To-End flowchart and the license for the software, you can provide a plan for the management of the third party code; you can identify an owner of the code in all stages of development, and if necessary specify a hand over process between development units, and eventually the hand over to your customers. You can define specific rules for the code and provide solutions before the issue becomes a problem. As an extra bonus you may also find and be able to eliminate conflicts between licenses – e.g. an Open Source license forcing disclosure of code and another license prohibiting such disclosure. Another Hydra-head down. Information is King A development process is only as waterproof for the in-sourced 3rd party software as the team members who are involved in this process. By identifying the different units that will be involved with the third party component, you will be able to locate who needs to be informed about the rights and obligations that come with the component. Knowledge about the restrictions for the component must be available for all those who actually access the code, all the time. A good way to achieve this is to keep a simple database where you register the third party IP included in the product, and specify rules for how the code may be spread and used. Access to the information in the database must be granted to all those who come in contact with the code and therefore you may not want to add the contract as such to the database. Do not confuse this with a contracts database; that serves a different purpose. Information in this database must be easy to find, easy to understand and interpret and it shall clearly explain what you must and must not do with the code. E.g. if you can only distribute to subcontractors or development sites that have been approved in the contract, these subcontractors and sites shall be listed in the database and if you can only distribute binaries to customers that must also be easy to detect for anyone who might need to release code to customers. It is, of course, crucial that the information in the database is always correct and updated. Therefore someone must be identified as the owner for the information. The owner of the contract is my first choice since that is the person with the most to gain on this; the one who will be hit by the Hydra if something goes wrong or have a smooth and easy day at the office when all the questions are answered by the database. Do’s and Don’t of software sourcing. When you accept the responsibility and liability that comes with signing a software license for source code – Open Source or closed Source, you must not let price and liability totally overshadow the rights to use and distribute. Do your homework and get a good understanding of the intended use of the code, before agreeing the license terms. Analyze the risk and avoid problems further down the road by planning the management of the software End-To-End and publish the restrictions for the code where it is easy to find and to update. – Get the team to help define the contractual needs for the sourced software. My experience is that insufficient or unclear license rights too often blocks engineers from doing their job. If you can get the engineering teams to help you prepare the sourcing better by providing relevant input to how your company will use the sourced component, you stand a far better chance getting the contract right from start, and it will save you from panic changes to both contracts and project plans. – Educate all on a need to know basis. My experience also tells me that engineers in general don’t love contracts, or rules restricting their creativity, but they do need to know the “Dos and Don´ts” with the third party code they handle. Just putting rules in a database is probably not sufficient; a verbal run-through of the rules is a good support for the information in the database, and an excellent way to test if the database works for its intended audience. You will probably also learn if the rules set up for the third party IP is too restrictive or too complicated. – Talk to all stakeholders. Sourcing of software is teamwork involving every department in your company – save possibly for janitors. It takes a team to find and fight all the heads of the Hydra. In other words; to provide all the input for the risk analysis leading you to a good license agreement and a safe passage for the software through your development process, you need help from all the stakeholders. Open Source is no different from any other third party software in this respect. – Ã…se Stiller [Ã…se has lived through the pains of licensing software both from the selling and the buying side of the cooperation as part of UIQ and Teleca, and has survived to tell the tale and educate the rest of the industry.]

[Is there a future for the SIM card in operator service delivery? Research Director Andreas Constantinou reviews the state of the SIM card industry, the commercial developments in the last 12 months and discusses why the role of the SIM may be indeed coming to a positive inflection point] Yet a paradox exists in the adoption of the SIM card for operator services. On one hand tier-2 and tier-3 operators have put the SIM card into innovative uses: the SIM has been used to deliver mobile banking in Czech Republic, mobile ads in Russia, ringtone downloads in Brazil, payphone use in Nigeria and automatic device detection in Austria. On the other hand tier-1 operators have used the SIM mostly for basic applications such as managing missed calls and roaming lists; At the same time handsets have developed far superior user interfaces to the SIM’s text-based UI and operators have invested in Java, Flash and on-device portals for advanced service delivery, as opposed to SIM based applications. It’s perhaps yet another reminder that innovation does not easily bubble up in large organisations like tier-1 operators, while cash-strapped tier-2 and tier-3 operators have been more resourceful and innovated using existing infrastructure. Still the issue of adoption of SIM cards for service delivery by tier-1 operators is indeed a fundamental one, given that tier-1 operators are the largest SIM customers. Furthermore with the commoditisation of SIM card functionality, SIM card manufacturers will need to continue delivering new value in order for the whole SIM ecosystem to survive – and a win-lose situation is not tenable. So is there a bright future for the evolution of the SIM card? The SIM industry backstage To understand the status quo, one needs to visit the backstages of the SIM card industry – an all-too-familiar site, for those observing the industry in the last few years. Operators (particularly tier-1s) have been clearly motivated to see SIM cards take a greater stake at service delivery, yet have been reluctant to invest in long-term initiatives without a business case for a 6-month RoI. As such, tier-1 operators have reverted to applications (e.g. on-device portals, active idle screens), platforms (Java, Flash Lite) or ‘container’ programs to further their service delivery aims. On the other hand, handset OEMs have until recently seen the SIM card evolution as a potential compromise to their own agenda and have been slow at adopting industry standards for SIM-enabled service delivery. And while SIM toolkit standards have been adopted in the vast majority of handsets (an estimated penetration of 95% of more), the potential of the SIM as a service enabler has been severely limited compared to the constantly increasing handset feature arsenal. Last but not least, SIM card manufacturers have since 2006 proposed significant technology advancements, in terms of ‘smarter’ SIM software, near-gigabyte capacity and creative new applications, from blogging and widgets to advertising and idle-screen promotions. Yet these advanced SIM cards have always demanded a significant per-unit price, while the average selling price of ordinary SIM cards has been dropping as much as 30% year-on-year during 2006. Yet the prospects for the SIM card evolution in early 2008 are not as dire as they may seem. A confluence of developments, both technology and commercial ones are marking an inflexion point for the advancement of the SIM card. Change of scenes Three major developments that took place in the last twelve months will likely impact the uptake of advanced SIM cards in 2008: 1. The price war on SIM cards has largely subsided; during 2006 competition from China and between the major SIM card manufacturers caused prices to drop by more than 30% year-on-year. Fortunately, the prices have now stabilised with Gemalto reporting a decline in average selling price ASP) of only 2% year-on-year for 4Q07 (compare this to declines of handset ASP of 5% or more for tier-1 OEMs in 2007). 2. The cost of NAND memory used in SIM cards (as well as PCs and removable storage media) has dropped dramatically in the last two years. The cost delta between a 256KB NOR memory and a 256MB NAND memory has dropped by an order of magnitude within the space of two years. This has helped make mega-SIM cards more affordable to mobile operators who are planning to source high capacity SIM cards. 3. The OMA standards body has been busy finalising the Smart Card Web Server (SCWS) specification, a technology for using the SIM card as an always-on web server that stores operator content, application settings and encrypted files for multiple applications. The SCWS specification is expected to be finalised soon, following three successful interoperability ‘testfests’ which took place between September 2007 and January 2008. The SCWS is a pragmatic specification that leverages the mature and widely used HTTP protocol to enable a range of solutions such as on-SIM portals, NFC, just-in-time customisation and DRM. The SCWS protocol requires a more advanced smart card OS, but no hardware upgrade and hence no increase in hardware BOM. This is in contract to high-capacity SIM cards, which impact not only the silicon BOM, but also add a requirement for two extra PINs to both the SIM cards and the reader terminal. We are not there yet.. Despite the recent developments, and the great potential to address the application distribution barrier, the role of the SIM card has not really advanced beyond that of an authentication mechanism, particularly for tier-1 operators who command scale and dominate OEM terminal requirements. A few tier-2 and tier-3 operators in Latin America and Europe have been using the SIM card in applications such as banking, automatic device detection, ringtone download and idle-screen promotions. Yet tier-1 operators still appear reticent and somewhat undecided as to whether to invest in advanced SIM cards with SCWS capability and/or high capacity. The primary reason has been pricing. SIM card OEMs have been bundling advanced OS capabilities in higher-end NOR cards of the 256KB and 512KB range, which operators don’t yet perceive the need for. The major handset OEMs have delayed plans to incorporate SCWS (and the necessary underlying BIP server support) due to the lack of an established base of SIM cards that support this functionality. Sagem and LG have committed to supporting SCWS within selected commercial handsets later in 2008, but there are no commitments as to the scale and the sustainability of OEM SCWS adoption. Overall, the industry has been caught up in a chicken and egg situation where no player has been willing to risk investment in advanced SIM cards. How to kickstart the system Pricing and addressable market are the fundamental criteria that have caused the Ferris wheel of the SIM card evolution to remain still. Yet, there are still ways to kickstart the wheel and push the industry inertia into motion. For that to happen, SIM card manufacturers need to redraw their pricing plans and figure out how to migrate the software BOM surcharge into service enablement post-sales revenues. Once SCWS capability is featured on standard SIM cards as a norm, the handset OEMs will be incentivised to support the many SCWS use cases and therefore produce compliant handsets via a software update. And once the industry Ferris wheel starts spinning into motion, the benefits for both handset OEMs and network operators will be compelling enough for the wheel to keep spinning for many years.. at least until the next S-curve arrives. – Andreas [Andreas is a moderator at the forthcoming SIMposium conference in Berlin, 22-23 April, the annual SIM mega-event for the mobile industry.]

[In part 2 of his predictions for MWC 2009, Research Director Andreas Constantinou talks about M&As amongst Linux vendors, OHA devices, enterprise UIs, the challenges for Modu, and the unstable future of UIQ] Check part 1 for learnings from MWC and predictions on a new Trolltech, the evolution of widget solutions, the relicensing of Qt, acquisitions of WebKit vendors and Danger devices. Prediction 6: M&As in the Linux vendor landscape Analysis: Mobile Linux has gone through two phases; the first phase (2000-2006) was the OEM in-house efforts from Motorola, NEC and Panasonic who developed their own middleware and applications on top of MontaVista (and Qt/E in the case of Motorola). The second phase (2004-now) has been the emergence of for-license Linux-based software stacks from MontaVista (the incumbent), WindRiver, OpenMoko, Mizi, Access ALP, Azingo, A la Mobile and Purple Labs. Many of these vendors also offer integration, customisation, productisation and certification services on top of their software stack, as shown in the next diagram. (note that Qt/Qtopia are missing from this chart because it is still not known whether they will be offered under commercial license terms following the Nokia acquisition). In practice the above taxonomy of mobile Linux vendors is rather simplified and the devil is in the details; OpenMoko is six months late and probably six more months before being mature enough for v1; Mizi has recently announced a re-developed version of a low cost stack and is looking for customers beyond Korea; Access ALP still has teething problems and its MWC demo was unimpressive; Azingo is well funded and has a quite stable & complete stack incl. WebKit, but has only started to build a services arm; A la Mobile is underfunded for its claim as the ‘Red Hat of Mobile’, while WindRiver appears to be taking on that role; and finaly Purple Labs is already on three European phones and has a single-core Linux stack ready for licensing. Naturally, ten mobile Linux vendors is far too many, while financial challenges will be setting in this year and OEMs will be making hard decisions about which stack/integration vendor to choose. MontaVista had a surprisingly small stand at MWC this year, while it has been losing many head-to-head bids to WindRiver. A la Mobile has publicly only been funded with $3million, a far cry from the $30million that Azingo (ex Celunite) has gotten to date. It is likely that MontaVista or A la Mobile would be looking for another financing round or for an exit. [update: A la Mobile secured a second round of 6.75 million from Venrock in February. That’s enough to power a startup for 2 years, but is it enough to build a services organisation?] On the other hand Purple Labs has been under the radar until recently when its ownership structure changed with the majority ownership moving from Vitelcom (Spanish ODM) to Sofinnova Ventures (altough details of the deal are sketchy). Interestingly, Purple Labs has a mature software stack already in three European Linux-based phones and claims to have the first single-core Linux stack already on a soon-to-be-commercial phone. Yet despite the strength of its technology (and the hardware design expertise of its team) PurpleLabs is lacking the professional services arm that will aid OEMs in integration and productisation projects (any takers out there?). [update: PurpleLabs has a proportionately-sized pre-sales team, but most importantly has a strong management team incl. the ex-head of Openwave prof. services]. Exits, IPR acquisitions and company acquisitions are therefore likely for mobile Linux vendors by MWC 2009. Prediction 7: OHA devices; cheap but ugly Analysis: The Open Handset Alliance and master-chef Google have been cooking the Android SDK for quite some time and the first development boards were shown running Android ‘officially’ at MWC. There are no counter-indications that Google will be able to hit its 2H08 promise for the first Android devices; several chipset vendors have been integrating the Android stack and HTC (followed by Samsung) has significant expertise in bringing up a ‘virgin’ OS into a mature phone software, as it did for Windows Mobile in 2002-4. The well-architected stack that is Android (incl. plug-and-play core apps and J2SE-like environment) will likely be targeting mass-market devices; having a US-based company as master-chef and HTC as the host, this probably means low-BOM devices with a PDA-like form factor. In other words, a low price and data-first design will be very much a priority compared to the looks and phone-first design (very much like Windows Mobile devices thus far). Prediction 8: Enterprise UIs emerging Analysis: The enteprise segment has traditionally been seen as completely contrary to consumer segment from a functional requirements perspective; consumer devices have to be fun and sexy, whereas enterprise devices have to be function-first and stripped down of most aesthetic features. But wait.. who said enterprise people are boring? I believe that some innovation on the user interface and the plastics of enterprise devices is in order. And while plastics innovation is too much of a gamble, UI innovation isn’t (you can change UIs easily with many of today’s UI frameworks). Therefore, I foresee that at least one vendor will be offering enterprise-targeted UI frameworks which provide both eye-candy and functionality such as word/Excel/PPT/PDF document viewing and rich email (Picsel comes to mind). This also means that the boundaries between enterprise-targeted mobile OSes and consumer-targeted OSes will be bluring, which is also the direction taken by Windows Mobile 7 featuring a customisable UI layer (a major functional delta from all previous Windows Mobile versions). Prediction 9: The challenges of Modu Analysis: Modu made big headlines at MWC, not only because of its huge marketing spend, but also the innovative nature of its connected device offering. Modu offers a mobile (cellular) building block which is at the center of a connected personal area network of mobile devices. Like many previous attempts at creating a distributed devices environment (most notably IXI), Modu is based on moving cellular connectivity into the centre of a connected devices framework and thereby making it much easier to design, develop and market mobile devices. In principle, the paradigm of a distributed devices environment is a win-win-win for operators, manufacturers and users (as I advocated on this 2002 IEEE paper on this very subject). However, the challenge is in bootstrapping the ecosystem of operators, device and services vendors to invest in this new paradigm of building connected devices. What Modu has done is quite clever; Modu did not develop just an OS for powering this connected device ecosystem (like IXI) or just the connected devices prototypes (see Motorola’s wearables distributed devices collection designed by Frog agency). Modu created the physical building block (a nano-phone, so to speak) that can form the nucleus of the distributed devices system, making it easier to bootstrap an ecosystem around it. However, Modu is still facing a major challenge; convincing OEMs to build devices around its building block. That means putting its money where its mouth is and funding (or at least part-funding) several handset projects, which is clearly a very expensive exercise. More importantly today’s handset OEMs are more keen to invest in services rather than over-innovative handset designs. To convince OEMs to build on top of its building blocks, Modu therefore has to create a framework for delivering services around it. There are a handful of service companies who are today creating service frameworks; Google, Yahoo and Nokia come to mind. Service frameworks are inherently a loss leader; there’s no money to be made by designing, developing and supporting the service connectivity framework (see Android, widgets/Go 3.0/OneConnect and Qt, respectively). However there is money to be made from enabling service delivery and access (see Google ads, Yahoo ads/services and Ovi, respectively). Therefore Modu’s challenge is in creating a loss-leader framework for delivering connected services around its cellular building block and convincing OEMs that this should form part of their service investment strategies. Prediction 10: The unstable future of UIQ Analysis: Motorola invested in a 50:50 ownership of UIQ alongside Motorola back in October, in what amounted to a diversion for the company’s Linux strategy. UIQ’s expansion (now 400+ people in Ronneby and Budapest – almost a tripling in numbers within a year) means that the venture has much higher costs than revenues. If you do the numbers, it turns out that UIQ must ship at least 6 million devices annually (at $3 per-unit royalty) in order to sustain its workforce OPEX. This is a far cry from the 1.2 million it shipped in 2006 but close to the 7.7 million estimated for 2008 (both figures from Nomura). UIQ must therefore ramp up volumes very fast in order to sustain its OPEX. The real challenge with UIQ however comes with sustaining the underlying Symbian OS strategy. With UIQ’s ownership transfered out of the Symbian, Sony Ericsson and Motorola are now arch-rivals for Nokia, who controls the majority of Symbian shares (and in practice most of the decisions taken by the Symbian board). It has for long been rumoured that Nokia has been working on a new-generation OS, but rumours aside, UIQ should have continual challenges in driving the features and strategic agenda for Symbian OS towards favouring UIQ. Moving UIQ to a different kernel support package (some Linux flavour) is a very expensive 2-year operation that UIQ would not easily venture into, given its financial state and the instability of its parent Motorola. Clearly a lot to look forward to until Mobile World Congress 2009.. – Andreas (while on the topic of predictions, make sure to check out our hugely successful Mobile Megatrends 2008 series. Full presentation below.) [slideshare id=209579&doc=mobile-megatrends-2008-vision-mobile-1198237688220186-3&w=425]

We ‘ve just launched our new website with a completely redesigned look & feel!. After a year in the making and working closely with two design agencies, the website is finally alive and kicking. Thanks to Paul at fifty50 and Savvas at Peel-Me for all their hard work. Feel free to browse through the site and let us know what you think! San Francisco next week? Continuing with our successful 360 degree workshop on Mobile Open Source, we ‘ll be delivering the workshop as part of Informa’s Open Source in Mobile conference in San Francisco conference on March 10. This one-day intensive workshop is a must for companies wanting to understand the economics, legal issues and complex landscape of Linux and open source software vendors in the mobile industry, and make informed decisions on their own positioning. Check here for more info on the Informa Open Source in Mobile conference, or drop us a line if you are in the San Francisco area and want to meet up. – Andreas