The first part of this article gave an overview of Google Inc. (GOOG) and how its business model affects its objectives in the device platform space. In that part of the article, Google was described as competing for the volume market by supplying low-cost operating system software under a strategy of deriving profits after the point of sale by directing device buyers into Google-directed revenue streams. Google's benefit in the platform war is thus driven by use, not (generally) device purchase. This article compares that plan with competitors' plans and explains what this means for Google in the platform war.
Competitors. What's a look at a market worth, without a look at competitors and the threats they pose? Google's competitors aren't all direct competitors, but there are several competing platforms.
Apple. Apple (AAPL) is the largest platform vendor by profit share, and operates the application store that does the most sales. As the reseller of all the applications on Apple's App Store, Apple keeps 30% of all sales - including little 99¢ sales whose transaction-processing overhead consumes most of that fee. But with application sales on the order of $4 Billion anually, there's a lot of post-sales revenue to share. Post-launch growth rate of app revenue dramatically exceeds music sales growth following the launch of Apple's music store. However, Apple's much more mature music store has not yet been matched in revenue by the App Store. (The Apple Music Store hasn't been thought of as a significant source of free content, but its podcast interface offers significant fee-free content similarly to the App Store.) And Apple's Music Store is still growing geographically. Some have theorized that Apple's revenue from the online store would eclipse its device sales, making it more like Google in reaping funds from use. However, Apple's hardware margins at present so dramatically exceed post-sales revenue per user account as to complicate "share" comparisons with Google; nevertheless, Apple is building a post-sales revenue model with which Google is in direct competition. So far, Apple seems to be getting the better of that competition: despite Google's platform outnumbering iOS units in the world and in new device sales, customers spend more buying iOS applications.
Google doesn't lose when Apple sells a unit, though: Google makes money on Apple's customers. Google chose not to allow Apple to access Google's APIs for producing turn-by-turn directions, so Apple replaced Google's maps service with its own maps. (To initially entertaining results: bring SCUBA gear on your search for the train station.) But let's face it: was Google making much money on maps? (How much do people really pay Google to put their store names or logos on maps drawn for people looking for some other destination?) Is the Maps app really a loss for Google on iOS? The larger risk is that Apple might direct users' search activity to a different source, thus eliminating Google's ability to reach those users with revenue-generating ads. Since Apple's customers are valuable ad customers - iOS users are disproportionately online and therefore able to see ads - Google may not want to lose contact with those customers.
Toward that end, Google offers an iOS app that connects users with voice-activated search. Wil Shipley, author of Monster Library and very longtime OS X developer (i.e., since before it was an Apple product), gave this review of Google's voice search on iOS:
Google and Apple are not competing for the same thing: Apple wants to sell units (and needs turn-by-turn directions in its maps app to keep competitive), and Google needs to have search users to hit with ads (which it can do on anyone's platform, though it knows many users don't change defaults and will use pre-loaded search tools). Google is reportedly bringing turn-by-turn directions to iOS through its own app, and Apple continues to can those who didn't stand behind Apple's maps and is recruiting experts to improve its maps.
But enough about consumers and their demands: what about big business, where Apple has enjoyed some real growth lately? Although Apple's earliest marketing effort focused on enterprise, it did not initially succeed in capturing business sales, and the majority of its sales in recent years have been dominated by consumer sales. Despite Apple's historic disinterest in enterprise, it has lately enjoyed accelerating acceptance in enterprise. However, as Apple has done little to support enterprise and nothing to offer off-migration paths from existing enterprise dependencies on non-Apple administrative technologies, it's unclear whether Apple's enterprise performance will survive attack by a new crop of products specifically targeting the enterprise market.
Apple's real strength in mobile devices is illustrated by its premium customer base. It parallels Apple's position in the PC market: Apple's not dominant by unit sales (a 15-year high left Apple with single-digit global share in PCs, unless iOS devices are included), but Apple's share is such a high-margin, desirable segment of the market that the company is dominant by profit despite its modest share. In mobile, iOS global unit sales share has been dwarfed by Android's; yet the platform use share numbers suggest Apple's customers are dramatically heavier Internet users, and thus in much more frequent contact on a per-user basis with drivers of use-based revenue. This is more immediately interesting to Google (whose current revenue depends on use) than to Apple (which profits nicely on unit sales and hasn't yet elevated use-based profit to a major segment). But the characteristics of Apple's users has real value to both Apple and Google over the long run. If Apple's segment of the market is a high-use segment, driving more use-based revenue, developing use-based revenue could secure for Apple some post-sales use-driven revenue to protect against future margins compressions - revenue that could greatly exceed what might be predicted from its raw share of unit sales. Apple likely fears that if Google gets all its users' searches, shopping, and maps inquiries, Apple will wake up to find its former partner knows more about its customers than it does itself - and more about how to poach them. Apple's move to services is thus in significant part defensive, though it has been working to build an advertising platform that theoretically might compete with Google in some markets (e.g., ad-supported software, in-map ad placement, and perhaps even third-party sites seeking ads for revenue on static pages).
Although a comparison of hardware is beyond the scope of this article, Android vendors who sell high-end devices to compete with iOS products do not sell those high-end units primarily; rather, those vendors generally sell a large volume of lower-end, lower-margin products to customers who may also use lower-end data plans and have lower-end shopping tastes for the products that drive post-sales revenue. The question for Android is whether this situation is a long-term problem, or whether eventual commoditization will drive former Apple customers to Android devices. And that question turns on whether Apple succeeds in maintaining the differentiation in the market now enjoyed by its products. As discussed below, this will in turn depend on hardware factors discussed in a sister article.
Microsoft. Microsoft (MSFT) made its fortune as a platform vendor: it never made PCs, it made PC operating systems. Microsoft, consequently, has little concern for the profit on a PC sale. Microsoft's concern is that PCs ship with a paid-for copy of a Microsoft operating system - preferably, lots of PCs. Microsoft's primary interest is unit volume - the metric that drives license count. This could not be more plainly reflected in Microsoft's express declaration that its objective in China is to be the leader in smartphone unit sales, and that it will lead unit sales in smartphones by driving down smartphone prices (and presumably margins, and manufacturers' willingness to take risks on R&D investment). Imagine you are a hardware vendor. Do you want this in your future? Although Microsoft took a test drive of the phone handset market when it briefly sold Microsoft Kin hardware (as it had also once sold music player hardware), Microsoft mostly quit as a mobile hardware vendor. (However, it did announce it would be an ARM tablet vendor. Microsoft reports its tablet is off to a "modest start because Surface is only available" from limited points of distribution.)
Although Microsoft has gained a double-digit share in Italy's smartphone market and appears poised to displace Research In Motion (RIMM) as the third-largest smartphone platform by unit volume in Europe, its global sales share stands behind Nokia's (NOK) now-defunct Symbian operating system. Samsung's Bada operating system managed to maintain its earlier lead on Microsoft with 2Q2012 unit sales share of about 2% of global smartphone market share, which Windows Phone reportedly passed in 3Q2012 while Android soared to 75% share. Outside Nokia, whichom Microsoft is paying to build Windows phones, little Windows phone volume has yet been reported. On the bright side, 140% annual growth to 2% share makes Microsoft's phone OS the fastest-growing smartphone platform. On the downside, Nokia's EOL'ed Symbian OS, despite losing over 77% of its sales volumes, still outsold Microsoft Windows Phone in the third quarter.
Microsoft's response to Google in mobile may look a lot like its response to Google in applications. Google's efforts to sell cloud-based solutions to enterprise customers - who historically deployed Microsoft-licensed applications running on customer-owned hardware - has been counterattacked by Microsoft on its own YouTube channel. (Entertainingly, YouTube is a Google property.) The gist? "These guys don't know Enterprise, and you should laugh at their effort to win your business." Enterprise decision makers may not watch the video, but they surely hear the message from Microsoft's sales team. Note that Microsoft didn't attack Google's delivery of applications as a tool for individual consumers or students: Microsoft directed its attack on preventing erosion of its enterprise base.
Direct attack on Android as a consumer product seems unlikely: Android has been proven for a few years, it has acceptance Microsoft's mobile OS presently lacks, and Android is now the market share leader. Microsoft's consumer market push will flow from advertising its novelty, appearance, and features as differentiating factors. In the enterprise, Microsoft will likely mount its attack at infrastructure support systems, and claim that administrative ease, compatibility, and other factors favor Microsoft platforms (of all kinds, including mobile) in enterprise. Microsoft's mobile device push includes an enormous and visible advertising campaign (striking on television, in print, and on billboards), but its stickiest efforts may be in its bastions of historic support in enterprise IT, where mangers' job security is enhanced by maintaining the value and relevance of Microsoft certifications that would be more relevant in an organization deploying Microsoft platforms than in an organization allowed to slide into deploying what was expedient.
Efforts to push Microsoft platforms into obsolescence should be expected to be greeted with great resistance from Redmond. While Microsoft may not appeal to everyone, intrusion on historically safe Microsoft markets will engender well-funded attacks. And not just on YouTube: Microsoft has been effective at preventing governments from requiring their departments use word processing tools compatible with open file formats. This kind of defense may not be directly relevant to mobile platform competition, but it gives insight into Microsoft's willingness and ability to spend money to influence decision-makers effectively.
Microsoft has modest global share in mobile platforms, but just wait. Microsoft lost billions of dollars for years to bring the XBox to the forefront of game consoles, and now look at it, gaining share in a declining hardware market, and selling software sufficient to easily cover its losses and more. Reading from the same playbook, Microsoft has already committed to paying Nokia a quarter of a billion dollars per quarter to pre-install a Microsoft smartphone OS in its money-losing hardware (that is, money-losing even after the billion-per-year subsidy from Redmond). Microsoft is in this for the long haul and has already opened its wallet in a renewed assault. Microsoft's competitors will sell into Microsoft-targeted mobile segments against a tide of subsidized competition.
Microsoft will take share; the question is, from whom? As hardware capabilities advance and the price of hardware suitable for its current mobile platforms becomes cheaper, the answer will become simpler: from everyone selling mobile devices at any price. (Though, perhaps not yet in tablets. And 300% growth in phones sounds less impressive when you recall Microsoft refused to admit how few Windows Phone 7 units it sold.) How soon Microsoft can obtain profit from the share it ultimately buys is a question for another article.
Research In Motion. Research In Motion , maker of the once-dominant BlackBerry operating system, is after some delays set to replace its platform in Janauary (on the 30th) with a derivative of the QNX operating system it bought from Harman International in 2010. Reviewers lauded the new OS, but current-product sales volumes have plummeted and some speculate that by the January launch RIMM may be unable to retain its once-broad customer reach due to retailers' disinterest in carrying unproductive inventory of what some characterize as a dying platform.
Whether RIMM's platform has a bright future will doubtless become clearer next quarter, when the company demonstrates (with sales, for good or ill) the nature of the market's demand for its future products. Some of those issues will revolve around hardware and hardware margins (and thus are beyond the scope of this article), but fundamentally the ability of BlackBerry to find a profitable market segment will turn on the quality of the platform. Blackberry's ability to attract a suitable customer base with its new product release will turn on the product's distinctiveness, which in turn will initially depend much more on the platform than the hardware. (Lots of folks presently make tolerable hardware. Absent hardware distinctiveness capable of resisting imitation, mobile device differentiation depends on platform.) The prospects for Research In Motion delivering a distinctive BlackBerry product fast enough to thrive again has been discussed in Research In Motion Over The Long Run, and revisited after a share slump in Research In Motion - Up, But Better?; the principal concerns facing the company remain those outlined there.
With respect to the platform war, however, it's important that Research In Motion historically sold a premium-priced retail product, and made profits on the sale of the hardware and platform units at the same time as Apple. However, the company also signed customers to a subscription service to obtain ongoing revenue over the life of a device. Now, Research In Motion has an application marketplace called BlackBerry App World, from which it is in a position (at least in theory) to earn revenue from third-party applications. The BlackBerry platform's historic advantage has been carrier-subsidized retail profitability coupled with post-sales service subscription revenue. Whether Research In Motion enjoys sufficient service subscription fees from customers, significant subsidy from carriers, and acceptance of premium retail pricing will all depend critically on the size of the market segment that views BlackBerry as positively differentiated from a host of competitors fighting to serve the same customers for less money over the long run.
The BlackBerry's historic source of support was in enterprises that issued them - and nothing else - to users. The Bring-Your-Own-Device trend, demand for iPhones in enterprise, and other factors have eliminated enterprise as a safe market. Research In Motion's decision to sell back-end software that supports its new OS but no current devices - and separately to prevent BlackBerry Enterprise Server from managing BlackBerry 10 despite customers' historic use of the product to manage BlackBerry devices - may drive customers into the arms of competitors at upgrade time. (BlackBerry devices of different OS versions can be managed from a common web interface, but this is only a front end to separate servers customers will have to deploy if they want to manage BB10 and also anything else.) In a bid to keep customers, Research In Motion has modified its device management software to support iOS devices - and no question why: the software not only costs $99 per managed device, but carries a $4/device/month post-installation licensing fee. Is it plausible no-one will deliver management software to enterprises for less? Nevertheless, Research In Motion is spinning the trend toward BYOD as a selling point for BB10, and is gearing up for the BB10 launch and its battle for survival.
It's hard to believe Research In Motion will ever see its old margins again.
Samsung. Samsung Electronics Co. Ltd. (SSNLF.PK) [whose ticker appears to trade in the U.S. despite Samsung's contrary FAQ], is mostly known in the mobile market as a powerful supplier of OEM parts and as the world's leading vendor of Android devices. The Android market is a highly divided and competitive commodity market, so selling Android affords Samsung no differentiation from competitors' devices. Samsung - covering its bases without any special loyalty - is therefore also a vendor of smartphones running Microsoft's mobile platform. As a hardware manufacturer, it sells both high-end Galaxy products and lower-end devices, but wants ways to boost its margins. And like Apple, it doesn't want to get the blame when its products randomly reboot from others' errs. So what is Samsung to do about this?
Though it is the world's leading vendor of Android hardware and a partner in Microsoft's Windows Phone 8 launch, Samsung is also responsible for the Bada platform. What's that? Never heard of Bada? Why, last year, the Bada platform garnered greater global share than Windows Phone, a position it still held earlier this year. With about 2% of global share, Bada holds twice the 1% share Steve Jobs initially targeted for the iPhone.
I hear the laughter already: two percent share? They're doomed! Doomed! Right? But think on this a bit. Unlike the PC platform wars of the last century, in which file-format lock-in precluded interoperation with competitors (thus driving platform monoculture through network effects), there is no reason users of different smartphones can't use communication standards to call one another, employ email and SMS standards to send text and images, or depend on standards like HTML5, H.264, MPEG-4, PDF, and so forth to view the same Internet content and read the same publications. (Proprietary file formats or single-vendor DRM schemes for e-readers are in most cases read by cross-platform applications, preventing device lock-in even if they create some dependence on use of a different e-reader for each store.) So what is it, exactly, that makes 2% untenable? Seemingly nothing. People who like their Microsoft phones haven't returned them simply because they are uncommon, and neither do drivers of exotic cars (which burn the same fuel on the same roads and function as advertised despite being offscale low on global sales share charts).
So, what does a 2% platform share afford Samsung? In a market the size of the global smartphone market, 2% is enough unit sales to render Samsung's OS overhead tiny on a per-unit basis. And with control over users' default search, maps, and other service offerings, Samsung stands to gain whatever benefit a platform vendor can gain from tens of millions of units in use.
This year, Bada is being combined with Samsung's Intel-backed, Linux-based Tizen. Hardware vendors like Intel and Samsung may want to promote open platforms that don't drive their customers into the arms of big competitors like Google or Microsoft, easily explaining the motivation behind developing Tizen into an open product suitable for driving products made by a whole universe of OEM manufacturers who would like to compete on hardware, and to offer platform choice without paying a fortune for it.
If successful, the effort will drive differentiation in the mass market back to hardware issues. The cost/feature war inherent the hardware competition can be expected to drive hardware vendors into a margins-sapping commodity competition that will impair participants' willingness to risk capital on device developments that may simply be copied by competitors, however. Vendors may seek to attract customers by offering a choice of platforms on similar hardware, but if every vendor has the same low-cost choices this won't be a differentiating factor. The result will be cut-throat commodity competition in hardware.
So, where's the percentage in yet another open mobile platform?
Through a deal with Baidu (BIDU) Samsung will provide Chinese users connection to non-Google cloud services. Why? Samsung wants some of that revenue Google's getting, and surely seeks to protect against risks arising from increasing Google's influence in China. Chinese customers have been habituated to non-Google data sources, and Samsung may share some of Apple's concerns about Google's access to customers' service use and search data. Open platforms not controlled by a data-and-software competitor afford post-sales revenue opportunities presumably valuable to offset thin margins at the time of sale.
The take-home lesson of Samsung as a platform vendor is that (A) even small market share is plausibly sustainable, and (B) Android is no more immune to competition than was iOS before it, or BlackBerry before that.
So Is Android Free, Or What?
No. Both Apple and Microsoft have litigated (and each has settled) disputes about intellectual property ("IP") that Google had no power to convey in its Android licenses. IP claims create costs - and in some cases damages awards or post-settlement licensing fees - for Android OEMs. So Android is not free of cost.
(Note that in software development circles, "free" can have a practical meaning more important than bare cost, which is the right of developers to inspect and modify code. "Free software" in that sense is also called 'software libre' in that it frees rather than confines developers. Many no-cost - or "free as in free beer" - operating system licenses are still not really free of cost because of the IP litigation overhead and patent licensing described above.)
But if Android isn't free, why wouldn't an OEM license something else? It is important for evaluating the competitive landscape that no one other than Apple can sell devices running iOS; no one other than Research In Motion can sell devices running the BlackBerry OS; and while Bada continues to exist independently of Tizen no one other than Samsung will sell Bada devices. For these platforms, vendors' only competition is an alternative OS such as Android, making Android the closest thing to direct competition for every differentiated platform in the market. Microsoft does license its mobile operating systems, but the reported licensing fees (as much as $30 per unit for the prior OS version, and $50 or more for Windows RT) are much higher than the reported costs of per-unit licensing to settle IP disputes; Android is thus cheaper than Microsoft's platform to anyone other than Microsoft. (Unless perhaps you are Nokia, and being paid a billion a year to ship Microsoft licenses. Don't expect that to continue long if Nokia gets into the black on those handsets.)
Consequently, in a commodity-hardware environment, Android is the price leader for everyone not shipping an in-house operating system. Whether developing and shipping a proprietary operating system is cheaper on a per-unit basis depends on a number of factors such as units shipped. With only 2% of the market, Bada's overhead is spread across many millions of units, making it cheaper than the Microsoft licenses that offer an alternative to Android. With the migration to a Linux kernel and a development partner in Intel, Samsung will bear maintenance overhead for even less code going forward. In the era of standards, the barrier to entry in the platform market would seem much lower than in the past.
But who will deliver the apps that make the thing useful to customers? As recently learned by such well-known players as Microsoft and Research In Motion, a new platform doesn't attract all the developers one wants. Both had to pay developers to build key apps. There is barrier to new platforms, and it's currently created by those who deliver tools to make platforms worth using.
Currently, the manufacturers willing to take the capital risk of their own operating systems are Apple, Research In Motion, and Samsung. Of these, Samsung is willing to sell any other OS available for licensing as well: it is a Microsoft licensee and the world's leading vendor of Android devices, both while maintaining a Bada platform share similar to Microsoft's at the beginning of 2012.
Android isn't without cost, but it's the least costly alternative to license and frees vendors from the capital risk of developing a platform that might not sell enough units to cover its costs. For risk-averse vendors, Android is the clear choice. And in a competitive commodity market, what vendor without an established history of profiting from its own platform won't be averse to the risk of launching a novel platform?
As the low-cost leader, a mature Android should be expected to dominate in the handset market. This is particularly the case in market segments in which commoditization compresses margins to the point platform-licensing overhead for another OS would consume most of the hardware margins. Given the increasing power of hardware - Apple's iPad doubled its performance in the roughly half-year between its third and fourth generations, without reducing battery duration - it's likely that the future holds broad availability of fully-functional high-end devices that require platform features to differentiate products from competitors. Thus, margins will compress, and manufacturers without an in-house OS will be driven to freeware like Android.
Freeware's place as the volume leader in smartphone platforms is no accident. It will continue, regardless who ends up leading in hardware. By seeding the world with Android devices that channel users' mobile activity into revenue-generating opportunities for Google - by using maps, searching for restaurants and other destinations, or even browsing the web and seeing Google's static advertisements on sites receiving ads from Google - Google will increase its mobile revenue. By how much? Google's current billions per year reflects lots of penetration at the low end where online use (and ad revenue) is lower. As Google's platform share marches up the ladder of product performance, taking more and more share of the growing market for phones and tablets people will actually use to read content, Google's access to eyeballs (and the value of its ad streams) will blossom. Over time, maintaining differentiation for a platform should become increasingly difficult, making freeware alternatives more and more interesting to customers comparing similar hardware.
Although discriminating customers who value Google's expertise in search and maps can access Google's data (and advertisements) from any device, Google need not spend a fortune on advertising Google search properties to users of hostile platforms if it effectively controls the default search and map functions of most of the world's mobile devices.
Microsoft's efforts to address this problem in smartphones - where it historically hasn't had significant sales - has led to a billion-dollar subsidy of the failing Nokia. In tablets, Microsoft has been driven to become a hardware vendor, though the effort drove OEMs out of the business of selling Microsoft's platform. Microsoft's success will turn on things like the staying power of its commitment to subsidize OEMs' hardware and advertising efforts, and the ability of Microsoft to create and sustain differentiation in enterprise through features related to Microsoft's existing enterprise efforts.
Google will dominate mobile platform volume, and the losers will be competing vendors whose hardware and software fail to differentiate their products.
One More Thing
Communication standards - including file format standards, but particularly web standards such as HTML5 - reduce the power of platforms to lock customers into repeat purchases.
Consider the example of Adobe (ADBE), whose once-ubiquitous Flash plug-in came under attack from standards. Flash is being replaced by content providers who once depended on it, the better to access mobile users who either can't install a Flash plug-in or refuse to accept the performance and security risks associated with it. And as Flash was once blamed for ARM notebook delays, dependence on it can hardly be considered by competitors with a possibility of pushing an open standard that can be implemented without dependence on a third party like Adobe. Adobe, which makes tools for content creators, has had to support HTML5 in order to keep customers who want to reach the world of real users - thus further reducing the barrier to abandoning the proprietary communication protocols and file formats involved in Flash lock-in. About a year and a half after Steve Jobs' letter explaining Apple's reasoning in excluding proprietary Flash plug-ins from mobile devices, the mobile Flash plug-in and 750 full-time jobs were reported killed by Adobe just over a year ago. The desktop plug-in still ships, and you can still find Flash-dependent video, but the trend seems pretty clear: Flash player is heading out.
With more and more work being done online, the amount of productivity that's being directed into web portals and open-standards mailserver infrastructure paints a dim future for proprietary third party protocols and file formats. People who need to exchange specialized word processing features without corruption probably won't risk file format converters that are iffy with features like tables of contents, change-tracking, formatting, and font; so for the near term the file formats of Microsoft Word (the leader by sales) and Word Perfect (still in shockingly high use in the legal niche) will be safe from obsolescence until collaboration on documents is more commonly a cloud-driven project. But many users create documents alone for homework, business displays, academic publication, and so forth - and will easily migrate to a cheaper or less buggy or more easily backed-up alternative.
Well, what about this?
Several years ago, Google announced it was working on a fully-open operating system project called ChromeOS. The whole point of ChromeOS is to give PC users access to the web, where most of their time is spent anyway. Like Android, Google licenses ChromeOS cost-free. ChromeOS leverages open web standards like HTML5 to provide access to the applications demanded by users. HTML5 includes things like offline local storage using lightweight database, which is good for caching and performance, but also supports offline work between synch opportunities. As proven on multiple mobile operating systems, HTML5 supports entire freestanding applications. Without needing the application programming interfaces sold by Apple, Microsoft, or Research In Motion, users can get fully-functional applications to run on their machines - connected to the net or not - and developers don't need anything more than the knowledge users can access Chrome in order to develop for ChromeOS.
Thinking back on the movement of Google's platform into the share-leading position in smartphones, and looking forward to its platform becoming the share-leader in tablets, one feels compelled to ask why the factors driving mobile unit sales share dominance won't operate in the notebook computer market. (Or the desktop market.) As in smartphones and tablets - in which Android first appeared in cheaper, lower-margin models - Google's PC platform is available as a pre-installed OS in small notebooks from Samsung (at the time of writing, from $250) and Acer (starting at $200). Samsung sells a $329 ChromeOS desktop with solid-state storage. In these low-end machines, one could potentially deliver a mainstream Linux distribution for the same licensing cost, but at what relative support cost? Google's effort to deliver user-friendliness, auto-updates, integrity checking, convenient application security, and other support-call-generating topics pay off in lowered vendor overhead. Since re-imaging of ChromeOS is non-destructive of user data and settings, users will learn not to fear maintenance (updates, etc.) and will reap the benefits of updated software. Hardware vendors will reap benefit from these things, too.
ChromeOS has been available on hardware for about two years. Although the first year didn't see explosive growth in ChromeOS, the Chrome browser has 12% market share, which is growing. This means 12% of the market is capable of running applications targeting Chrome, and fully supported by ChromeOS. This represents a significant off-migration tool from anyone else's operating system to Google's. Naturally, Google doesn't need to capture OS share to capture the eyeballs that drive ad revenue. But as Google begins supplying more and more of the back-end of users' technology experience, Google is in an increasingly powerful position to provide the online data products that drive its revenue.
As with smartphones - after one year on the market, Android had low-single-digit market share but now outnumber all other platform shipments - Android's tablets have moved from trivial share to nearly half the market in a fairly short time. The low-end market where Chromebooks are marketed isn't a market in which users are locked in by thousands of dollars of software investment. Moreover, it's a market in which users are sensitive to price and manufacturers are sensitive to factors impacting margins. ChromeOS will ride a wave of development for Google platforms driven by Android, and benefit from users' increasing expectation that their data should be available to them on any device. The PC replacement cycle may be a bit longer than the smartphone replacement cycle, slowing the speed with which successive generations of users will likely experience (and recommend) Google's platform.
Microsoft acknowledges the risk to its desktop and applications platforms by offering web-centric alternatives. The crippled Microsoft Works that used to come on low-end machines once required users to pay costly software licenses to enable file-format compatibility with a world full of Microsoft Office users; but now anyone with a Hotmail account (sorry, a 'Windows Live ID') can get Office by browser cost-free. It should be evident that protecting the MS-Office franchise by making it available free online will impact licensing revenue for new machines, and ultimately Microsoft's ability to sell an OS to users who understand they can do everything they want online. The movement of work to the cloud has driven nearly every Microsoft property toward accessibility through a standards-based browser. The message may not be out, but it's coming: Google and the champions of open standards will render client operating systems irrelevant for quite a few PC users.
Not today, no. But it's coming.
Platforms Over The Long Term: Google
Proprietary platform vendors can either embrace standards and compete on performance and user-friendliness (while lowering the barriers to cross-platform migration) or they can be left in the past. Apple's moves to make its platform depend on PDF, HTML5, MPEG-4, H.264, and other open standards (which include some non-discriminatorily-licensed open standards) are movements in this direction, even as Apple has continued to improve its developers' experience with its native Objective C programming language. The standards help Apple customers (and developers) interoperate with a world full of machines that Apple doesn't influence, thereby lowering both customer lock-in and the barriers to adoption facing prospective customers. Apple has gone so far as to eliminate the DRM-driven lock-in formerly driven by purchases from its MPEG-4-driven music store, despite being the world's dominant music player vendor and its largest music retailer. (Apple hasn't been able to negotiate as favorable terms from video vendors as from music vendors, so DRM and its associated lock-in persist in video at Apple as elsewhere). The same is true of with Bada's combination with Tizen, where display programming now depends on HTML5 rather than the proprietary C++-driven Touchwiz interface peculiar to Samsung. Specialty niches may continue to be dominated by proprietary platforms supporting the software built for markets not yet overtaken by standards, but the vast majority of email-and-photo-album users will be free to move anywhere they like.
And Google will be there, offering a platform for the commodity machines fighting to serve the undifferentiated middle of that market at the lowest cost. As the commodity market migrates toward a free-as-in-free-beer platform to preserve margins, software dependent on proprietary platforms' programming interfaces will survive only to the extent their genuine differentiation is sufficient to drive a decision to upsell a customer to an entirely different (costlier) platform. The platform licensing model that made Microsoft a consumer desktop powerhouse in the 1990s seems strikingly inconsistent with this long-term trend. And in the competition between Google and Microsoft to profit from use-based revenue streams, we've seen the result before. In the PC market, Microsoft has been selling server-based products like Sharepoint to enterprises so their in-house application investments will lead to ongoing license fees that don't depend on specific client desktops. (And for customers without a Microsoft-licensed back-end to support Sharepoint, Microsoft offers Sharepoint Online.) It appears Microsoft is preparing itself for a day that OEMs in the PC business - or their enterprise customers - won't pay for desktop OS licenses ... and maybe not so many servers, either. Microsoft is maturing a model for making money without control of the OS market, by selling cloud services, developing licensing revenues, and expanding into things like entertainment and search. Microsoft's new platform is the domain currently occupied by the likes of Oracle, Google, Apple, and every application vendor Microsoft can replace or turn into a Microsoft VAR.
This doesn't mean that non-Google platforms are doomed. Rather, it means that the movement toward standards has a stickiness that will drive more and more applications toward platform-agnostic standards, lowering the barrier to platform migration. It means that in the commodity market, fee-free software licensing will displace competitors due to narrowing hardware margins in the face of predicted consumer price inelasticity. It means that proprietary device platforms will be driven to support the very standards that undermine the lock-in that once married customers to those platforms for life. Mostly, it means that manufacturers' struggle to differentiate computing products will be very interesting for the next few decades - and that companies able to offer hardware differentiation will benefit greatly as standards' support for cross-platform software diminishes the differentiating power of platforms. It means that applications will increase in importance compared with operating systems in driving revenue.
As is under way in mobile, Google will come to dominate market share of platforms shipping on commodity consumer hardware in the traditional market for personal computers. This will occur less swiftly than in mobile due to customer habituation and the file format lock-in affecting a significant existing customer base. Competing vendors will succeed only to the extent their hardware and software differentiate their products from the vast commodity market in which Google's pricing will give it advantages in obtaining premium positioning from OEMs among their product lineup.
And the bigger the pie of net-connected devices, and the more dependent the world's consumers become on third-party search and ad-supported content (whether static or bespoke), the bigger the revenue stream Google will be chasing with its platform program. As mobile devices and PCs become increasingly commoditized, Google will enjoy a dominant position in offering the platform of choice for the global mass market.