Forecasting Windows market share

Last week Frank X. Shaw, VP of corporate communications at Microsoft stated:

 … most of the people around me were using their iPads exactly as they would a laptop – physical keyboard attached, typing away, connected to a network of some kind, creating a document or tweet or blog or article. In that context, it’s hard to distinguish between a tablet and a notebook or laptop. The form factors are different, but let’s be clear, each is a PC.

Actually this “admission” that iPads are PCs is not something new. Steve Ballmer made the same assertion in 2010 pre-iPad (though calling them slates). Arguably, the notion that tablets are PCs has been dogma at Microsoft for over a decade and Windows running on all form factors has been a strategic guiding principle.

Which is why I’ve always added the tablet data to the PC data to create a picture of the “personal computing” market. And this is what that picture looks like today:

Screen Shot 2013-06-03 at 6-3-5.31.21 PM

Note how the share of various platforms has evolved over this brief time span:

Screen Shot 2013-06-03 at 6-3-5.32.43 PM

Seen this way, Windows has now reached 60% market share and it’s likely to dip below 50% during this year. What happens beyond then is harder to imagine. If Windows tablets start growing as fast as the tablet market overall then Windows could stabilize in share. But if Android and iOS tablets follow their phone brethren in growth then it will be far harder for Microsoft to maintain share. But is that cause for concern?

Not necessarily.

The total computing market[1] is likely to expand to over 4 billion users with 1.5 devices per user in the next five years. That expansion implies that 20% share equals more than one billion devices, making such an ecosystem “good enough” for the average developer. It certainly has been good enough for Windows developers to date and they have kept hiring it throughout the new mobile app revolution.

So even if Windows dips to only 20% of the world’s computing market it will still be perfectly “viable” for some time to come.

  1. I define the computing market as the total number of devices which have (a) a CPU (b) a broadband connection (c) a native application execution environment which is open to third party apps. This definition implies the presence of an “ecosystem” which is bound specifically to a platform.
  • Will

    Just curious, how do you compute market share from sale share?

  • Walt French

    Courtesy of Somebody Else on Twitter*: Microsoft has re-defined PCs so as to make Apple the #1 PC brand. Quite the change.

    * Somebody else should provide a proper cite.

  • It would be interesting to see the market share of the desktop form factor over that same timeframe. I think Windows’ fate is tied, more than any other OS, to that of the desktop form factor. Apple has made a distinct and obvious effort to move their users to the laptop.

    I believe the desktop computer experience is one that is ripe for disruption. There is still a clear need for desktop computing… but the current 30 year old paradigm can’t compete with more portable means for convenience. There are computationally intensive tasks (3D rendering for example) that can’t be served by laptops or mobile devices.

    Perhaps fully cloud-based computation backend services in support of mobile devices is the answer?

    • Kenton Douglas

      You answer your own question. Technologies like WebGL/CL on the browser, along with cloud based backends for 3D rendering, etc means current desktops will become moot as a requirement for “real work”. The real battle is for the browser. This will utilise thin clients (desktop or otherwise). The key OSs to watch are the likes of Windows RT and Chrome OS. In my humble opinion 🙂

      • I don’t see Autodesk racing to make web-based versions of Maya nor Newtek Lightwave. I don’t think a web-based app could duplicate the front-end requirements for modeling, rigging, lighting etc. I doubt that any browser-based scripting language has the complexity to produce such applications that perform with the required speed… without being computationally intensive, thus defeating the “thin client” goal.

        No i think the answer is a robust native application connected to a backend bank of scalable processors. Perhaps the computation time could be an in-app purchase.

        This week only 3 trillion polys only 99 cents.

      • Prasanna Kumar Ranjan

        Perhaps a CloudTop or CloudTap!!

        Thence, all one need is a virtual (yet, native & local) control to the app process running on a VM, hosted in a data center.

      • Render farms aren’t anything new, so making them cloud based is just an extension. Although you are probably right (for now) that the front ends will need fairly beefy desktops with big screens.

      • Kenton Douglas

        It’s certainly not anything we’re going to see overnight. I think we have to look a few years down the line with faster and more resilient data pipes before we see any kind of an emergence.

        In referencing Windows RT and Chrome OS I’m talking about sandboxed runtime environments rather than ‘just’ the browser. Clearly, they both need plenty of work, and they take different approaches (Windows Runtime v Chrome Packaged Apps/NaCl), but in each case the browser is front and centre. In not suggesting all apps will/could be browser based (the majority might well be thin client hybrid). The fat client (PC-based) app isn’t going anywhere either, and by definition niche application areas will remain on the PC. But, by the same token, if the market does exist, is an architecture like Nividia’s VCA not possible for cloud based rendering? There’s plenty of scope for properly architected client/server apps utilising cloud based compute and client side draw/presentation functionality. The obvious benefit here is scable on-demand compute as you mention. I would agree that the underlying programming technologies will need to evolve (not a fan of JavaScript). I can’t be confident of what will emerge in terms of the next set of web programming, but I personally like the look of stuff like Dart (with good SIMD support) alongside NaCl, and ASM.js. To reiterate though, these technologies (including WebGL/CL) are barely out of the cradle. Let’s see how things evolve 🙂

      • JohnDoey

        Web apps conquer all is an antique idea. That is 2003 right there, when HTML5 was called Web Apps 1.0. In 2008 this thing called App Store brought native C/C++ OpenGL apps into the Internet era, and made them as easy to install and use as Web apps, and even made them safer to use. Today, consumers expect fast 3D graphics, rich media, and highly sophisticated interactivity from their apps which the Web has not even begun the process of eventually supporting.

        WebGL allows arbitrary code to run native on the device — that breaks the architecture of the World Wide Web and is highly insecure. App Store type certifications are required for that kind of graphics to be secure.

        HTML5 is replacing Microsoft Word and RTF and PDF and Flash and Java and HTML4, but not C/C++ apps. The Web is not even close to running Instagram (except the viewer,) let alone Photoshop.

      • David Leppik

        Actually, WebGL is a JavaScript front-end to OpenGL. It does not run native code. It does allow programs written in OpenGL Shader Language (a variant of C) to be run in the web browser; but since the browser compiles it, and the language is already limited, it poses no more security risks than JavaScript.

        For a while Microsoft claimed that WebGL was insecure. What they were really saying was they didn’t want to spend the time to write a secure version– particularly when they have been telling developers for years to use their own proprietary 3D language instead.

        Browser vendors take security extremely seriously. I’ve seen them implement some remarkable stuff over the years to keep user data private and client computers safe. They wouldn’t implement OpenGL if it were inherently insecure.

      • Kenton Douglas

        “WebGL allows arbitrary code to run native on the device — that breaks the architecture of the World Wide Web and is highly insecure” … which of the major browsers (Chrome, Firefox, Opera, IE, Safari) excludes support for WebGL? Agreed on the extra performance advantage of native C/C++, but that gap will close and become largely moot.

    • JohnDoey

      Laptops are just portable desktops. Big desktops without batteries are just antiques. Laptops have been disrupted by iPad. Netbooks (mini laptops) were killed by iPad inside 2 years.

      • obarthelemy

        I really don’t think the part about the iPad, or even tablets in général, killing Netbooks is true.
        – tablets do overlap with netbooks, but 50% rather than 100%. Netbooks do Windows/Office, TB of storage, windowing, standard I/O ports for screen and peripherals… plenty of use cases recquires at leats one of those.
        – before tablets, many people around me got netbooks. I’m looking for a cheap one now, so I’ve asked around: nobody want to part with theirs, because they still use it, though much more occasionally (mainly holidays)
        – what killed the netbook was 0 hardware evolution. Intel and MS blocked specs (10″, 1024×758, Max RAM at 1 GB or 2)… Once everyone interested got a netbook, there was no reason to upgrade, ever. I still got a 4yo one… only now are better ones popping up (similar price, better specs), such as the Asus 1015E.

      • Kizedek

        It’s the “doing” that is a matter of debate (as in “Netbooks *do* Windows/Office”).

        It would be interesting to see if there are a similar number of users actively and regularly using Pages on iOS than Office on a netbook or tablet.

        MS may have done well to create an iOS version of Office early on. But rather than play to their strengths, MS seem to want to try and do an Apple and make Office the differentiator for their own hardware, which is understandable (though the hardware is lackluster).

        Interesting take on the death of the netbook: It wasn’t the iPad, it was the fact that netbooks didn’t improve for four years; and only now are new and improved netbooks beginning to come out?

        Well, if you are correct, then the slump in PC sales will turn right around. Even though PC makers can’t make “Ultrabooks” to save their lives either (which have come and gone since the netbook), and OEMs are jumping ship. I am looking forward to seeing the revival of the netbook.

      • I’ve never seen a laptop that I could outfit with as much RAM or 3rd party card options as I can with my Mac Pro. Perhaps some of that can be handled through breakout boxes but I doubt it. For a laptop to fill the needs of power users new services and accessories would be required. To some extent those needs are being developed… but a backend processing service would eliminate the need for 32GB of onboard RAM.

  • I’m confused by your last graph, personal computer platform market share. I’m assuming those numbers include all iOS and Android devices, not just tablets? If just tablets, Android tablets surpassed iOS tablets? I thought iOS tablets were outselling Android 2 to 1.

    • The last graph is of the same data as the first two. It does not include phones. The data from IDC suggests that iOS tablets are about 40% of the market. Of course we don’t have much to go on in terms of actual purchase data. The biggest problem with market measurement is that there are many millions of devices sold in China as “grey market” tablets.

      • melci

        Analysys International believes that the iPad captured 71% marketshare in China in Q4 2012, so I wonder just how many grey market tablets there actually were?

      • Jake_in_Seoul

        Yes, IDC needs to be challenged on this nonintuitive and (to me) highly dubious claim. Even here in South Korea, in the supposed citadel of Samsung and Android power, I would estimate iPad vs other tablets at 60/40. And if there really are 10s of millions of no-name cheap tablets floating around China, they likely aren’t connecting up to Google.

      • obarthelemy

        At least on the foreigner-accessible stores, pretty much all tablets on sale have the PlayStore, and have had it for the last 6 months. Before that, the Market was almost always available (via the right .apk), and if not other markets.

      • Jake_in_Seoul

        From a March 29, 2023 Silicon Beat report: “In the fourth quarter ipad and ipad mini represented 83% of tablet sales in China . . . . When people [in China] think of tablets they buy ipads”

      • obarthelemy

        I think you should also look into the “grey Android desktop” market. I’m fairly sure devices such as the MK808 are selling by the 10K.

  • Sacto_Joe

    Arguably, devices like iPod Touches and iPhones are also mobile computers, especially where they serve as links to a larger OS ecosystem. I’m just sayin’….

    • Yes, but no, not in the sense that they can replace a desktop PC or a laptop, at least not yet.

    • JohnDoey

      Cars are also computers. Cars are not PC’s. Many watches are computers but not PC’s. Many phones are computers but not PC’s. Many media players are computers but not PC’s. A DVD Player is a computer but not a PC.

      It is really hard to make a PC. So hard that there have only been a small handful of them in the 36 year history of PC’s. You don’t get to just declare your watch or media player is a PC because it feels good or helps you bash Apple.

      • obarthelemy

        by definition, a *Personal* Computer lets users (or at least, owners) run whatever code and use whatever data they want. Walled-gardened iOS devices barely qualify.

      • Kizedek

        Pity those 1000-dollar Nespresso machines don’t allow their owners to put coffee “pads” or instant coffee powder in them, they can only put “pods” of a certain shape and size in them. Oh well, I guess it isn’t a coffee maker after all, even though the owner buys it to make coffee and it makes a near perfect cup of coffee each and every time.

  • willo

    These graphs are amazing when you think about it, a transformation at a very rapid rate.

    I think smart phones should be included as well, as they really are just computers with a phone capability. There are few things my smartphone can´t do that my iPad can do.

    Next, we´ll see wearable devices explode, and then embedded systems like windows, tables, tvs, etc. They´ll all be fully capable computers and easily comparable to “PCs” 5 years ago. We are living in a paradigm shift in computer technology right now.

    • JohnDoey

      An iPad can run PC apps like Keynote in a PC view (8–10 inch 1024×768) in their original PC source code (native C/C++.) A user can switch from a Mac or Windows PC to an iPad and keep all their same apps. I used to use 2 Macs, one of which only ran Pages and GarageBand, and I replaced that Mac with an iPad that runs Pages and GarageBand.

      A PC has always had a full-size view and C apps. There is an argument to be made that iPhone is a PC because it runs C apps on the same core OS from the Mac, but it doesn’t run a PC view. PC apps have to be significantly redesigned in their UI to run on iPhone. On iPad, you just replace mouse with touch as you move from mouse PC to touch PC.

  • normm

    Two things. First of all, right now it’s not so easy to decide which tablets or phones are PC replacements. Are eBook readers PC’s? Are feature phones? Eventually all phones and tablets will be PC’s, but if you count them all that way now then Windows PC’s have already been swamped, and it becomes a question of whether they can get back some market share as mobile devices become smarter.

    Secondly, your diagrams only show market share, but it would be even more interesting to graph profit share. Of course if you include all mobile devices your market share diagram looks worse for Apple, and the profit share diagram looks better.

    • Did you read the footnote?

      Non tablet EBook readers lack (c), so they would never count as a PC or PC replacement.

      • JohnDoey

        iPad is the one and only successful tablet PC platform so far because it is the only tablet that runs PC apps. The other tablets are media players, equivalent to iPod touch. They run movies, music, books, Web pages, and phone apps, just like iPod touch. iPad runs native C/C++ apps that came from other PC’s like Cubasis and Avid and DOOM and Keynote. Other tablets run Java phone apps that don’t even have PC-sized views.

        You have to run HTML to be a Web browser. You have to run C/C++ to be a PC.

        Microsoft’s tablets are next in line in tablet PC’s, but they have sold so few and have so few touch apps and require a mouse for even basic things, so they are still in the product demo or beta stage. Nobody else is even trying to make tablet PC’s yet. That is why iPad leads so much in tablet usage — iPad usage replaces not just an iPod or book reader but also a Mac or Windows PC. I commonly work all day with only an iPad, iPhone, and a few accessories. Nobody works all day on a Kindle.

        PC apps are all in C. Rewriting in Jaca takes years, which is longer than an Android device generation. And Android lacks PC class subsystems. You have to build the car and the roads. That is the number one reason Android lacks PC apps. The original developer of VNC took years to get his app onto Android, and it is just a media player. An unrelated developer ported VNC to iOS from source he was not familiar with in 3 months because he was able to reuse the majority of the original C source code as-is.

        PC is an abused term, but it has to have some meaning, some connection to all previous PC’s. It has to run PC class apps to be in a class called PC’s. A Palm V or BlackBerry Bold or Android device is not PC class. They all run Java applets.

      • Mike Brown

        iPad is the only tablet that runs PC apps? You have to run C/C++ to be a PC? LOL whut?
        1. What exactly is a PC app and what is not a PC app? You do realize that all the major mobile operating systems have app stores that allow you to install native applications…right?
        2. The vast majority of applications(native and 3rd party) in the Apple i-ecosystem are not written in C/C++….they are written in Obj-C. The vast majority of applications ever written on Microsoft Platforms are in Visual Basic.
        3. The future of computing is the network…one of the many reasons why Microsoft is circling the drain. They no longer control the API…further more, I fail to see how running an app in a browser makes your mobile device “not a PC”
        4. You don’t know what you are talking about.

      • Kizedek

        I think John is including Obj-C in the C family. Whereas many MS apps, particularly for Metro/RT, are created in .net, which seems to require an interpretive engine much like Java, and therefore they are not native, but more like java applets. The .net apps are more akin to OS X desktop widgets (under Dashboard) that have existed for sometime and would not be considered native apps being a kind of applet.

      • obarthelemy

        This is grasping at straws in order to try and find an iOS advantage.

      • Kizedek

        Where are the more complex and polished Android and RT apps: video editors, multi-track audio sequencers, multi-layered paint programs, etc.

      • obarthelemy

        I think the whole “the future of computing is the network” is misunderstanding:
        – computing has become a consumer market
        – consumer don’t know squat, and especially are confusing “on-line” with “in the cloud”
        – con men are convincing customers to hand over their apps and data while committing to nothing in exchange.

        The fallout will be fun to watch.

      • obarthelemy

        You have a weird C fetish.

      • David Leppik

        You can program in C for Android via NDK, it’s just not recommended:

        The main use is in porting OpenGL-based games, or if you need cross-platform support for a large chunk of your code base. I don’t know why VNC would have taken so long to be ported, since NDK has been around about as long as Android. Presumably the developer saw some advantage to a Java rewrite.

        You can also program in C and compile to JavaScript if you want your code to run in a web browser, via LLVM:

        The value of C-to-JavaScript is limited, since the main advantage of C (producing native code, which is usually faster and less resource-intensive) goes away when the output is JavaScript. Its only use is to port code already written in C to the browser.

        Long story short: if you have C or C++ code, you can run it on iOS or Android as native code. Or you can run it non-native on anything with a web browser.

      • normm

        So it’s a PC if Amazon lets third parties write apps for it, and not a PC if it only runs web apps? And then if Amazon changes its policies, the device changes into a PC? (This is exactly what happened to iPhone!) What if Amazon lets just one third party write apps for it — is it a PC?

        We’re in an age of marginal definitions. Are phones “smart” if they run Android, even if they are only usable as feature phones? Are devices that can’t run C apps really replacements for today’s PC’s? This kind of hair-splitting isn’t very useful, because with a few more years of evolution all the phone-like and tablet-like devices will be very smart and we’ll probably have to call them all PC’s. So I’m suggesting we define a mobile market that includes all the phones and tablets today. Then seeing who’s winning or losing market share doesn’t depend on exactly how you define a smart device or a PC!

      • I think you might be setting up a strawman argument. You specifically asked about Ebook readers, which I understood as being devices like the non-tablet Kindles. Clearly those are not PC replacements. A feature phone is clearly not a PC replacement, and by definition, an Android phone that is only being used as a feature phone is not being used in place of a PC, thus not a PC replacement.

        I don’t think I am engaging in any hairsplitting. On the contrary, you seem to be the one so engaged.

  • Your numbers indeed show that the external factors may not be a concern for the viability of MS computing platform for quite some time. But based on the lessons of so many previous market leader tech companies, the danger may come from within. When the market share goes that deep downhill in any stock exchange quoted Big Corp for so long, all kinds of (hidden) stuff hits the fan. And not all of that is rational, but fully behavioral. Stuff like consumer/customer perception and petty politics of blame game or distribution of spending cuts (or sparse investments) can have huge impact. Add to the equation simultaneous problems in the other businesses of the financial performance portfolio (Office business fat margins being a prime concern for MS), and one could end up in the vicious cycle of restructuring in which the short-term steadying actions have the risk of deepening the slide.

    • Walt French

      At the time Microsoft beat the anti-trust rap, I thought it a beginning to their end, allowing Microsoft’s many clever engineers to emphasize the Microsoft stranglehold on the desktop, and turning a blind eye to innovations outside of Windows. How much better the software would’ve been had “Office, Inc.” competed on a level playing field against Lotus, Borland, WordPerfect and others, rather than relying on cute tricks and special info about “Windows Inc”’s OS to make it appear that the other products weren’t as tuned-in.

      Well, nobody else saw it that way at the time. Yet Another Crackpot Idea. So instead, we have Office ubiquity sacrificed to try to shore up Windows on mobile, Office trying to move, about 10 years too late, to a subscription/cloud model that Google was incented to attack since Office would never have partnered with them as long as they were tied to Windows. Maybe the clash was inevitable, but it seems it was accelerated by such indiscriminate efforts to control the world.

      • Tatil_S

        About 13 years ago, I’ve used Lotus presentation and email programs when I was an intern at probably the only company still using Lotus as its office suite. Back then MS Office products had even more glaring usability issues, but Lotus turned out horribly worse. This is in addition to requiring a technician taking over your laptop for up to half a day just to install it. I have to say Lotus did not lose its market share because of cute tricks and lack of inside info on the next gen Windows OS.

      • Walt French

        US v Microsoft was 1998. At that time, Microsoft had already eviscerated Lotus et al.

        Notes had morphed into a complex monstrosity that the big consulting shops built their workflow on but yes, it was impossible for anybody else. I worked for a shop that used it for email, arguably the least attractive feature of it, but my then-employer got it “free” as part of a big iron IBM hardware bundle. I dunno whether IT was smart not to lock in, or whether they were just lazy — my group tried in vain to use the workflow stuff that one user was expert in.

        But mostly, Borland, Novell and others had obviously responsive, very good solutions that failed all the InfoWorld comparisons because they assumed Excel, Word et al had an inside track to Windows development. By the time of the suit, it was already too late.

      • JohnDoey

        It is well known that Microsoft told all their MS-DOS developers that to get onto OS/2 because Microsoft was 100% behind it, but all the while Microsoft put all their software on Windows.

      • Tatil_S

        I am sure MS had some “unfair” advantage in its Office development and that may have even been the primary reason for some competitors to disappear, but I don’t think that was the case with Lotus. With that horrible user interface. it would have lost out to MS anyways.

  • jeff g
    • It was better when it was a Businessweek article.

    • Isn’t Android without apps (i.e. embedded) effectively Linux?

      • obarthelemy

        there’s (at least) a Dalvik of difference between Linux and Android

      • masquisieras

        and a GPL licence of difference.

  • John


    As an ecosystem, Wintel is here to stay in foreseeable future. That’s a certainty. Wintel is entrenched in the business world. More statistics won’t provide additional insight.

    Look at it from another angle: Microsoft’s bonds are safe, but equity value in its stocks is uncertain.

    The real questions are:

    – How will the consumer market fair? Can Microsoft stop the exodus in its consumer market?

    – How far will the total market shrink?

    – Can the shrunk market support Microsoft’s current valuation?

    • The foreseeable future is not what it used to be. Capital equipment in the 1800’s lasted 100 years. Around WW II the equipment installed at factories was expected to last 30 years. With the advent of computers, replacement of equipment has accelerated. We are in an era of increasingly short life for fixed capital.

      So what seems obvious now — that corporations will always be built around winter networks and desktops — may not be common wisdom for as long as we’d imagine.

      • John

        Hi Jon,

        Do you know how much it will cost for a corporate to migrate to a different IT platform, with all the custom built enterprise software stacked on the top?

        An example: COBOL is an antique software platform. We are talking about something became a standard in 1960’s. It underpinned the first wave of enterprise software in the commercial world. Yet, it’s still holding 25% of the market share according to one report:

      • Do you know how much it will cost not to?

        I am not saying this change will take place overnight. But I think your statement underestimates the current disruption taking place.

        COBOL’s longevity has to do with the fact that it is simple and effective. It is an awesome language for accounting functions. Even in a rapidly changing world, things that work will stick around for a long time.

        But that’s the point. The Wintel computer paradigm is going to be rapidly impacted by mobile (and wearable) computing. There will simply be things that you can not do effectively with the old paradigm — the same way COBOL can not do everything you may want to do today. No one starts a new project in COBOL. It only exists as legacy software. It won’t disappear overnight, but it surely has no place to go but down.

        And like COBOL, the Wintel model may have an entrenched 25-50% for many years, but that would be a big change from its corporate dominance today. The amount of money corporations spend on non Wintel computing is already huge compared to three years ago. As this trend continues, it suggests less money spent on maintaining the old Wintel paradigm.

      • LTMP

        I’d go one step further and say that companies that are too heavily tied to incumbent platforms will find themselves unable to keep up with disruptive upstarts.

      • JohnDoey

        How much will it cost to train a generation of workers who have never used Windows to use Windows? People in their 30’s and 40’s already struggle with Windows XP and nobody knows where the money or downtime will come from to train them in Windows 7 or 8. People younger than that have even lower levels of Windows skills. Sure, people will fake their way through on a Windows/Office system, but if you out them on an iPad they will do more and better work in less time and the cost of the iPad is a small fraction of the cost of a Windows system plus I-T maintenance and training.

        At one office that I consulted at, the workers were refusing to accept Windows 7 upgrades and the managers were refusing to send their workers to Windows 7 training, and at the exact same time, everybody was bringing in their own iPads because they were getting owned in meetings by other workers with iPads. Nobody asked for any iPad training at all and everyone was immediately more productive.

      • californian

        “(A Wintel system uses 35x the power of an iPad and requires 35x the air conditioning.) ”

        “the workers were refusing to accept Windows 7 upgrades and the managers were refusing to send their workers to Windows 7 training”

        Nothing to write back. But any Microsoft business or technology article on any tech blog appears to bring out the worst hating replies and comments from posters without reliance on ground facts, statistical inferences, hypothesis tests, regressions etc. Not even use of epistemology. Or logic.
        Something to do with groupthink or herdthink. Social psychology should have an answer as to why hatred can exist against technology. Of all things, some people hate technology. That is actually very funny.
        Ultimately companies are made of people. Good organizational (from view of productivity which is again abstract) structures, good use of input resources and ability to keep in tandem with tecnological trending or usage should just about help any average technological company to survive generational changes. Why would not Oracle survive? Or Microsoft? Or SAP?
        Ultimately again, lack of goal directed behavior is what causes individuals or groups to so called ‘fail’. People should ask if Microsoft is not displaying goal directed behavior. I think they are displaying traits of it organizationally. So they will exist. They will have dominance in a few markets while only minimal presence in a few others.
        To write any company off as irrelevant or using a singular product point as being representative of 1 billion information workers around the world is simply put a sad generalization.
        More maturity and more analysis please.

      • It wouldn’t cost very much if the company encourages employees to bring their own devices (BYOD). In fact, by shifting the cost to employees companies could see huge savings by eliminating costly IT departments.

      • obarthelemy

        Yep. Employees should also BYOS (bring their own SAP), database, line pf business software…

    • melci

      John, the thing is that the Business PC market is actually only 15% of total worldwide PC sales now, far smaller a percentage than back in 2000 when it represented 60% of all PC sales according to Goldman Sachs.

      Likewise, business is moving to mobile devices at a frenetic pace and Apple’s iOS is far out in front of the pack with its mobile business market share increasing from 69% to 78% in Q1 2013 while Android declined from 30% last year to only 22% with Windows relegated to less than 1% of tablets and smartphones in Business according to Egnyte.

      The iPad captured 93.2% of the Business tablet market in Q4 2012 while the iPhone accounted for 73% of all non-BB business smartphones according to Good Technology. In total, Good reports that iOS devices represented 77% of mobile device activations in the enterprise market in Q4 2012.

      As such, Business is not nearly the safe haven for Microsoft that you think.

      • John

        Hi melci,

        The 15% is an interesting figure. I didn’t know that. Thanks for the information. Is this statistics available on the web?

  • Pingback: Microsoft Will Still Be Fine Even If Windows PCs Keep Tanking, Says Apple Blogger | Business Insider Australia()

  • Pingback: Microsoft Will Still Be Fine Even If Windows PCs Keep Tanking, Says Apple Blogger (MSFT) | Digital Wealth()

  • Pingback: Microsoft Will Still Be Fine Even If Windows PCs Keep Tanking, Says Apple Blogger (MSFT) |

  • Pingback: Can Microsoft thrive with just 20 percent of … – CNET | My blog()

  • obarthelemy

    It’ll be interesting to see if MS manage to become relevant in Mobile… and not just by renaming “laptops” to “tablets” after making the KB removable. Intel’s new CPUs seem much stronger for Mobile now, so surely MS can get a foot in, on the back of x86 legacy apps. Probably won’t be enough to give a leg up to the ARM stuff.

    I’m still wondering why Google aren’t making a more aggressive move in peripheral markets; consoles, desktops, laptop… Android already supports all of those, small outfits are moving into that space… it would take very little effort (mainly publishing guidelines and getting a handful of major OEMs on board) to dramatically widen Android’s hunting grounds.

    • Space Gorilla

      Tough to go hunting when you have no money. Other than Samsung the outlook re: profit and Android makers is bleak.

    • James King

      I don’t think Microsoft has to become relevant in “mobile” per se. It just has to make a version of Windows that works decently on laptops. Hell, it’s moved 100 million copies of Windows 8 just because there isn’t a popular alternative.

      The real threat to Microsoft is that a company may get bold enough to compete directly against the Windows/Office franchise once Microsoft is no longer the 800lb gorilla in the room. Google is already making forays into Microsoft’s traditional strongholds with Chrome OS and its cloud apps. The threat to Microsoft’s core business rises with the decline in its overall relevance as a platform.

      • Tatil_S

        I agree. The most commonly cited reason to buy MS Office is compatibility with everybody else. When more than half the computers cannot run it, will people feel the same need to spend hundreds of dollars to stay compatible with only the other half the world? If they need to find an office suite solution for their non-Windows devices and if that solution works on Windows just as well, will they need to pay that money even for the half of the world who could run MS Office?

      • Shameer Mulji

        I can’t speak for the average consumer, but businesses will spend the money to ensure compatibility with their documents / spreadsheets, etc.. MS Office is THE standard in office productivity as far as businesses are concerned.

        If they can’t get Office (or any other suite compatible with it) on Google or Apple devices, make no mistake, they will get it on Windows devices. Why do you think MS is dragging their heels in porting Office to the iPad or Android tablets? They know once they do, Windows won’t be necessary.

      • Tatil_S

        There are already relatively large corporations using Google Docs as an office suite, as well as a large number of iPad and Android using workers in many others. MS Office *was* the standard when almost every computer could run it and were running it. When more than half the computers sold today cannot run it at all and many Win8 devices cannot run it well, MS Office is no longer the same must have standard.

        Of course, corporations will keep buying MS Office to at least support all the legacy documentation they have already generated. Yet, many will also be on the lookout for alternative solutions for the devices that cannot run MS Office. That is the beachhead competitors did not have before the mobile revolution. Do not underestimate the impact new graduates who will bring their experience with non-MS presentation or collaborative writing tools. Once an acceptable solution is found, it will not take too long for IT to realize the alternative solution runs on Windows as well. MS may rue the day it decided to sacrifice one cash cow to artificially prop up another.

  • Whether the Windows OS “survives” is irrelevant. Without a monopoly, Microsoft loses its bully puplit and cannot dictate technology direction.

    Microsoft traditionally does not perform well when it does not have the ability to leverage its monopoly. They would need to actually write good software — something they have demonstrated they are utterly incapable of.

    • californian

      This is a non-engineering business viewpoint. Historically each giant technology vendor has come up with contributions of its own to the overall technology stack. And Microsoft has its own. Like – SOAP, AJAX, .NET, C#, DirectX, etc etc
      We are still at the beginning of the cloud transformation or cloud revolution if you talk to marketers. The direction of the enterprise hardware/software/it/systems industry from its previous quasi-equilibrium state with client/server systems is still unknown and uncharted. We still do not know where and how enterprise or business data will persistently reside adhering to all sorts of restraints and constraints. Mobility only affects the client side. The more interesting story is on the server/storage/network/compute side.
      At this point of time, it appears both Oracle and Microsoft have very long and hard roles to play in that transition. Microsoft especially holds the keys since it understands the data formats better than most.
      None of the consumer cloud companies like Apple, Google etc have even begun to understand this problem yet. And this problem will impact American and world economy in a massive way. Only Amazon has shown the guts and grit to get to these problems. But it has to be seen if they can be persistent. Even IBM is getting scared of Amazon these days as their lawsuits indicate.
      You will understand this better after Microsoft announces a new reorganization where a separate division called Enterprise will be created. And this will include all business software, cloud, hardware integration and systems management responsibilities. That will be one big mega division with annual revenues touching $60 billion or more with average annual growth rates of atleast 5% in revenue. That does say something, doesnt it?

      • Your viewpoint is extremely Microsoft-centric and will not fly in the real world. Over the last few decades I have consistently seen Microsoft using bully tactics to achieve and maintain its monopoly, and nearly all of my fellow IT executives feel the same way about using Microsoft products: they feel as if “they have to.” As that perception begins to fade, so will Microsoft.

        It is indeed likely that an “Enterprise” division will be created, selling old world products such as Exchange and SQL Server to customers who are too entrenched to escape. Meanwhile, everyone’s focus will continue to move on to more interesting things. Microsoft has followed in IBM’s footsteps since the very beginning, and there is no sign that they are diverting from that path now.

        A decline in relevance precedes a decline in revenue, for a company that size, by a decade or more. We can expect the pink slips to begin appearing in 2025 or so.

      • Des Akkari

        You are absolutely correct from my perspective as well. The world is changing and now that Android and Apple are gobbling up the rec/normal user with quality code that actually works, soon Linux is going to be the sole platform for enterprise. Sure some will say that Micro$oft will survive in the enterprise space but there are 2 things working against this….cost and reliability. As a developer I cannot for the life of me understand why developers still use Microsoft, it just gets in the way of what I am trying to do, hardily ever works the way it should….most of the time. Now that I have been coding on Linux solely for 7 years, I would not take MS crap if you paid me to use it. Even Skype, owned by MS won’t use their crap servers…Skype uses linux. Yet MS fanboys will pay for a license that MS isn’t stupid enough to use on mission critical systems, lol. Example being, would you buy a Ford from the guy driving the BMW? As this fallacy fades, so will the MS….and as a Java guy, I can’t wait to dance on the grave of C#.

      • davel

        As Des below states you are excluding the significant share linux plays here. With IBM support linux plays a significant role in server side computing. After all do you need to pay Microsoft a license to run a web server? There are many server alternatives that run very well on linux which means all you need to pay for is a good white box. In fact I bet Google, Amazon, Apple, etc do not run Microsoft for their server farms.

    • davel

      The problem for Microsoft is the world has run mobile computing on non Microsoft platforms for over 5 years now. Intel has produced Haswell and perhaps with that chip and Windows 8 ( in whatever form ) will bring a new platform to allow Microsoft to compete here. The question is does the consumer want the services that Microsoft offers or are they content with non Microsoft offering?

      As you say the web offers the user to be platform agnostic.

      Horace. thanks for the changes to your site.

  • Also keep in mind that the current Windows market share is even worse than the graph depicts. About 10% of it is computers that were purchased with a Windows license that was thrown away and replaced with Linux.

    • Mark

      That seems rather high. Source?

  • Microsoft execs claim that tablets are just smaller PCs. Whether or not this is true is largely irrelevant though, because most tablets aren’t running Windows. Which is the real problem MS has, not semantics.

    Though a simple way to show that tablets aren’t PCs *is* the fact that they aren’t running Windows. MS managed to beat everyone off every single PC platform in history (even Netbooks by slapping XP on them). Honestly I think a PC today means “generic x86 chipset compatible device”. Most tablets are custom ARM chips. No binary compatibility with x86, which means even if they did run Windows they wouldn’t run any apps. Microsoft missed this boat big time.

  • Pingback: 如果失去对桌面的统治 微软是否能生存下去?()

  • Pingback: 如果失去对桌面的统治 微软是否能生存下去? | 爱WP7 – 爱WinPhone()

  • Richard Love

    Several of the comments feel rather too polarised to retain a factual balanced view.

    While it’s certainly true (the numbers are clear) that there is a significant movement to adopt (add) new highly mobile form factors (that were initially typically Apple iOS, and are now more often Android based) if you look at the 1st actual ‘traditional’ PC market chart, then the volume is actually flat at circa 80 million a quarter… as one or two other posters have noted tablet (and now convertible) form factors are driving incremental growth.

    That’s good for all vendors who sell devices, and that includes Microsoft and their OEM partners (now that they have an actually very effective mobile/tablet as well as desktop OS (and Intel a low power high performance CPU platform) on which it can compete very effectively with iPad / Android.
    Microsoft and Intel have both been undoubtedly late to the ultra portable all day battery tablet market, but now they are there, don’t just assume that 100% of those incremental tablet/convertible devices will be Apple or Google based.
    (especially those used in business where Microsoft’s announcements this week about Windows 8.1 put it ahead of competitors for many business users).
    In the consumer space, it’s a tougher fight, as Apps are more critical, but the volume of W8 devices sold (and quite attractive profit share at 20/80) means Microsoft store Apps are steadily growing in volume (and more importantly quality)
    Don’t count this game done just yet, I personally think we’re going to see three very strong and different ecosystems that will fight for share.. Microsoft, Apple and Google… I dot think any one of them will ever dominate as MS did (which will be great for users as it will keep all three innovating fast and hard)

    • Shameer Mulji

      “Don’t count this game done just yet, I personally think we’re going to see three very strong and different ecosystems that will fight for share.. Microsoft, Apple and Google… I dot think any one of them will ever dominate as MS did (which will be great for users as it will keep all three innovating fast and hard)”

      Completely agree with this statement.

  • Pingback: 【悲報】Windowsのシェアが急降下するとの予測 | ソフトアンテナブログ()