Thursday, May 02, 2013
« Talking about your own open source proje... | Main | Good Windows 8 sales »

This question keeps floating around the Internet, and I thought I’d give my viewpoint.

Obviously I have no more intel on sales numbers than anyone else (which means pretty much nothing). So anything anyone says at this point is pretty much speculation, and that includes me too.

So rather than speculate uselessly, let me provide a bit of narrative.

Almost nobody buys operating systems. This is a truth that has existed since the dawn of the concept of an operating system. Operating systems exist as an abstraction over the hardware that allow developers to build applications without having to deal with the low-level details of interacting with hardware.

Of course over the past couple decades “operating system” has come to include a lot of things. At a minimum these days people expect the OS to include drivers for every type of device ever invented (or to be invented), a nice user interface, and a set of pre-installed “first party” apps. None of that stuff actually is the OS, but most people view it as part and parcel.

Even so, very few people buy an OS. They buy a device (computer, ultrabook, tablet, phone) because the device enables them to do something useful. More precisely the device allows them to run software (apps) that do something useful.

(I’ll freely admit that I’m biased. As a software developer, I do tend to see software as the most important part of this whole picture. And I think I’m correct, because if you give anyone a device and tell them they can’t install any software on that device I can pretty much guarantee you that they won’t be able to use the device for much of anything.)

I’d also suggest that there are two broad categories of “person” in this discussion.

There are regular individuals like you and me. We buy devices ourselves, investing hundreds or thousands of dollars so we can do things like access the Internet, write letters, manage home budgets, send email, play games, etc. Our motivations in buying a device are to gain access to the apps that allow us to do those things we consider worthwhile in life – whatever that might be for each of us. And yes, people like me also get joy out of the device itself because we’re geeks, but most people just see these things as extremely useful tools or toys.

There are organizations (which are also apparently “people” thanks to Citizens United). They buy devices for their employees so the employees can do things that provide productivity and value to the organization. Some organizations are OK with employees using those devices for personal reasons, others aren’t. In almost no case does an organization buy devices other than because an employee needs the device to perform important aspects of their job.

I keep running into people who think there’s no difference between these scenarios. And maybe I’m a little slow, but I really struggle to see how organizations are going to start buying (and supporting) devices outside the scope of enabling employees to be productive. Nor do I see how dock workers, administrative assistants, lab technicians, and other employees are going to start purchasing devices for the intent of using them at work. Actual human people buy devices to make their personal lives better, not because they intend on using them as an alternative for a work-supplied device.

(In other words BYOD is pretty much bunk – but that’s another blog post.)

To judge the success of Windows 8 then, one must evaluate it in the context of people buying devices for their own use, and organizations buying devices for their employee’s use.


From what I’ve seen there is some resistance on the part of people in terms of buying Windows 8. On the tablet side of things there aren’t a lot of apps, and people buy these devices for apps, not the device itself. On the ultrabook, laptop, and desktop side of things people buy a PC because they want to run PC software – all of which runs on Windows 7 just fine, so there’s no obvious reason to go to Windows 8. But there is an obvious reason not to go to Windows 8: everyone knows it is different, and people fear change.

So if I’m a regular person looking to buy a tablet, I’ll probably gravitate toward the tablet with the most apps – hence an iPad or perhaps a Kindle Fire. And if I’m a regular person looking for personal productivity with Office, CAD software, or gaming, I’ll be perfectly happy with Windows 7 (or even Vista) as long as my current computer keeps working.

Please note that I am personally not a regular person in this context. I’m a geek, and thus have been running Windows 8 since long before it was released. But I have sat in restaurants and overheard conversations about Windows 8 by random people – conversations that lead me to be pretty confident that my previous paragraph is correct.

None of this says that people won’t slowly adopt Windows 8 as their existing computers need replacing, because they probably will. And as more people actually start using Windows 8 and tell their friends and neighbors that it really isn’t that scary, then the fear of change will fade. And with any luck the number of apps available for WinRT will grow relatively fast so people will consider buying Windows 8 tablets because they feel confident those devices have good and useful apps.

Does this mean Windows 8 is a success or failure? I don’t know. We are in uncharted territory to some degree, because today’s computers (with their i3, i5 and i7 processors) are essentially identical in performance to computers from three years ago. For the first time in recent memory (and perhaps ever) computers have stopped getting faster, eliminating one of the primary reasons why people would buy a new computer. Now the only reason to buy a new computer is a complete failure of your existing computer, and computers often last a very long time…

In other words, Windows 8 adoption in the personal space might be slower than in the past (we don’t really know). But if it is, I strongly suspect one major factor has less to do with Windows 8 than the reality that few people are motivated to spend hundreds of dollars to buy a computer that isn’t any faster than the one they already have.


On the organizational side of things the dynamic is entirely different. Organizations try to minimize the number of device types, operating systems, and operating system versions because it is extremely expensive to support more than one. Organizations use “apps” (applications, or enterprise software) that is required for the organization to function. When those apps fail the business loses money by the second – often many thousands of dollars per second. Upgrading from Windows X to Windows Y is never done without extensive testing to ensure those important apps work on the new operating system version. And such upgrades are done according to an orchestrated plan that minimizes the time the organization is forced to support both versions.

This is nothing new. This is the reality of enterprise computing that has existed for as long as I can recall (thus dating back well more than 2 decades).

Strangely, I’ve had people argue that this dynamic is no longer true. That organizations are now going to adopt BYOD, even though that directly means supporting numerous devices, operating systems, and operating system versions all at the same time. I’m not sure what these people are smoking, nor am I sure I want to try it because I think whatever they’ve smoked caused some brain damage. But I could be wrong – perhaps organizations are ready to radically increase their IT support costs in order to allow employees to use random hardware devices and operating systems? Or perhaps there is no increased cost to IT because (as one person told me) it will now be the employee’s responsibility to ensure they have a working device at all times – thus all IT support costs will be born by the end users. Something I’m sure will thrill the minimum wage workers in the warehouses who’ll apparently now have to buy and support computers they’d never have purchased before?!?

Back in the world of the sane, what is actually happening is that a great many (most?) organizations are just now migrating from Windows 2000 or Windows XP to Windows 7. This is because those older operating systems are off support, or will be off support in April 2014. No more bug fixes. No more security patches. Nothing. I truly pity any poor souls left on XP a year from now.

This migration from XP to 7 is not cheap. It is not only an IT issue in terms of upgrading hardware, drivers, and operating system installs. It usually also means updating or replacing ancient enterprise software that was written in VB6 or PowerBuilder or other technologies that haven’t been current for many years.

As a result, almost no organizations even have Windows 8 on their radar at all. Virtually nobody is planning for a Windows 8 migration, because they are just now getting to Windows 7. In fact, my informal polling while speaking at conferences around the world is that nobody expects to move to Windows 8 until 3+ years from now.

Does this mean Windows 8 is a failure? Of course not. Remember, these organizations are just now moving from XP to 7. Windows 7 was released in 2009, and only eclipsed XP in 2012 in terms of installs.

Following that time schedule, we won’t know of Windows 8 is a success or failure until around 3 years from now: in 2015.


In summary, is Windows 8 a success or failure? I can’t say. Nor can anyone else, even though a lot of people (including myself) speculate about it quite a lot Smile

The space to watch is the personal/individual computing space, because that’ll move somewhat faster than the organizational space.

Like every previous version of Windows, we won’t know the success or failure of this version until 3+ years after its launch, because enterprises always move at a stately (if not glacial) pace.