If we weren’t already drowning in articles defending Windows 8, or claiming it was going to be a bigger disaster than even Windows Vista, for the last year this week makes you want to hop on Noah’s Ark. From Ed Bott’s “Windows 8 is the new XP” to this rebuttal (which reminded me of SNL’s parody of Point/Counterpoint) we are awash in speculation about Windows 8’s success. Rather than speculate on Windows 8’s success or failure, which I believe will be far more nuanced than most commentators, I want to explore what Microsoft’s real strategic goal likely is with the release. And I think it is different from what most people, even most Microsoft employees, may believe.
Let’s start with what is definitely not a goal for Windows 8, broad-based Enterprise adoption. By that I mean, the rollout of Windows 8 to traditional desktop and notebook computing. Keep in mind that back when Windows 8 was being planned Windows 7 had just shipped. Because of the time, cost, and complexity of enterprise-wide operating system rollouts most enterprises would either still be rolling out Windows 7 or have just completed their rollout around the time Windows 8 shipped. Indeed Windows 7 just recently passed Windows XP as the most popular version of the operating system in use. No enterprise would have the appetite to immediately start the cycle over again so quickly, and so (I believe) the prevailing wisdom inside Microsoft was that they would skip Windows 8. In other words, before Microsoft had decided on the details of a new app model, the Windows Store, the Start Screen, or removing the Start button they knew there was no point in targeting enterprise desktops as Windows 8 upgrade targets. The next window of opportunity for enterprise desktop upgrades doesn’t open until 2015 or 2016, giving Microsoft plenty of time to refine what they’ve done in Windows 8 to make it attractive for that upgrade cycle.
So if broad enterprise adoption of Windows 8 isn’t a goal then what is?
Minutes per Day (MpD) of usage is the ultimate leading indicator on the health of a product. If it drops low enough then customers wonder why they are spending money on your product at all. If it keeps increasing then the perceived value of your product to the customer goes up and they want more of it.
Microsoft faces MpD challenges across its product lines. MpD for Microsoft Word has been dropping for a decade as users stopped writing memos in favor of writing emails. Then Outlook/Exchange and Hotmail MpD dropped as users move to Instant Messaging and SMS for short messages. Windows MpD drops as consumers and information workers do more communications, information consumption, and entertainment on iPads and smartphones.
The overwhelmingly number one goal for Windows 8 (and Windows RT) is to reverse the downward MpD trend. So how does it do that? Let’s go back to the enterprise for a moment. Have there been changes in how many MpD a task worker, for example someone answering phones in a call center, spends on Microsoft products? No. How about a product engineer designing aircraft parts? Nope, no significant change. Could changes to Windows increase the MpD for these users? Probably not by a material amount, and certainly not by changing their desktop computing environment. To reverse the MpD trend you have to attack two things, the cause of the shift in existing MpD from Windows to other products and new usage scenarios that creates MpD that has yet to be captured.
Since the bulk of the downward pressure on Windows MpD comes from tablets, Windows 8 is primarily focused on addressing that trend. It tries to do that both by directly going after the tablet market that Apple created with the iPad (e.g., Surface with Windows RT), and by addressing the market segment that is being pulled to the iPad due to lack of alternatives that meet their current computing needs (Windows 8 on hybrids/convertibles).
In some cases Microsoft seeks to capture the entire MpD that is going to iPads, which is not just what has shifted over from PCs, and shifted from TV/Radio/Books/Magazines/Newspapers to iPads, but what has been created by entirely new usage scenarios that tablets enabled. Because this market is still young Microsoft has a good chance at obtaining a significant, though not leading, market share over the next couple of years. The iPad will still rule, but Microsoft might get enough share to reverse the overall MpD decline and position themselves for increased MpD over the long haul.
In other cases Microsoft is just seeking to shift some of the MpD back from iPads to Windows-based devices. Let’s assume that a user owns an iPad and also gets a Windows 8 Convertible as their next notebook. Every time they pull the Windows 8 tablet off its base and take it to a meeting instead of grabbing their iPad it increases Windows MpD. Every time they go out to lunch and take that tablet instead of their iPad to read the paper or answer email it increases Windows MpD. Every time they are sitting on an airplane and decide to read a book or watch a movie on their Windows 8 device after finishing working on a sales report, instead of putting it back in their briefcase and grabbing their iPad, it increases Windows MpD. And if you follow the MpD argument, then maybe not this year, or next year, or even three years from now, the MpD on their iPad will drop low enough that they wonder why they even bother to carry it. And then Windows becomes their full-time tablet OS.
The place where Microsoft does expect Windows 8 to succeed in the enterprise is in scenarios where tablets or other NUI-driven devices are called for. I’ve used the scenario before where a retail chain decides to give all its associates tablets so they can help customers on the floor instead of dragging them back to a desk. Having someone walking the floor with a Windows 8 device will dramatically increase MpD compared to usage from behind a desk, as well as stemming the loss of MpD that occurs when an iPad is used to implement this scenario. Telepresence is another area where I think Windows 8 will shine and lead to increases in MpD. In other words, I expect that most enterprises will adopt Windows 8 for specific scenarios while using Windows 7 as their primary client computing operating system. And that should add to overall Windows MpD.
There are similar ways to increase Windows MpD by enabling or improving consumer scenarios that benefit from NUI. From monitoring your baby’s room or home security system, to video calls with your friends and family, to checking your voicemail, to looking up recipes, to immersive entertainment, all without having to sit down and grab a mouse, there are numerous scenarios that Windows 8 targets better than any other operating system. And each leads to an opportunity to capture (often totally new) MpD.
Financial analysts, industry analysts, press, and pundits will focus on market share, units sold, revenues, and profits. But if you are really thinking about the long-term and the ultimate success or failure of Windows 8, look to MpD. I wouldn’t be surprised if when CEO Steve Ballmer and Chairman Bill Gates have strategic discussions that’s what they do.
Hal's (Im)Perfect Vision 
I’d add that another goal is to improve the underlying structure or at least address the fact that Windows’s needed to get rid of legacy software that is no longer used
WinRT is a wrapper for Win32, which was introduced in 1993, and a new version of COM that adds a new interface, IInspectable, to handle new metadata required to allow WinRT to have multi-language support similar to .Net.
No, it is not a wrapper but rather lives parallel to Win32. At least that is the intent.
Hal, I’m sorry I can’t reply directly to your reply.
If you’ll look on Page 6 of this article,
http://arstechnica.com/features/2012/10/windows-8-and-winrt-everything-old-is-new-again/
you will see why I say WinRT is a wrapper for Win32.
Simply put, does WinRT’s Windows.Storage.StorageFile.createStreamedFileAsync call Win32’s CreateFile function, NT’s NtCreateFile, or an entirely new system function? Only the first one would indicate a layering on top of Win32. The later two are not.
I also think there may be some confusion if WinRT APIs are implemented in the Windows Subsystem (e.g. User32) DLLs. Some equate that with them being layered on top of Win32, when in reality they aren’t going through the Win32 API at all but rather just sitting in the same DLL that implements Win32 functions.
The bottom line though is that Microsoft says they aren’t a layer, they are native. Maybe when the next version of Windows Internals comes out we’ll get a clear picture.
It calls CreateFileW.
There are many goals for Windows 8, I concentrated on the existential one.
Yep, and note that this also applies to some extent even to conventional laptop and desktop PCs (in the home) as well – faster startup time, better battery life (on laptops), (the hope for) more/better entertainment apps and more immersive experience using them, an app store / safe install & isolation getting people to install more apps, live tiles reminding people to use those apps, more integrated sharing, all designed to get people using their PCs more often and their phones less often, by reducing the “overhead” of using the PC. Touch and tablets can be seen as a special case of that.
Very well thought out article, and I entirely agree – this ties together my own scattered thoughts perfect. Because of the strange headline that I don’t understand the reference too, I unfortunately cannot share it on Twitter or Facebook because others may also not understand and wonder why I’m linking to a vulgar article.
Using American cultural references does have it’s downside! In the early days of the Saturday Night Live TV show they would include a segment that parodied the 60 Minutes Point/Counter Point political commentary segment. In the real segment James J. Kilpatrick was so nasty towards Shana Alexander that in the SNL parody Dan Akroyd always opens his counter point to Jane Curtain with “Jane you ignorant slut” since that seemed to be what Kilpatrick was implying about Alexander. For Americans, at least those in the Baby Boom generation, the phrase is now generally used to point out Ad Hominem (emotion-based) attacks.
I’m sorry you are uncomfortable linking to the article based on the title.
So with the BYOD trend, are people just using their own devices for specific functions or would they rahter do all their work, if possible on a tablet (with a keyboard?).
If they want to do all their work on a tablet, and if there is such a tablet that will allow them to do all that work, why is it that the enterprise needs to buy and maintain a desktop PC for them?
Of course the answer is in the numbers – how fast will enough people move to using tablets whenever possible at work?
At the moment I think BYOD applies primarily to people who are using a personal device for occasional work use. For example using your iPhone or iPad to do email when you are not in the office is probably the most wide-spread example. Being able to do other things, like a manager approving an expense report, on a personal device is a more complex example since it involves a more general piercing of the corporate firewall and perhaps placing bits on the device. Being able to use a personal device as a full function work computer under some circumstances (e.g., a developer needing to fix a bug while they are on vacation) is another common example. You can scale it from there to the point where the user supplies their own primary computing device, but that is currently a very rare case.
I don’t think tablet use is a matter of “whenever possible” so much as “whenever suitable”, and that’s a hugely different thing. BYOD or not, devices like a Surface Pro with Touch Cover or HP Envy x2, are going to be a lot more suitable for a lot more work scenarios than pure tablets like the Windows RT-based Surface or an iPad. While a lot of users in the BYOD environment might decide they want to focus on the iPad as their primary computing device, I think reality will set in when they actually try it.
Oh, my mistake, Process Monitor is producing some slightly misleading stack traces (maybe it has the wrong dbghelp.dll or something). It’s actually using the (new) Win32 function CreateFile2. CreateFile2 is CreateFileW with slightly reorganized arguments. Both functions do the same thing; they slightly alter the layout of their arguments (again) and then call CreateFileInternal.
CreateFileInternal which then proceeds to call ZwCreateFile.
Process Monitor wasn’t seeing the CreateFileInternal, I guess because it’s not public.
Either way; it’s still going through Win32. And not just kernelbase.dll Win32; it’s using things like shell32.dll, presumably because shell32.dll already had code to manage Known Folders (and maybe isolated storage too).
Thanks for the info, Peter.
Pingback: Where are the 7″ Microsoft tablets? | Hal's (Im)Perfect Vision
I’m a programmer and Windows 7 is my workhorse. I work with three monitors, which is probably just slightly more unusual than a power users two, which is quite mainstream.
I tried for a week to learn and work with Windows 8. My conclusion was that I just couldn’t work with it. It is totally counter intuitive to any workflow that I tried to develop. I gave it a chance and I learned the shortcuts.
I am truly disappointed in how sane people could not have realised just how alienating and hard to use Windows 8 is for productivity. If an alien came down and studied that two systems I am certain that he would give Windows 8 a bad review – it is not just because of what we are all familiar with.
I use VM’s for my actual programming platforms (Ubuntu etc). I also own an iPad, which I love but am slightly critical of, and several Android devices.
Windows 8 is a disaster on a *worse* scale than Vista. I am dumbfounded on how people couldn’t see this. Maybe on a tablet it is great (though I still haven’t seen one in real life), but that is where this interface should have stayed.
Please don’t anyone tell me that I can find the desktop by performing a series of tricks and keyboard combinations. It’s not just that singularly infuriating goose chase, it is a meriad of design decisions that clearly were 100% focused on a tablet experience to the total exclusion of consideration for the desktop. They should have just put FU on the desktop.
I am now, for the first time, seriously considering running a separate drive for Windows and going mainstream with Linux. I see the writing on the wall for Windows, and maybe the sooner the better for me to adjust.
Most of all, I can’t understand why MS are chasing what is fast becoming a duality of a commodity vs niche market. That is a bad strategy. Look at TI, amongst others, getting out of it. The mobile phone market is very close to a commodity market in the component parts.
I am really puzzled and am hoping against hope that I can have, somewhere, a desktop continuing to be like my current Windows 7 + Stardock Fences/Object Desktop – for me that is the best combination for productivity. With Windows 8 my productivity was hindered at every small opportunity.