Tech on the Side: Windows 8 is a flop, and there is little they could have done differently
Apr 17, 2013

Tech on the Side: Windows 8 is a flop, and there is little they could have done differently

The recent release of an IDC report shines a pretty bad light on Microsoft. Windows 8 is selling very poorly with the steepest declines ever in the PC market, and in particular it has entirely failed to gain traction in tablets, with companies like Samsung even pulling their Windows RT products in some places. Especially given the enormous importance of succeeding in tablets, this has caused many commentators to bluntly call Windows 8 a flop. They are probably right. However, I submit in this post that many of the big decisions made by Microsoft were ones where they could not have reasonably acted differently, and that the fact that Windows 8 was a flop was at least in part determined by the evolving market conditions and positioning of other companies. There simply was no clear alternate path to doing pretty close to precisely what they did.

Microsoft's market position:
Microsoft's defining competitive advantage has always been its enormous "moat". That is, Windows is the de facto operating system for desktops and laptops that everyone uses, everyone thus knows how to use Windows, is trained in its use, can be found everywhere, has every program imaginable written for it, etc. And since it has all those things, it is thus the de facto operating system. It is this feedback cycle which has allowed Microsoft to make relatively small evolutions time after time in each iteration of its OS with little worry as any newcomer, even if they were to offer a superior OS, has to get over this enormous moat. Incidentally, Facebook has a similarly powerful moat in that everyone continues to use Facebook because everyone continues to use Facebook; breaking that cycle even by a superior platform will be extremely difficult.

Going into Windows 8, Microsoft had little to fear on the desktop front. Linux, while getting better and more user friendly was still not gaining meaingful market share. Apple desktops and laptops do well, but their growth is limited by their business model. The big new entrant was Chrome, presumably for the ultra low end even if they position themselves differently at times (see the Chromebook Pixel) for example. But for the most part, the Microsoft position on desktops and laptops was reasonably secure. They can hope to provide a better product that quickens the upgrade cycles, but they were going to remain desktop king no matter what. 

The big threat was a shift away from desktops and laptops, where Microsoft faced little threat, to an entirely new form of computing on smartphones and tablets where Microsoft had little traction. Thus the core judgement on the success of Windows 8 should be in whether it was able to move into tablets. While a successful OS might drive some momentum in the desktop space, the desktop market was destined to decline no matter what, and that this happened can only partly be blamed on Microsoft. Given the truly phenomenal success of the iPad, and subsequent Andriod tablets, Microsoft found its position reversed. It had no software for these popular devices, and Apple and Google were quickly growing precisely the kind of moats in things like number of apps that once gave Microsoft its advantage. The only reason this shift happened was because Windows fundamentally does not work on a tablet and need a fundamentally new operating system. 

How Microsoft responded, and had to respond:
So what is Microsoft to do? To keep their advantage, they need to have access to the old way of doing things with all the apps (or programs, as they used to be called) and power that Windows, together with Intel chips, provided. Scrap that and they are starting from ground zero. So you had to keep the old system, the old "moat". They also had to make a completely new OS designed for touch screens, since the old system simply didn't work in this domain. So some combination of the old and new was simply mandatory, the only question is how the combination works.

The problem turned out to be a hardware one, as much as a software one. The kind of chips capable of running "full" windows need to be fast and this needs tonnes of power and thus resulted in bulky, battery filled cases, too bulky for a tablet, or computers with horrible battery life. A cheap, lightweight tablet with a decent battery life just could not run full windows from the hardware standpoint. Perhaps in another year this will be possible, but not in 2012. So Microsoft had no choice but to release one version that only did the tablet centric stuff. Something that could run on a cheap ARM mobile chip, the type of chips that can't run photoshop. Hence Windows RT.

So Microsoft had no choice but to have Windows be both tablet and desktop centric, and no choice but to release a completely tablet version. So the next question is whether they should have made the other version be the weird combo of a tablet side (or Metro as it used it be called) and the original desktop side. Could they instead have done what Apple does, with two entirely separate and incompatible operating systems?

Again, the answer is no, they were forced to do what they did. If it was just Windows RT, a satisfactory ARM tablet OS, and a separate Windows 8 desktop that was a small upgrade of Windows 7, there is no justification to jump from Apple and Andriod to Microsoft in the tablet space. While they needed Windows RT because of the hardware problems, they had no ability to compete in tablets if they just had Windows RT. The advantage of Microsoft is the legacy moat, if you don't have that, you are left fighting a tablet war from scratch against the stalwarts. The other two had enormous app libraries and install bases, enormous popularity and support, and there would then be little to no justification for someone to jump ship to the windows tablet OS. Andriod in specific is being offered with a less than free model (see below) and thus Windows RT would have to offer enormous benefits to be able to justify itself. It is a good tablet OS, perhaps a better tablet OS, but it can be expected to have - and did have - trouble getting over the moat by itself. By creating an OS that works as it does on the Surface Pro, they present a legitimate argument and appeal by having the needed tablet centered OS, with the ability to run all the windows legacy apps which is so necessary.

Now that the split of having an ARM tablet without legacy windows functionality and an Intel tablet with it was established, the only remaining question is what to do with the desktop side of things? One is hardly hurt by having the Metro tablet-centric stuff in the Windows 8 OS. I never use it, it is just there, but I am not hurt and in fact slightly prefer the superior start button. In order to preserve continuity, and in order to appeal to all the touch screen hybrid devices that are not just tablets, it makes sense to include both functionality in the desktop etc versions.

The last question, perhaps, is whether the legacy and Metro components should be as segregated as they are or whether it should have been more tightly integrated. This point is a bit more debatable, but I think what they did is the best option. In a normal desktop situation  one is not hurt by having the touch-centric stuff in the Metro UI available, it just might not be used. Most seem to prefer just using it as if it was Windows 7. In a touch configuration, none of the legacy software was going to make much sense in a touch interface anyways, so you can't cross it over. Any suggestion I have seen on how to more tightly integrate the two very different favors of windows 8 into a single thing are generally very bad. So I think Microsoft's decision was correct and forced here too.

The main point here is that the major decisions in Windows 8 (to make a tablet-centric OS, to make an ARM version of that and an Intel version, to make the Intel versions be this combination of the two, and have that be only loosely integrated) were all decision that are largely forced. They had little choice but to do what they did. As the market numbers indicate, it didn't work out great, and Microsoft really lost out on the tablet market in particular, even as the desktop market declines across the board. But just because it didn't work out great, doesn't mean there was any easily identifiable thing they could have done differently. This was just how it played out given the market forces and company positions.

One might object about this or that detail of Windows 8. Of course there are many details Microsoft could have done differently, some of which would have made it a more desirable and successful product. I don't mean to take away from all the good commentary done by others in this direction. Focusing on the details, however, and all the little ways Microsoft could have been more innovative and successful can, at times, miss that the broader picture and big decisions was largely determined for them.

It's all about the business models:
The old clash between Microsoft and Apple in the desktop computing market was primarily a contest of business models. Apple created software exclusively for its own hardware devices and sold a complete package. Microsoft let any hardware company use its software. They could charge a high premium for doing so based on the strength of their moat creating a de facto monopoly outside of Apple. Today, that business model difference remains as Apple pushed into iPhones and iPads. Apple creates an entire integrated platform of software, hardware and services and tries to justify its amazingly high premiums based on the appeal of this platform. Microsoft today keeps its old business model with Windows 8, letting any OEM use it.

Google's biggest innovation, however, was to entirely transform the business model in a way that makes the old difference between Apple and Microsoft look trivial in comparison. Instead of charging high premiums, Andriod operates on a "less than free" business model. For a company to use Andriod, they not only don't have to pay for it (ie free), they get to share in revenues generated by the OS (so less than free). Of course, a company like Samsung has some development costs as well, particularly as they try to add special differentiating features and skins over top of vanilla Andriod, but the basic less than free model nonetheless applies. It is a risky strategy in the sense of opportunity costs. Android is wildly successful, and probably wouldn't have been as successful as fast were it not for the less than free business model, but one might argue that Google could have made vastly more money if they had charged Microsoft prices for Andriod. Rightly or wrongly, the difference between the three companies is primarily one of business models.

Incidentally, many will suggest that Andriod is a copy of iOS, and that many of the ten inch tablets are clearly copying the ideas behind the iPad. Whether any actual patents have been violated (or what patent law should actually be) is a separate question, but I think the general idea that Google and Samsung and the rest did certainly attempt to emulate Apple is unquestionable. This isn't a bad thing, in a market the good ideas survive and coalesce as companies identify what works and then everyone is doing that. The big innovation, however, was not in difference between the OS or the hardware from the OEMs, it was the business model and in that sense Google didn't copy a thing.

On innovation:
Has Microsoft been innovative with Windows 8? Partially. It recognized the fairly obvious fact that they had to make a tablet centric OS, and went and made one in more or less the only reasonable way. Windows 8 is certainly a big change from its predecessor in this way. But it was made in the context of keeping an identical business model in a changing time, and as argued above, most of the decisions were forced on them. This is not necessarily bad, they had an amazing business model for a long time and it shouldn't be changed just for the sake of change. Apple, for instance, kept its old business model and just translated it into what it did for phones and tablets and experienced remarkable success. However, Microsoft simply made the correct moves possible given the existing market. Apple genuinely took a huge risk in the development of the iPhone and iPad. Google genuinely took a huge risk in the development of Andriod, which while not truly innovative in itself was very innovative in its business model. Microsoft simply hasn't done something to shake up and change the market, merely reacted correctly given the market as it has been defined, largely by others. Microsoft's mistake was not being the innovator to create the tablet (and phone) market that Apple and Google made possible. After that, everything was more or less determined.

Thoughts on this post? Comment below!

Share this post:

Tweet It! Facebook Add Feed Reddit! Digg It! Stumble Delicious Follow

6 comments:

CQ said...

I disagree. As a self-organized user, I'd notice how Windows had consistently bloated in a GUI sense from 98 to 7. Just doing the basics - like a (very) simple manual editing of a program bar submenu(s) became near impossible.
I use a variety of programs (still in XP) without the keyboard input. For Win8, there should have been a Screen Touch modification applied for the cursor mouse input. Thicker desktop row bar items were all that were really needed. There have long been onscreen keyboard programs (pre-touch). Any number of later and/or snazzy GUI updates are also installed - except as freely programmed by others.
As for legacy, if older programs really can't run with new hardware and underlying processors, then there wouldn't also be emulation modes and programs to run them. Microsoft forces the obsolescence issue. Even with 95 and the switch to 32-bit, they named a single catch-all subdirectory Program_Files, with both a space and a longname. With the Fat32 hard drive system they 'later' implemented its maximum size limit to push for that general change. Even Solitaire suffers when they recreate a new general card.dll file to the exact same file name as its prior version.

bazie said...

If I understand you correctly, your proposition is that they should have just released the classic style windows desktop with a sort of touch sensitive overlay to it (thicker bars, touch modification to mouse input, etc), and not tried to create a separate touch-centric OS? I am rather skeptical that this could be made to be an optimal operating system for a tablet that compete with the ease of use of iOS/Andriod, even if it had vastly more functionality. I submit that the interfaces are so sufficiently different that one truly does need an entirely new interface designed to work for touch, and that a "Normal windows + touch tweaks" is never going to be optimal.

I will also note that several of your points I don't necessarily agree or disagree with, they are just smaller points than the rather big scale comments I was trying to get at in this post. As in there could very well be many different smaller things they could do differently and should have done differently, even if the really big decisions were forced as I suggest.

Anonymous said...
This comment has been removed by a blog administrator.
CQ said...

Also of note is that Microsoft HAD a XP Tablet o/s in 2005. They also had a 'Media Centre XP' and a 'Home XP' and a 'Professional XP'. Instead of expanding a customer's user experience, parcelling everything out restricts it. A clean set-up option would have sufficed.
If it can reboot into a Safe Mode; maybe a setup option could have allowed a modest mid-range of Lower Power Modes? A classic style might have had a keyboard attachment boot check for whether row bars and icons are initially doubled/halved (or sized to a second setting). Instead is an entirely separate o/s GUI called Metro and not called Metro, with apps that apparently can run and apps that cannot.
When I was 'installing' a few tiny program items on another's laptop machine with Win 7 'Home', I accepted its start-up with Windows option. It was a very simple little thing. The machine next re-booted into a forced, unknown 90 minute diagnostic evaluation. The specific item was correctible with a 2 second manual edit. How many customers, after purchase, would have discovered that buying a lesser O/S version meant that various 'smaller things' could NEVER be done.
When I look at Win 8 (from indirect reviews and ads etc.), whether with updated and new or legacy apps/programs, I see more of what I cannot do, rather than what I still can do.

Elipsis said...

First of all I think it's important to note that you have a typo in the first sentence of this article. I want you to be shamed and then correct it for me.

Secondly, I completely disagree with most of what you've said. Microsoft had plenty of choices, and what they did was choose to shovel a tablet operating system into their desktop product. It's the misguided "Windows Everywhere" mentality that lead them to the erroneous conclusion that if the just cram this loathsome interface into the faces of their users, they will learn to love it anyway.

The correct maneuver would have been "Microsoft Everywhere". Keep Windows 8 as a lovely high functioning desktop operating system without all of the wasted screen real-estate and obfuscated functionality that the tablet environment demands. Separately, release a product called "Microsoft Metro" and let that be a marginal success for the tablet. Desktop users and mobile users have different needs, but Microsoft tried to use ~market forces~ to brute force their way through the need for any specialized design considerations. It didn't work mostly because it was an awful idea.

bazie said...

It seems to me that this is largely what they did. Namely they kept the legacy desktop more or less entirely as is only occasionally leaving it for Metro with a (slightly improved, imo) app launcher and the like. And then they made an entirely separate touch centric OS, in RT having no desktop at all.

I completely agree that desktop and mobile have different needs and need different OSes, which is why I suggested Microsoft SHOULD make two, and to keep them largely separate from each other. However, in the interest of the range of higher end tablet and hybrid devices, they ship both of them "together" but together is a rather loose sense. I do effectively all my computing entirely in the (marginally improved) legacy desktop and almost never leave. If someone likes or wants to use Metro, or if they get a touch device in the future, or like a particular app on a secondary touch device they want to use on the desktop, it is all there for them. But it isn't mandatory.

I certainly think the higher end tablet/hybrid devices need to have both. Microsoft's biggest advantage was in the size and familiarity of its legacy desktop, and while it doesn't work for touch it needs to be at least available on touch devices in some form if they can ever hope to have an advantage in this space. I think the main point of contentions seems to be that for a desktop version, there should have been a way to shut metro out entirely, and not shut it out 95%+ the way I do. I suspect from a standpoint of growing their business, having both available in the loosely integrated sense is going to drive familiarity of Metro and thus of Windows tablets in the future and is a good business decision, but this is getting to be a much more minor point in my view.

Post a Comment

Frequent Topics: