A leading research firm suggests that Windows is on track to lose relevancy in the years ahead if it doesnt make a major course correction.
As Gartner analyst Neil MacDonald spoke with a roomful of IT managers about the future of Windows, he asked two key questions. To learn how these tech professionals view Windows, he requested a show of hands:
First, he asked how many in the room felt that the current Windows course and speed is sustainable: Is it tenable, is it something we can actually live with?
No one raised their hand, MacDonald tells me.
Then, given that response, he asked: is radical change is necessary?
And half the room raised their hand.
MacDonald, who along with Gartner analyst Michael Silver authored Windows is Collapsing: How What Comes Next Will Improve, wasnt surprised by the robust skepticism.
As he tells it, dissatisfaction with the industry-dominant Windows OS is deep and pervasive. We hear this from clients every day. Same types of problems: too hard to secure, too hard to manage. Struggling with the upgrades to Windows Vista its painful.
The core problem, MacDonald says, is that Windows gargantuan, monolithic code base makes it unwieldy, less able to adapt to current demands. The baggage of its long legacy makes it far from quick and nimble in todays world.
When Windows was first designed, being connected was the exception; now being disconnected is the exception. When Windows was developed, malware was largely unknown and the anti-virus industry hardly existed. Now we realize that theres a lot of bad guys trying to get our stuff. And security should be baked in from the very beginning, not bolted on as an afterthought, he says.
To remain a top player in the years ahead, Microsoft needs to change considerably. What were saying is, its time for Microsoft to do something radical, to go back to the drawing board, to reinvent the operating system in an era when network computing is nearly ubiquitous.
The Ideal Windows OS
An ideal Windows OS would be modular, targeted for the specific environment it lives in. Not one sprawling code base, but different versions that are fully optimized for different machines.
MacDonald points to the example of Apple, whose OS X system fits perfectly in the limited world of mobile devices. Consequently, the iPhone grabbed hold of the mobile market in a remarkably short period.
Thats the benefit of having an adaptable operating system. You can thin it down as needed which is not something you can do today with Windows.
But Apple, of course, has a completely different business model. Its software and hardware are fully married, as opposed to Windows, which sits on top of commoditized machines. Is MacDonald actually saying that Microsoft do the unthinkable offer a hardware-software combination in the PC world?
Hes not ruling it out. Has Apple, with its innovation, shown that you can produce a better experience for users if you do the hardware and software yourself? he asks.
Microsoft insists that this isnt the business model its wants, yet the companys Xbox 360 benefits from it. There theyve shown: when Microsoft does hardware and software, they do something cool. Microsofts Zune also offers a software-hardware synthesis, though with less success than the competing iPod.
Next page: Will Microsoft Change Before the Tipping Point?
One Size Fits All?
When MacDonald refers to the need for Microsoft to develop a modular approach, hes not referring to, say, the umpteen versions of Vista. The list
is long: Ultimate, Home Premium, Home Basic, Business, Enterprise.
Thats a fairly superficial top down packaging exercise, he says. That had gotten a little ridiculous.
The real challenge goes deeper. What the company is struggling with is: how do you make a business model out of something thats as fluid as todays computing landscape? A world in which a legion of mobile devices communicates with an army of desktops, which communicates with a heterogeneous data center?
In this world, a single, lumbering code base first conceived of a generation ago necessarily struggles.
The possible changes could take many forms. For instance: Maybe they consider forking the enterprise code base from the consumer code base, MacDonald says. So that enterprises ultimately get something thats more manageable and customizable to what they need. If one set of workers only runs a specific set of applications, why cant I thin it down and configure it much more like a kiosk?
This one size fits all its designed for everyone, but works perfectly well for no one because its trying to be all things to all people.
Will Microsoft Change Before the Tipping Point?
Although MacDonald (and the IT managers he addressed) feel that Windows needs a course correction, its unclear whether Redmond is of the same opinion.
Microsoft has the ability, if it felt threatened, to make these changes, he says. However, There is currently nothing serious that threatens their position on the desktop, or their position with Office, and therefore they are not highly motivated to embark on such radical innovation.
While the company feels some pressure to change, its the depth of the change thats in question. Microsoft is aware of the trends that we talked about, they acknowledge that the trends exist, MacDonald says. They disagree about whether radical change is necessary
I think they feel they can contain this, with continued incremental improvements in the modularity of the underlying operating system, repackaging different SKUs.
But I think theres a disconnect between Microsoft and its customers on how bad its gotten.
The difficulty, in essence, is one of legacy business model. I think the problem is, they hold themselves hostage to the business model, where they have to ship with the hardware, and anything that breaks up this monolithic stack is a threat.
So the next major release, Windows 7, is on track to be more of the same incremental changes.
Yet the company may not have unlimited time to change course. The sands are shifting underneath its feet. Our projection is, 2011, theres a tipping point where more organization have applications that are designed to be OS agnostic. And it really does change the equation, so backward compatibility is not the big issue that it was, because were trying to wean ourselves slowly off Win32 apps.
Next page: Who Gains if the Giant Falters?
Who Gains if the Giant Falters?
If indeed Microsoft were to suffer in the marketplace, at least three players would benefit:
Google: The search giant has done something deeply intelligent by developing its Android mobile phone platform. Its smart that they go first for the mobile market, and grow up into ultra-mobile devices, MacDonald says. And ultimately, why not get into the more general PC environment? In the future (and even now) the line is blurring between mobile computing and the desktop PC its all one environment. So entering the mobile arena first (which is not Windows strength) is savvy.
Apple: Apple continues to grow in enterprise market share, he says.
Linux: You cant rule out desktop Linux. It continues to evolve and get better, MacDonald says. OpenOffice, a free, open source alternative to leading Microsoft cash cow Office, keeps on improving, and is already perfectly suitable for most uses, he says.
Advice for Enterprise Users
MacDonald and Silvers presentation offers a number of recommendations for enterprise clients as they consider their future relationship with Windows:
On an immediate basis, take a look at the extent of Windows OS instances in your infrastructure. Create a forecast for your tipping point, at which youll likely become OS-agnostic.
Over the next year, take a look at where virtualization technologies and OS-independent applications can give you a competitive or cost advantage. At the same time, MacDonald recommends proceeding with Vista deployments; even if Microsoft were to change course, it would be a couple years down the road.
Longer term, keep track of Microsofts road map. Based on how satisfied you are, vote with your wallet, they advise. Dont hesitate to consider alternative computing models. And if you want to extend the life of your current Windows instances, use virtualization to squeeze more bang for your buck.