Has Microsoft Gone Nuts?

The rumors floating out of Microsoft these days are remarkable:

—Microsoft’s Nokia business is working on an Android phone, and Microsoft might let them sell it (link).

—Contrary to its entire business history, Microsoft may give away Windows phone and Windows RT for free (link)

—The Start menu will be restored in at least some versions of Windows 8.1, and it will be possible to run Metro apps in floating windows inside the traditional Windows interface (link)

Those are just rumors, of course, but they’re coming from multiple reporters who have strong ties to Microsoft, which means they’re probably genuine plans or at least serious trial balloons. Taken together, the reports give a picture of a company that’s laudably willing to revisit its assumptions, but that also seems to have lost track of what it’s trying to accomplish. Specifically:

—Making Windows Phone free is a nice idea and would have been a clever response to Android six years ago. Back then, the idea of a free mobile OS was appealing to handset companies, which are at heart surpassingly cheap. Faced with rampant iPhone sales and no good alternative, the phone companies were very willing at that time to give a try to the low-cost Google-backed approach.

But today, Android has huge market momentum, so a phone vendor switching off it would be abandoning most of the available customers, something they are extremely reluctant to do. Besides, getting a free OS is not currently the main concern for most handset companies – the big worry is finding a way not to be crushed by the dominant Android vendor, Samsung. Windows Phone has its own dominant vendor, Nokia, so it has all the negative aspects of Android without the customer base.

Nice idea, Microsoft, but you’re closing the barn door not only after the horses left, but after the barn burned down.

—I think a more interesting licensing strategy for Microsoft would be forking Android: Take an open source version of it and add your own services on top. That’s apparently the idea behind Nokia’s OS plan, in which Nokia would supposedly replace its low-end phone OS, Series 40, with a modified version of Android. Those phones could then tap into the Android app base, making them more attractive to low-end customers. (Heck, they might even be more attractive to high-end customers as well.)

But if that’s the right strategy for the Nokia business unit, it’s also the right strategy for Microsoft’s OS licensing. If you’re giving away Windows Phone for free, the only way you’ll make money from it is through bundled services. You could just as easily bundle those services on a forked version of Android, save the expense of creating and maintaining all the low-level OS plumbing, and get access to the Android customer and app base. That sounds like a much more appealing licensed OS than Windows Phone, even though you still have the problem of Nokia competing with your other licensees.

But that’s not the roadmap we’re hearing from the Microsoft OS team. Instead, they’re talking about creating a single Windows code base that runs across all types of devices, something that’s technically appealing if you’re a Microsoft engineer but thoroughly uninteresting to customers.

So the various parts of Microsoft appear to be working at cross-purposes. The Nokia unit has a nicer mobile OS plan but no intent to license it, and the OS group has a licensed mobile OS with almost no customers. Something’s got to give.

Meanwhile, the rumored changes to Windows 8 are, to me, a mix of sensible ideas and bizarre improvisation. The word is that Microsoft’s going to offer three versions of Windows (link):

A phone/tablet version that runs Metro-style apps. Although I’m sure Microsoft will save money by unifying development, customers don’t care about that. They care about features. Unless there’s some dramatic change in functionality that we haven’t been told about, I think this new product will have as much appeal to licensees and customers as the current version of Windows Phone. So you can think of this as the version of Windows that no one wants.

A “consumer PC” version, which may or may not be able to run existing Windows 32 applications. If it can’t, I think it will sell as well as Windows RT did.

An “enterprise PC” version, which would run today’s Windows 32 applications in addition to “modern” (Metro-style) touch apps. The Start menu would apparently be restored in Win32 compatibility mode, and you’d be able to run Metro-style touch apps in windows inside the Win32 interface.

The only configurations likely to sell in volume are the ones that let customers run their familiar Win32 apps. I think reviving the Start menu in these configurations is smart; it makes it easier for current Windows users to move up to the new OS. That’s such a no-brainer that Microsoft should have done it in the first version of Windows 8. The question is not why they’re doing it but what took them so long.

On the other hand, running Metro apps in a traditional Windows frame is...I don’t even know what word to use. Bizarre? Crazy? I’d even say frightening, because it implies that Microsoft has lost track of why it did Windows 8 in the first place. The idea behind Windows 8 was to build a full-function PC that could also switch to work as a no-compromises tablet. I don’t think a lot of customers actually want that, but at least it’s a clear direction.

Mixing the two metaphors on the same screen completely undercuts Microsoft’s basic idea. Instead of switching between tablet and PC mode, you’re mixing two totally different usage paradigms on the same screen. How does the user know when to touch and when to click? It’s like driving a car that has both a steering wheel and a joystick. Instead of giving users a tablet and a PC that you can jump between, Microsoft is at risk of giving users an awkward combination of the worst of both worlds.

It feels like the people who were responsible for the original Windows 8 vision and strategy have left the scene, replaced by folks who are tactically tweaking the products they inherited, with no sense of where they’re going long term. 

The focus on rationalizing code bases feels symptomatic of this. It’s a sensible thing to do and will cut Microsoft’s costs, but it does nothing to increase demand. In the tech industry we overuse the phrase “rearranging deck chairs on the Titanic,” but in this case it really seems to fit.


What it means

New CEO needed. It’s important to remember that Microsoft is in the middle of its biggest business transition ever. It’s simultaneously getting ready to merge with an enormous Finnish phone company, and hiring its first CEO who wasn’t a cofounder. That sort of situation typically encourages bizarre behavior. For example, groups will try to lobby for their favorite projects by leaking information about them, trying to build up public support that will influence the new CEO.

Think about it, why would Nokia possibly want to leak the news that it’s replacing the OS in its low-end phones? That’s very likely to stall sales of the current phones, driving down revenues. Nokia just made that mistake with Symbian, and now it’s repeating it? Somebody in the Nokia business apparently feels it’s more important to lobby for its OS vision than to protect current sales. It’s a triumph of business politics over short-term business sense.

Another behavior we should expect to see is business managers pushing forward aggressively with their plans, looking to prove that they’re dynamic leaders who don’t need to be replaced. It doesn’t matter if those plans contradict company strategy; the point is to look dynamic. The strategy’s going to change anyway. And what’s Ballmer going to do, fire you?

We should expect to see more odd behavior until MS picks a new CEO. Then it’ll be several months of strategic reviews, followed by ritual bloodletting and reorganization. So Microsoft is likely to continue to be confused for at least the first half of 2014, and that’s assuming they can choose a new CEO quickly, something they may not be able to do to do (link).

Windows Metro = OS/2. The big bet in Windows 8 was that Microsoft could re-ignite sales growth for Windows by tapping into the tablet market. A PC that could also work as a tablet, Microsoft reasoned, would be more attractive to customers than either product alone. I think it’s pretty clear that the Windows 8 bet is failing. Windows 8 is being pre-installed on a lot of PCs, but that’s because Microsoft is pushing it through the OEMs. Microsoft could ship a hamster wrapped in duct tape, label it Windows, and a lot of OEMs would bundle it. What hasn’t happened is Microsoft’s promised explosion in user demand for convertible Windows computers, followed by an explosion in developer activity that might drive future demand.

I think it’s increasingly likely that the tablet interface formerly known as Metro will be more or less stillborn as a development platform. It will linger for a long time as Microsoft’s software for touchscreen devices, but I don’t see it being embraced by the leading-edge developers who can drive new demand to a platform. Instead, it’s kind of the third option for developers who have already built for iOS and Android.

I think most PC users will stick with the traditional Windows interface, most Windows developers will follow them, and most people who want tablets will get iPad or Android or Kindle.

So the challenge for Microsoft’s new CEO is the same one Steve Ballmer has tried and failed to answer for years: Demand for Windows is declining because the platform hasn’t done anything new for a decade, while Microsoft doesn’t control the fast-growing segments in tablets and smartphones. Microsoft tried to use Windows 8 to take over tablets. That failed. What do you do now?

The traditional answer would be to break up the company and try to salvage parts of the business that can grow on their own. It’s the kind of big deal that consulting companies love to recommend and investment firms love to broker. Besides, it looks bold, even though it’s actually the path of least resistance. So I wouldn’t be at all surprised if the new CEO chose that path.

The alternative to a breakup is to actually fix the product problem: to offer new functionality for Windows that’s more attractive than the competition. That would mean new mobile software that’s substantially better than iOS and Android, and/or new PC features that will give people a compelling reason to recommit to personal computing. There are plenty of opportunities to create that sort of innovation (link), but Microsoft doesn’t have a great record as that kind of product leader.

So if Microsoft is to stay together, the new CEO needs to be either a product visionary or know where to find one. I wish them luck.