Microsoft needs to be a "mobile-first, cloud-first" organization; it's Satya Nadella's mission statement. But while Microsoft is doing pretty well transitioning to the cloud, this week brought another Microsoft disaster in mobile, and it's one that could imperil the overall Windows 10 strategy.

Microsoft is swerving all over the road so hard that it's impossible to tell where it's going.

Nadella laid off 7,800 people, took a huge write-down, and declared a new mobile strategy yesterday. Microsoft seems to have a new mobile strategy annually. It worked with big-name, first-world partners like Samsung and LG and charged them license fees. Then it focused on Nokia, which it bought, so it became primarily a first-party phone maker. Then it tried to nurture lesser-known, developing-world partners and didn't charge them. It kept its core experiences like Office to itself, and then it deployed them across all the competing major OSes. It tried to appeal to less tech-savvy consumers and then, at MWC this year, said it was going big for business.

So now Nadella is saying that he no longer wants to "grow a standalone phone business" and instead "create a vibrant Windows ecosystem." Okay, sounds like he's backing off from hardware. But wait! He is "committed to our first-party devices," and will fill "three customer segments" with them. (In other words, he's going to do a low-cost Lumia, a midrange business Lumia, and a high-end Lumia.)

This way and that. Whiplash. We're going to make our own phones, but we're not actually interested in selling them. We'll have partners, but we'll be best in the key customer segments. It's a huge mess, and it's a huge mess that looks like the mess Microsoft has been in since Windows Phone 7 launched in 2010, spinning and spinning in the widening gyre.

There's one little light of hope: yes, Windows 10 is going to be much more compellingly integrated across desktop and mobile than Windows 7 and 8 ever were. Continuum and universal apps are a new game. But nobody's going to play that game if there's nothing they really want to play it on.

What Hasn't Microsoft Tried?
As I've reported before, the high-end Lumia 940 doesn't look like a game changer. With a rumored 5.7-inch quad-HD screen and Qualcomm Snapdragon 810 processor, it will be a respectable phone, but Samsung will be coming out with a better Galaxy Note 5 and Apple will have a better-marketed and more app-friendly iPhone 6s Plus at roughly the same time.

Paul Thurrott outlines a Surface-like approach for Windows Phone success: make devices with unique capabilities that inspire third parties to do better. Yes, that would be a good strategy. Microsoft hasn't said it's doing that, and it hasn't shown any ability to execute on that kind of strategy in mobile. It squandered the Lumia 1020's camera advantage, and the 940 is not going to set the world on fire, even if Windows 10 has great, unique features. Meanwhile, Microsoft seems to have either bored or alienated most of the hardware partners it's announced in the past.

For Windows 10 to succeed in a mobile-first, cloud-first context, Windows Mobile 10 either needs to be on compelling hardware people want to own (which the company doesn't seem to be capable of making), or Microsoft needs to break completely free of its own OS and become a provider of cloud services and apps, which Nadella just said it won't do.

I don't have any solutions here, just frustration. Microsoft doesn't seem ready to make great and highly differentiated Windows Phone hardware, to nurture great Windows Phone partners, or to abandon the Windows Phone platform. And yet it wants to be "mobile-first, cloud-first." Can someone over there explain how this is supposed to happen?