Andromeduh? (Premium)

I’ve been trying to set expectations for Andromeda accordingly. But there’s something about this quirky and erstwhile product that makes fans forget the past. And the reality of the situation Microsoft is facing.

Which is this: Microsoft, which has consistently failed in both mobile and hardware, is quietly working on a mobile hardware project called Andromeda that will somehow change the world.

Not helping matters, most of the reporting on Andromeda so far has been both unprofessional and uninformed. And that’s true across the board: Everything written about Andromeda so far has been pure speculation, and that’s true even when a blogger or reporter has received first-hand information from actual sources, as we have. Or by viewing internal Microsoft documentation. Which we also have.

So I’ve tried to maintain this perspective, to be clear about what is and what is not speculation. To not irresponsibly stoke the hopes and dreams of fanboys to drive page views and somehow attach me, or this site, to the story. I find this entire situation to be depressing, frankly. And irresponsible.

But it was with great interest to see, yesterday, that an actual reporter with credentials and decades of experience finally provided both the perspective and the reality that I think has been largely missing from the Andromeda story: As Mehedi discussed yesterday, Mary Jo Foley has finally weighed in on Andromeda and has delivered a sobering report that Andromeda, in her sources’ words, is going “back to the drawing board.”

She writes that the Andromeda OS bits have been pulled from Redstone 5, the version of Windows 10 that Microsoft expects to ship in September 2018. And this means that Andromeda, the quirky, two-screen pocketable mobile device that requires this software support, probably won’t happen until sometime in 2019, if it all.

What I find particularly appealing here—and to be clear, yes, we all love being right—is that the reasons for this delay are pragmatic: “There’s still no compelling reason for Microsoft to come to market with its current iteration of a small, dual-screen mobile device.”

Exactly. This is what I’ve been saying all along. That no matter how “cool” this thing looks, no matter how innovative the hardware design is, none of it matters if the device has no real purpose. And Andromeda makes no sense if it is not backed by a reasonable software and services ecosystem.

Critics, detractors, and, yes, haters have already provided some overly-obvious counterpoints. My favorite: The iPhone wasn’t by a reasonable software and services ecosystem when it arrived in 2007, either, and it changed the world, I’ve been told multiple times. It could happen again.

Yes, it could. But it’s not going to happen in mobile: The next change-the-world moment, the next wave, will happen in a different market. It’s not going to happen with a device that is really just a Windows-powered PC; that ship has sailed. And, sorry folks, but it’s not going to happen to Microsoft, a company that is well-regarded by the enterprise and a non-event to the consumers that are needed to make this device a success.

When it comes to speculation, though, I like to do it through (what I hope is) an educated lens of how today’s Microsoft operates. And that makes Andromeda an iffy proposition.

On the one hand, Satya Nadella’s Microsoft has OK’d vertical market moon shots like HoloLens, a device type that will never be mainstream but speaks to the company’s vision and ability to innovate new hardware form factors.

On the other, Satya Nadella’s Microsoft shit-canned Surface mini at literally the 11th hour because of the same concerns I now have about Andromeda. This is interesting for a number of reasons, but here’s a basic truth: That Surface mini, while absolutely doomed, still would have sold in numbers far exceeding HoloLens over the same time period. It would have made more money too.

How Nadella and his now Terry Myerson-less Senior Leadership Team might vote on Andromeda is an open question. But when you add in the software “scheduling and quality” issues that Ms. Foley’s sources described, a postponement is warranted. So the only real question here is the same one I’ve been asking all along: Should Microsoft even release this thing?

To understand the decision-making process here, it’s important to understand the mentality of the people making the decisions. Microsoft, like its most ardent fans, still feels institutionally that it can make a difference. And there are people at the company who believe this can happen with hardware.

Microsoft has never been successful with hardware. And yet, there are positive signs that cannot be dismissed. Its Surface PC business, for all the controversies around speed to market, technology (USB-C/Thunderbolt 3), and tiny market share, is both aspirational and of high-quality. It’s Xbox consoles are technically superior to the competition. And the next-generation Surface Hub 2? Are you kidding me? Amazing.

These positive developments do suggest that Microsoft could deliver on a unique new mobile device, one that is not quite a phone and not quite a PC. And that this thing might be both beautiful and well-made. Hope springs eternal. And those voices at Microsoft that want the firm to press forward may rise above the noise of logic and reason. It happens.

But the software and services ecosystem issue looms over the entire project, and it should be the deciding factor, I think. If Andromeda is positioned as a new kind of mobile device, it will fail. If it is positioned as a new kind of PC, like Surface Pro … Hm. Within the constrained expectations of this much smaller market, what will obviously be low sales could be a bit more sustainable. Right?

Maybe.

That line of thinking didn’t help the HP Elite x3, of course. That device was really just a phone with no software, but HP tried to market it to businesses as a complete PC system that also happened to be a phone. And it’s hard not to view Andromeda as a slightly more futuristic take on the same idea. But it’s not like the Windows 10 software market has grown much in the two years since HP announced that mistake. Certainly not when it comes to software that might run well on a small mobile device.

I’m also reminded of Steve Jobs’ tortured rationale for the iPad he introduced in 2010. Apple, he said, had asked itself whether there was room for a third category of device, something between a laptop and a smartphone. That device, he intoned, had to be better at key tasks—browsing the web, doing email, sharing photos, watching videos, and so on—than a laptop or a smartphone.

Whether he achieved that goal is debatable. But Apple, which is the richest company on earth because it sells highly-desirable mobile products and software and services that runs on those devices, was able to single-handily create a new market for that third category of devices.

And that means two things for Microsoft and Andromeda today.

First, the question is no longer whether there is room for a third category of device. The question is whether there is room, any room at all, for a fourth category of device, one that would be a hybrid, of sorts, of two of the other categories. Another thing that is sort of a PC. And sort of something else. This time a phone.

Second, with developers already fatigued by supporting too many popular platforms, is there any rationale at all for Microsoft to foist yet another platform on the world at this point in time? Technologies like Xamarin, Flutter, PWAs, and others are working to make the differences less problematic. But the real developer push and focus are on the volume parts of the market. And that’s Android, iOS, and web. Not Windows 10.

“For the record, I think Microsoft is doing the right thing here,” Foley writes, referring to its decision to at least delay Andromeda. “Microsoft would be badly served coming to market with something that’s more a prototype than a well-thought-out finished product, as its competitors could take the concept and beat Microsoft at its own game.”

I couldn’t agree more.

As I note above, there are some signs of life for Microsoft hardware. But this project needs both a clear rationale and a viable supporting ecosystem—plus a reasonable price, which Foley says is another issue with the current design—before it can survive, let alone thrive.

So here’s the bone that I’ll throw to you Windows enthusiasts in the audience, the people who are pining for Microsoft to ship this device because it’s cool-looking and has a Microsoft logo on it. Waiting is the best thing that ever happened to Andromeda because the alternative, today, is to simply kill it. I happen to believe it’s going to fail no matter when it ships. But shipping it this year would guarantee that outcome. Waiting is the right decision.