Microsoft, Back on Track (Premium)

So what did I think of Build 2018?

It got off to a tough start, for sure. The high-mindedness of Satya Nadella’s vision fell flat for many of the Build attendees I spoke with this week. It did so for me as well: While it’s hard to argue against his pro-accessibility and pro-privacy stance, the case could have been made in a far more expedient fashion. This part of the keynote should have been shorter, but it also should have come at the conclusion.

But if you can get past that, Build 2018 fulfilled the show’s central goal, which dates back to previous Microsoft developer events like PDC and WinHEC as well. It provided developers and decision makers with marching orders for the near- and long-term future, and provided insights about where Microsoft believes things are headed.

And unlike the previous few Build events, that direction is true.

Think about it. The Universal Windows Platform (UWP) is an abject failure, as is the cross-devices promise of Windows 10, which remains relevant on PCs only. The firm has no mobile platform to flout, so it has mostly missed a generation, and its developer base has stuck with “legacy”—let’s say “proven”—technologies as a result. Consumers? That market remains a theory to those in Redmond.

So instead of flouting consumer features that no one will ever use—and that, more important, do not benefit developers in the slightest—Microsoft has instead moved decisively to embrace a new generation of ubiquitous/ambient computing in which it can play a dominant role, if all goes well. And it has finally stopped pretending that its developer base will simply adopt new, unproven technologies just because Microsoft made them.

Perhaps not coincidentally, the software giant has also recast previous consumer-focused technologies for its core commercial customer base, making them more relevant to developers.

For example, the AI, bot, cognition, and speech services that back Cortana, the digital personal assistant that debuted in Windows phone, of all places, now play major roles as Azure services. HoloLens and Mixed Reality are recast from niche solutions for individuals into visualization and recognition platforms for meetings and first-line workers. And even Kinect, one of Microsoft’s biggest consumer failures of all, is brought back to the dead as a sensor set that makes sense in Internet of Things (IoT) implementations.

And then there’s Windows 10.

As I noted previously, Microsoft is slowing down the addition of new consumer features to a system that simply does not have an engaged user base. Instead, it is doing what it should have always done: Focusing on productivity. The handful of features it showed off at Build—Timeline with cross-device compatibility, Sets, and Adaptive Cards—are firmly rooted in their productivity aims, and not in the flashy nonsense that marked the previous few Builds.

That Windows 10 has been demoted to a component of Microsoft 365 will be alarming to some. But it makes sense: Microsoft 365 is the future of Office 365, which is itself the successor to Office, which was itself the successor to Windows as Microsoft’s primary revenues generator. Windows is infrastructure. We don’t “use” Windows. We use Windows applications to get work done.

Still, you might think that Joe B’s Microsoft 365 keynote was a disjointed mess of most disparate technologies and features across several different platforms. But there is, in fact, one cohesive signal running through almost all of the newly announced features. They’re data points for the Microsoft Graph, which we might view as the heart of Microsoft 365.

Conceptualizing the Microsoft Graph can be difficult. (It is for me.) But this platform, which was once limited only to Office 365 enterprise data locked within a single organization, has been expanded to include a wide variety of disparate data sources, and it can be used to build truly intelligent apps and services.

Microsoft is also working to bring technologies that were previously limited to UWP to more developers. Yes, these features will still only work with Windows 10, but with an audience of nearly 700 million, I’m not sure that “limited” is the right term. The more important point is that Microsoft’s developer base has stuck with WinForms, WPF, and other non-UWP technologies. And they’ve been largely left behind as Microsoft put its efforts elsewhere. That’s changing, and Microsoft is once again meeting its developers where they are.

But Microsoft’s most important developer initiatives, I think, are tied to its “intelligent cloud and intelligent edge” efforts. This speaks to a future in which ubiquitous computing makes smartphone addiction look cute and to a potential growth market for both Microsoft and its developer base.

That future is unwritten, of course. But there are only a handful of companies who can deliver on a complete platform as Microsoft has done. And so far, none have.

Likewise, only a handful of companies have the capacity—really, the potential capacity—to both store and process the reams of data this coming wave will require. Here, Microsoft does have some competition, of course, primarily AWS. But it is still a major player in a market in which there will only be a few major players. The future is there for the taking.

In my darkest moments, I complain about the “nonsense” that Microsoft has added to Windows 10 over the past few years, and about the lack of direction that these mistakes seem to indicate. But Build 2018 served as a wake-up call that Microsoft could put that behind them and deliver on a more realistic and even exciting future for itself, its developers, and its customers.

I will still wrestle with the strange emotional gap that was made by the firm’s recent demotion of Windows. I did, after all, grow up in a world that Microsoft is now busy dismantling. But it’s doing so for the right reasons. And the direction it has charted is, I think, the right one.

Quibble all you want about the too-lengthy day one keynote. (I did.) But Microsoft hit it out of the park with Build 2018.