When it initially launched, Android was meant to be the next great coming of open source innovation. Google’s handpicked manufacturers would contribute to the community, everyone could pull the source code at will, and Google would be the grand maestro of it all. Along the way, it was going to beat Apple, replace Windows, and make everyone a lot of money.

The only problem was, the plan didn’t really work.

There’s not much evidence that Google is making a lot of money off of Android, the Open Handset Alliance is in shambles, and the idea that Android is truly an open source and community effort was thrown out the window a while ago. Yes, Android does serve its purpose in that it is a counterbalance to iOS, but with the mobile landscape in a constant state of flux, who knows how long that will last without economic incentive?

For a bit of history, there are two main reasons that Android hasn’t ended up being the savior that Google wanted it to be at the beginning.

The first is that the association that oversees Android, the Open Handset Alliance, was comprised of companies that directly compete with each other. Not just generally competing “in the technology industry” but in the same exact market, often with the same exact product. With such a close competition, cooperation between companies like HTC, Samsung, and Motorola was a naive dream.

The second is that, while Android may be an open source project according to the legal definition, it isn’t truly open in the same way that Linux and even Chromium are. Instead of truly being a community effort, the Android alliance is a set of brokered deals between OEMs and Google, with revenue, partnerships, and the integration of proprietary services like Google Search being put at the forefront of development. As is normally the case with companies, revenue comes before community.

Essentially, Android wasn’t “open” in the way that Google promised it would be, and that killed its biggest advantage: being an open source project. Since it wasn’t succeeding in an “open” capacity, Google decided that it needed to play hardball.

As business priorities shifted away from creating an open source community and moved towards a more traditional, revenue-focused model, Google and its partners became stuck with Android as an open source project. That’s where the company is today, having an open source project that it needs to maintain in order to keep face, while also wanting to create a better product without the community’s involvement.

What Google wants, and what it needs to do for Android, is to do what Microsoft just announced on Monday. Microsoft unveiled the Surface, the company’s in-house tablet and iPad competitor. The tablet is entirely a Microsoft production, and cuts out the third-party manufacturers in a way that Microsoft has rarely done in the past. Essentially, Microsoft has made it clear that it doesn’t trust its hardware partners to deliver the most sublime experience, and so it is taking things into its own hands.

Microsoft will now have third-party manufacturers creating tablets and computers running Windows in a number of configurations, while Microsoft itself manufactures competing products. It’s a tough pill to swallow for OEMs, but the added marketing power of Microsoft should make the brand itself stronger, which helps all companies.

This hybrid solution of a partnership with manufacturers and in-house hardware means that Microsoft can have the wide scattershot effect that Google currently has with endless Android configurations, while at the same time having the integrated experience of iOS. It’s the best of both worlds, with the entire market being firmly controlled by Microsoft. The company is cherry-picking the parts of its strategy, and it may have a chance at pulling it off successfully.

This strategy has two primary advantages. The first is that Microsoft controls everything that happens with Windows, and makes sure that the third-party manufacturers that do make Windows-enabled computers and tablets do it according to Microsoft’s directions. The second advantage is that Microsoft is able to have a single device it can feature as the pinnacle of Windows — in the Microsoft store, no less — while at the same time having manufacturers design hundreds of configurations for niche use cases.

The hybrid in-house/OEM solution also happens to be the strategy that Google should be pursuing. Yes, the company has purchased Motorola and could use it to build hardware in-house, but there’s no evidence that the two companies will be tightly integrated as of yet.

The best solution for Google is to continue to create the best licensable software for OEMs, while also using its close ties with Motorola to make truly definitive hardware for the Android ecosystem. We shouldn’t be surprised when that happens. The Open Handset Alliance would still get free, up-to-date, and customizable software, while Google continues to widen its margins and grow the ecosystem.

In an odd turn of events, it looks like Microsoft just out-innovated Google in the technology world. Talk about a reversal of roles.