This morning, Microsoft and AT&T announced the first smartphones based on Windows Phone 7, ushering in a new era for the software giant's mobile efforts. T-Mobile is also expected to announce devices later in the day, and in a separate UK-based event, Microsoft and its partners will announce Windows Phone 7 devices for international markets.

For Microsoft, the stakes are high. Although the company has dominated the traditional PC market, it has stumbled badly in the growing market for highly mobile and connected devices. Its original system for this market, Windows Mobile, was strong enough to beat Palm years ago, but it has faltered in the face of dramatically superior competition from Apple's iPhone, Google's Android, and Research in Motion's (RIM's) BlackBerry.

So Microsoft started over. Windows Phone 7 isn't a new version of Windows Mobile but is instead a completely new system, built from scratch and designed to meet the needs of tomorrow's market—not yesterday's. Windows Phone was also designed to address some head-scratching limitations in the iPhone and Android, which are widely seen as the most important mobile platforms going forward.

These limitations included an over-emphasis on the apps model, in which each function of the phone requires another app. In Windows Phone, Microsoft has created an extensible model of experiences, called hubs, which allow related functions from multiple sources—like digital photos—to be enjoyed from a single experience. On the iPhone, for example, users need to manage multiple photo apps, remembering which one does what. On Windows Phone, these experiences work the way you do, and can occur in a single, cohesive experience.

Microsoft has also engineered Windows Phone to be fairly PC antagonistic, which could surprise and confuse users. Unlike most mobile devices, Windows Phone 7 doesn't require any PC connectivity and indeed does very little to integrate or sync with PCs and PC-based applications. It even ignores Microsoft-centric PC solutions like Outlook and the Windows photo-acquisition software. Instead, Windows Phone is aimed squarely at the future, integrating deeply with Microsoft and third-party online services almost exclusively.

Whether this kind of unexpected innovation resonates with consumers remains to be seen, of course. But the timing couldn't be better for dramatic change. Microsoft's current mobile platform, Windows Mobile, has been hemorrhaging market share, dropping from 9 percent last year to just 5 percent this year. Meanwhile, Google Android has surged from 2 percent share to 17 in the same time frame.

And smartphones are big business, with total annual unit sales closing in on those of the PC industry. This year, the industry will sell about 270 million smartphones, up 56 percent, as consumers upgrade far more aggressively and frequently than they do with PCs. (On the PC side of the fence, the industry is expected to sell almost 370 million units this year. That's up 19 percent from 2009.) Annual sales of smartphones could hit 900 million units within five years.

Microsoft could be helped by some competitor mistakes (and hubris). Apple's iPhone 4 has continued to sell well, but it's been a PR disaster for the company with its multiple hardware defects, including a faulty antenna and proximity sensor. Apple's handling of the gaffes—the company doesn't handle criticism well—has proven even more problematic, with Apple basically wishing away the problems without actually solving them. In the case of the antenna, which often doesn't work when the iPhone 4 is held normally, Apple simply recommended that its customers hold the device differently.

And Android is pushing toward the low end of the market with an online store that's more like an illegal sidewalk sale than a respectable retail outlet; Android's Marketplace is marred by intellectual-property abuses and other low-quality wares. It's unclear how a company as big and powerful as Google has escaped any legal attention for these transgressions.

My expectation is that Windows Phone will find a seat at the smartphone table alongside iPhone and Android, both in the United States and internationally. The key for Microsoft is that it is splitting the difference between these two platforms, offering a curated and high-quality apps platform like that found on iPhone but sold via a plethora of devices from multiple carriers, like Android. By poaching the best parts of each company's strategies and avoiding their mistakes, Microsoft should be able to establish a third viable modern smartphone platform. It helps that the system the company has created is genuinely innovative and fun to use, as well.

It should be noted that the number-one smartphone platform in the world is still Nokia, whereas the number-one player in the United States is RIM. But neither has what it takes, I believe, to be viable in the long term, and Nokia in particular is going to have to start over from scratch, as did Microsoft, if it intends to even be considered part of the conversation: Most of the so-called smartphones it sells today are nothing of the kind and are more akin to "feature phones."

I'm in New York today to cover the announcement and related launch events. I'll be blogging about these events from my Windows Phone Secrets blog.