We live in a post-something era. This much, Microsoft is willing to concede; the iPad's thundering success changed the landscape. It has shown that the buyer is willing to imagine a different form factor than the PC commanding her principal information delivery platform. Apple has yet to conquer that platform, but it has fired its third round of volleys and the castle walls have been breached.

For Windows 7 to succeed, it needed to do one thing swiftly and unquestionably: kick Vista to the curb. It did not have to be different from Windows as we had come to know it, to be better than Vista. It needed to be comfortable, almost from the first moment of its existence. Windows 7 met that goal. Windows 8 has a much steeper mountain to climb.

Last September, just prior to the Developer Preview's release, I set some fairly high expectations for what I believed Microsoft would have to accomplish with Windows 8 to assure its success. These things are not going to happen - there won't be "one Windows" that you "get" on your phone, your PC, and your TV as a kind of subscriber (a unified kernel is not quite the same thing); there won't be one screen that you can become familiar with and that's portable across devices, it won't embrace hypervisors or virtual envelopes as the equivalent of devices.

The value proposition, which I've gathered from various sources including Windows Division President Steven Sinofsky speaking to my face, is this: You, the consumer, prefer a rearranged environment that at the very least reminds you of a multitouch device, even if it can't yet be one, and that emulates many of the thrills of using your smartphone. Owning a modern smartphone or tablet reminds you of just how dull and commonplace a PC has become. You would pay money to feel that your PC - whose very name already seems antique - was built in the same decade as your communications device. A marketing campaign that fits the flavor of Windows 8 thus far might begin with someone saying, "All right, all right... you win."

There is no question that the Windows 8 Consumer Preview, launched earlier this month, feels dramatically different from its predecessor. For Windows 8 to be a successful product, the feeling it delivers must be something that the ordinary consumer will want.

So without snark, with zero sarcasm, with eyes facing forward, let's assess the path to victory for Windows 8. We'll start by defining what "victory" means. Microsoft has sold many "successful" products without them being particularly well-liked. Vista was deemed a success for several years before the company allowed its marketers to attribute the word "mistake" to it. Even Windows Me (I bet you forgot) was proclaimed a smash success while it was being sold, though its success was marked by privately recorded surges in sales of the service pack for its predecessor. "Victory" in this context does not mean the usual wave of early adopters.

I'll set the bar comfortably low: Let's call "victory" for Windows 8 a state of customer satisfaction that is at least equivalent to that for Windows 7. For Windows 8 to meet that bar, the following things will need to happen over the course of the following year:

Windows Phone 8 will need to merit a respectable third place in the public mind, completely supplanting BlackBerry. This may be the easiest and most likely objective to achieve, thanks in large part to the rapidly corroding agenda for the BlackBerry platform. If the everyday consumer perceives the PC as yesterday's news, he perceives BlackBerry as last year's.

At least one virally popular, cross-platform app will need to be created for Windows 8 and Windows Phone 8. Right now, there is no solid reason to run Metro as opposed to some other application or Web service on the Desktop. Metro's best single use is as a staging ground for the new Start menu. The fact that Microsoft continues to use the phrase "Metro-style apps" the way happy families in TV commercials refer to "Jell-O brand gelatin" denotes the difficulty Microsoft continues to have in giving the newest of Windows 8's two worlds an embraceable, adoptable identity. For 30 years, operating systems have always been launched with "killer apps" as their booster rockets - and this fact has not changed in the Post-Something Era. Metro, or "Metro-style Gelatin," does not have one.

Intel's "Ivy Bridge" processors can't come a moment too soon. The new Ultrabook form factor requires lower power and longer lifespan in order to eke out a thinner form factor. Consumers will expect a very thin form factor for them to accept the idea of an Ultrabook as opposed to a tablet. Intel's latest reference models of Ivy Bridge laptops demonstrated at CeBIT in Hannover last week feature multitouch screens that rival the latest iPad. Windows 8 will not be perceived by a majority of users as improving the experience for mouse- or touchpad-based PCs. So it has to make the case that an investment in a new Ultrabook will pay off better than an investment in an iPad. Not even the sleekest Sony Vaio models currently on the market capture the imagination like an iPad today. No matter how much Windows 8 improves between now and general release (my thinking is, mid-October), if the machine it runs on feels like a hand-me-down, it won't win over the tablet buyer.

Microsoft needs to catch Apple on the "up-beat" of its cadence. If Apple wanted to ensure that the noise around its latest iPad drowned out any buzz Windows 8 might generate, Tim Cook might have opted to delay that tablet just a few months. That is, unless he has a new iPhone or other enhancement for the early-fall timeframe. If the new iPad is indeed the big Apple product until at least the holidays, then Microsoft may yet have a small window of opportunity - although October would be on the late side. The longer Microsoft waits, the more the almost Pavlovian anticipation for the next iSomething will rise toward a crescendo, and Windows 8 cannot afford to be drowned out.

There needs to be one Microsoft design department finalizing the architecture choices for everything that uses an account. Sure, the folks attending Microsoft conventions are tethered to sixteen different accounts and any number of concurrent identities. But an ordinary fellow who just wants to play the music he already owns does not expect, after tripping through a few meaningless warning screens and Cancel buttons, to have signed up for Xbox Live under an assumed name. I don't own an Xbox and hadn't really planned on it. I wanted to try to play a song, so from the Start menu, I chose Music to see what would happen. I got a menu full of music I didn't own yet. Obviously, this was from Microsoft's Store, so I tried to exit or find a settings screen or locate some hint as to where my own music is located. When I bring up Pictures from Metro, I get my pictures; why, when I bring up Music from Metro, don't I get my music?

In the end, there needs to be one account. Simply by trying to cancel out of Music (the standard way to get out of a Metro app is to pretend you're using Android and leave it running), I found I had signed up for a "gamertag" on Xbox Live, with the name BlabbyScarab5 generated for me because I was evidently too stupid to know I was supposed to sign up as a gamer before I could cancel out of the act of listening to someone else's music.

This is the type of crap that sensible, regular people will not tolerate. When someone at the Best Buy demonstrates an iPad, he's asked to pull up a song. Two pokes, and there's a song. Plug the iPad into a dock, and you have crisp, clear stereo. Consumers will compare an Ultrabook to an iPad. They will ask the Best Buy guy to poke up a song. If the Best Buy guy responds, "Do you want to choose an Xbox Live gamertag to go with your Microsoft Account, or would you prefer to generate a username and avatar at random?" they will walk out of Best Buy and into an Apple Store.

Microsoft is capable of making a comfortable, workable system for a new class of PC. It's done it before, and it's been successful. But more than any company I know, Microsoft is capable of unlearning from its own success. It needs to find the exit button for this pathology and relearn the lessons of Windows 7.