FANTASY M&A —The buzz is all about Microsoft, or possibly Google, taking a stake in Facebook, the popular social network, at a lofty valuation as high as $15 billion. But the logic of those deals is driven by advertising — the more targeted, the better. But what, exactly, are advertisers hoping to target, and why? Besides crude demographics and geographies, the most logical hooks for ads are Facebook users' expressed preferences — the books, music, and movies they're increasingly listing on their profiles. And who has the best data on what consumers will buy? Why, Amazon.com, of course. The logic of a combination — a merger of the two giant databases of consumer preferences is, at least on the surface, compelling.

The companies, in fact, are such natural partners that Facebook CEO Mark Zuckerberg invited Amazon.com to be one of the first to develop an application for Facebook — for book reviews, naturally, complete with one-click buying. With Amazon's existing video store and today's launch of an online music store selling MP3 files, the opportunities for direct, hypertargeted selling only grow.

As well, both companies see their future in building platforms that allow developers to build Web-based applications on top of their websites. Amazon is, arguably, ahead of Facebook in its efforts, despite the latter's buzz, and an e-commerce element, allowing application developers to connect to users and ring up sales would make Facebook's platform more compelling.

What makes this deal fantastically implausible, of course, are the financials. Amazon's market capitalization today is around $37 billion; to take over Facebook, adding a control premium on top of its mooted $15 billion value, seems beyond Jeff Bezos's capacities. As well, Amazon.com remains saddled with physical infrastructure built during the boom; that dreaded brick-and-mortar is a deterrent to investors who favor Facebook's virtual model.

But a deal is not completely unthinkable. Amazon has already taken technical steps to open up its warehouses to other merchants. Could it complete that work, and spin off or sell its fulfillment systems, leaving it as a pure Web business? And could that smaller but more profitable operation merge with Facebook?

Succession might play a factor. Zuckerberg, who has resisted selling, seems unlikely to give up control even for a large stake in a combined company. But Jeff Bezos, who has held on to the company he founded, might view Zuckerberg a worthy successor, who could prosper under his tutelage. Their visions for their companies' futures as software platforms, after all, are eerily similar. And with Facebook's valuation up tenfold in a year, anything seems possible. So why not contemplate the improbable?