(Business 2.0) – Let's consider a few fun stats. First, broadband has reached nearly 39 percent of all Internet-connected households and is expected to be in 79 percent in five years. Next, research firm In-Stat/MDR predicts that corporate spending on broadband content-delivery services, essential to high-quality video streaming on the Web, will almost triple by 2007. Third, the vast majority (72 percent) of work connections are high-speed. Throw in the forecast that TiVo-like devices will be in 20 percent of American homes (and certainly a higher percentage of wealthy ones) by 2007. And finally, 99 percent of homes with incomes over $100,000 have Net access, and it's a safe bet that most of those have broadband connections.

Add it all together and here's what it means: A massive broadband marketing opportunity is upon us--and, as a recent Yankee Group report concludes, the 30-second TV ad will soon lose top billing as our most valuable marketing vehicle. What will replace it? Web-based video advertising.

This form of advertising will be vastly different from its TV ancestors: It will be measurable, trackable, and targetable to the nth degree of demographics. And to the extent that a given audience member cares to invite an advertiser into his or her private database of likes, dislikes, and desires (a profile on Amazon or Yahoo, for instance), marketers will be able to target specific intent to buy--and psychographics to boot.

In essence, Web-based television advertising will be grafted onto the already booming market for streaming video on the Web, and instead of buying mass-market television shows (content, for lack of a better word), marketers will soon learn to buy pure demographics and consumer intent, courtesy of Web-based networks like MSN, AOL, and Yahoo; preference engines like those in place at retailers like Amazon; and new distribution channels like TiVo.

Yes, I'm on my hobbyhorse of content over intent again, but this is worth repeating: Marketers are in a slow, denial-laden shift from buying content-attached audiences, like those of TV shows, to buying intent-attached audiences, like those of search engines and personal video recorders. But imagine what happens when the two merge: Advertisers can buy the audience they want, just when the advertisers are most wanted by the audience.

I admit that the idea of television advertising migrating to the Web has always been a bit suspect. After all, the Web was supposed to provide a targeted, one-to-one medium that would far surpass the relatively blunt instrument of the 30-second TV spot. What's interesting is that it need not be an either-or proposition. It's still flawed, but video on the Net is wildly popular--MSNBC alone streams between 14 million and 25 million clips a month on average (and 87 million in the first four weeks of the Iraq war). I'll bet there's a decent business selling that streaming-media audience to advertisers.

If you're looking for evidence, consider the case of MSNBC.com. The Microsoft-operated site has sold out its inventory of 15-second advertising spots, which run adjacent to many of the news clips it streams each month. These clips are precursors of the multiple forms of marketing-driven video we will one day see on the Web--the first fish with feet, so to speak. And because it can tie in to Microsoft's Passport registration technology, MSNBC.com can (and will) someday link that ad inventory to specific demographics or audience preferences. Advertisers that buy these spots are not buying shows; they're buying pure audience, as well as specific knowledge of that audience's intents. And that's the beginning of something revolutionary.

There are any number of problems with selling television advertising on the Web, of course. Will Web surfers accept it? Will they hit the fast-forward button (or will they be able to)? Will advertisers see the Web as a valuable medium? Can they even buy it in a way that makes sense to them--using the same reach and frequency metrics they apply to buying television and print? Can they let go of their decades-long dependence on the up-front and learn to love the Web?

I'd bet that the answer to all these questions is yes. But don't take my word for it. We're going to find out sooner than most people think.

John Battelle directs the business reporting program at the University of California at Berkeley's Graduate School of Journalism.