• by Aaron Goldman , Op-Ed Contributor, July 26, 2006

I recently came across a paper written by Google and the Center for Neural Consumption at Hebrew University describing "social and interactive television applications" that allow your computer to listen to your television, determine what you're watching, and display relevant Web applications.

Imagine watching Rachael Ray on the Food Network and having a Web site pop up with the recipes she's using, along with a chat room for people who have cooked that particular dish.

Now we're talking--er, listening.

The paper outlines a methodology by which a computer can pick up and decipher the audio signal from a TV. In turn, the authors introduce four potential Web applications: "personalized content layers, ad-hoc social communities, real-time popularity ratings and virtual media library services."

And how would these applications be monetized? You guessed itthrough audio keyword sponsored listings.

Seems to be a no-brainer to me. I'm sure companies like Peapod would salivate over the opportunity to reach people watching that Food Network show with an ad to "Buy All the Ingredients Here."

But, as with any new technology that promises advanced ad-targeting features, the biggest question marketers will want to know is: can it scale?

Well, we already know that TV and Internet are often consumed simultaneously16 percent of all Web use occurs while viewing TV, per the Online Publishers Association/Ball State Study released last month. That's still a relatively small number, though, so let's consider what it would take to reach critical mass and make this a meaningful advertising vehicle.

One option would be to entice consumers by offering commercial-free content to those who adopt the technology. Why would the networks ever agree to this? Well, they already face stiff challenges from TiVo and DVR penetration. If they could use this platform to drive viewers to their Web properties and share in the ad revenue from sponsored listings, it might be worthwhile.

Another option would be extending this technology to cell phones and PDAs. While there are certainly a fair amount of people who chat on the phone while watching TV (unfortunately, my wife is one of them), most people will at least have their cell with them in the room.

Imagine if people watching MTV could download a ringtone from the artist whose video they just watched. Or if the voting options for "American Idol" popped up right on your phone as you were watching the show.

As the paper points out, this technology could also be used to interpret radio signals--another reason why the cell phone might be the right device to tap. Nearly all handsets being manufactured these days are Web and video-ready, so the robust applications are still possible. And certainly, subtle keyword text ads would be suitable in this environment.

So, enough about personalization and advanced search tools. How long before search engines literally start listening to us? And will it make AdSense?