According to eMarketer mobile ad spending will reach almost $5 billion this year, with the lion's share of that going to "direct response ads," which are what location based advertising is best suited for. Location based ads are very attractive to advertisers because they add a personal level of targeting that's already available and used effectively on the Internet. When I search for "toyota" on Google, I'm served ads for my local dealerships. GPS technology can target mobile ads even more precisely and make them even more relevant to where you are at that moment.

But there are plenty of potential hangups. Take the deal that CBS announced this morning with mobile social network Loopt. CBS plans to use Loopt's GPS technology to deliver location based ads to CBS mobile users. These aren't exactly like the scenario I announced above, since CBS will be displaying location-aware ads on top of mobile content rather than deliver them via text message. That actually highlights the first problem with location based advertising.

It would take a perfect confluence of events in order for many ads to make sense. Not only do you have to be near the thing that is being advertised, but you also need to be viewing the CBS mobile site. For entities like Starbucks that exist on every street corner, that might not be an issue, but in practice how often do you think you'll be in the proximity of one of CBS' advertisers while you're viewing the site? It'll happen, sure, but it drastically cuts down on the number of opportunities to deliver location aware mobile ads when you have to be viewing a specific mobile web page at just the right time to receive an ad.

CBS can, of course, deliver more general ads fixed to your location -- but is that really taking advantage of the GPS capabilities that Loopt offers? Yeah, it's neat to see ads for things in the city I am traveling in, but not as neat as seeing ads for the businesses on the street I'm walking down.

The fix for that, of course, is to deliver the ads by text message, triggered by proximity to the advertiser's storefront. We questioned in December whether mobile ad startup Fluc would be able to fill inventory for a similar type of ad scheme (though is isn't clear if Fluc is actually targeting ads by GPS or if they're just asking users where they are located). "If the GAP knows you're near a mall where they have an anchor store, and they know from your Fluc account that you fit their consumer profile, then they might pay to send you an ad," we wrote. "That's a lot of 'ifs,' though."

Further, ads like this have to be opt-in. Not only is there a privacy issue involved, but text message ads are also the least likely to be trusted by consumers, according to a Nielsen Internet survey. The CBS-Loopt ads are opt-in.

Another hangup with the idea of location-based ads is reach. Right now the CBS ads are only available to customers using a GPS-enabled phone on a network that Loopt has a partnership with -- so far that means just people on Sprint Nextel or Boost Mobile. Loopt's CEO predicts that by the end of 2008 there could be 50 million mobile phones in the United States equipped to receive this type of advertising, but for now the audience remains relatively small.

Even so, location based advertising is a tantalizing vision for the industry. Google CEO Eric Schmidt said last week that location based ads are the future and will lead a revolution in mobile advertising. Last year Google launched a mobile version of AdSense and their own mobile OS. I think we can expect location aware ads from them in the near future. And whatever Google does in the world of advertising, you can bet others will follow suit.

The ad scenario I described above is possible (and it can get even spookier and more finely targeted when mashed up with other data -- say, your social networking profiles), but it's probably not quite here yet. At least, I haven't seen it. The question is -- do we even want that? Should we potentially trade more of our privacy for more relevant advertising? What does a dollar off a latte mean to you?