In many of these apps, a user "checks in" at restaurants or coffee shops in order to pose a review or let friends know he or she is here. Foursquare is the hot property in the category at the moment, in part because it turned the concept into a bit of a game, where frequent visitors to a locale can float to the top of some kind of abstract virtual community and declare themselves the "mayor" of a bistro, an office building, maybe even a hot dog stand. More power to you, oh Mayor of Ace Hardware in Fairfax, Del. Boasting different functionality and features are similar schemes like Yelp!, Loopt, Buzzd and countless others with equally goofy names.

One of the more established players in the field, Brightkite, claims over 2 million users of its mobile apps on iPhone and Android as well as the mobile Web. CMO Rob Lawson tells me that his company tries to differentiate itself in an admittedly cluttered field by focusing on real-world friends and groups who are keeping up with one another and discovering new places. "We are not as fixed on [U.S.] geography and technology, but on friends and groups," he says.

Geography does play an important role here, in that people do check in from locations and the system is designed to track people as they physically navigate their social lives and discuss it with one another. "We have data around who people are and the things that they do," says Lawson. If you go to coffee shops a few times a week or go to the movies more than one a month, the Brightkite system knows this and now is targeting ad campaigns against this knowledge.

Calling the approach "ultratargeting," Brightkite says it has already done over 20 pilot campaigns that can take into account a user's location, recent behaviors, the content of their conversations, the time of day, even the local weather. Most of the data comes from individual check-ins, but also inferences the system can make off of that data and historical knowledge of activity.

Brightkite says it recently finished a campaign for Chevrolet that targeted people within a three-mile radius of a dealership only on the weekend. It claims success in driving traffic to the dealership. For a Las Vegas "Jersey Boys" show, Brightkite targeted only users it knew came from out of town with a special offer. It claims the redemption rates were over 2%.

Lawson admits that even with a couple million users, parsing the crowd down to time, locale and behaviors creates a natural issue of scale -- you have to slice the data wisely to get desired ROI. He also insists that the user data is fully anonymized and the members are only sharing what they designate to be shared. For instance, whenever I make a post in the iPhone app, it asks whether to make the message available to everyone on the network or only to my declared friends, whether to send it to Facebook and Twitter accounts as well.

How well users are aware of what is being tracked for marketing purposes is less clear to me, however. Almost all of these local mobile social nets will have to negotiate thorny privacy and data usage issues over time. Sharing information with my friends is a good deal different from choosing to share that data with a marketer, and we have already seen how far it gets technology companies to say the information is "anonymized."