Here’s an interesting market test for online content — selectively subject your audience to it based on geo-location.You largely don’t hear about geo-location testing anymore, yet Politco has decided to adopt this artifact of market research to test its audience’s receptivity to a metered paywall. Starting next week, readers in Iowa, North Dakota, Vermont, Mississippi, New Mexico and Wyoming will be subjected to a meter count when viewing the site’s currently free content (Politico already has a premium subscription area called Politico Pro).The blog post announcing the test also said that while the site is testing a meter for its readers nationwide, its DC-based audience will NEVER be subjected to a meter since the site often sells out its ad inventory for this market.So apparently, “separate and not equal” is the new model. And that’s a risky bet for Politico — especially in a virtual world where audience members could be in DC one week and New York the next.Clearly the site has chosen its six test markets based on size — small size, that is. By choosing small audiences, the site doesn’t risk alienating the majority of its users. But is that good market research? Probably not, especially considering Vermont is the only “blue state” in the list of six (and not a very “typical” blue state).Politico might have been wiser to adopt Consumer Reports’s model when it wanted to test its television ad across the political spectrum and socioeconomic status. CRO chose to test its ad spot in Madison, Wisc., which is known for having both rural and urban populations along with a healthy mix of registered Republicans and Democrats. Results from one test in a market that mirrored the general population are more easily generalizable than any of the results Politico will likely get from its six-state test.