A system for measuring audiences of outdoor advertising in specific areas is based on the combination of mobile phone location estimations with Internet listings of social events

O nline advertising is the fastestgrowing advertising medium, not least because it can track not only how long someone was on a webpage with an advertisement but also how many times someone clicked on that ad. By contrast, outdoor advertising has not reached its full potential because it cannot measure return on investment. People spend 27 percent of their time exposed to outdoor advertising, but such forms of advertising attracted only 5 percent of US media spending in 2008.1 The problem is that, for measuring advertising effectiveness, media planners currently rely on gross traffic numbers or circulation counts from the Traffic Audit Bureau, which represent historical data that has never been audited. As expected, unvalidated measurements do not call for marketing dollars. The industry would embrace outdoor advertising only if credible audience measurements were introduced. A more credible way to measure audiences for a billboard should include both the number of people in front of the billboard and the likelihood that those people might like a specific ad shown on the billboard. We propose a system that estimates the number of people based on the number of mobile phones near the billboard and infers people’s preferences by combining location estimations from the mobile phones with listings of social events (such as a football game or music festival) that are freely available on the Internet.