Abstract: For an advertisement opportunity near a travel region, advertisements may be selected that are targeted to individuals who are likely to view the advertisement. However, travel patterns among individuals sharing particular traits may exist that facilitate targeted advertising, but may be non-intuitive and therefore difficult to predict, and other techniques, such as population surveys, may be costly and inaccurate. Presented herein are techniques for automatically evaluating travel patterns by tracking the routes of particular individuals, and inferring demographics for such individuals based on the locations of their routes (e.g., an individual whose route frequently includes a residence may be presumed to share the population demographics of the residential neighborhood). Extrapolating such individual demographics may enable inference of shared demographics at particular advertisement opportunities (e.g.