Purpose – The purpose of this paper is to provide an evaluation of a newspaper insert survey and web-based voter surveys associated with the same newspaper, conducted before and after the 2004 US presidential election.

Design/methodology/approach – The study compares response rates, demographics, and political profiles of voters responding to these two different channels of communication and evaluates the success of each in predicting the election outcome.

Findings – The study results show some significant differences between the two methods of voter data collection; nevertheless, each is useful in a comprehensive system that attempts to follow voter attitudes and intentions before the election and predict election outcome.

Origiality/value – The study relies on the use of an innovative marketing poll that goes beyond simple prediction of a voter's behavior and offers an explanatory component useful in the development of marketing strategies during a campaign.