Effects of configuration and exposure levels on responses to web advertisements

The debate about which media metric efficiently measures the effectiveness of a web-based advertisement, such as banners, is still alive and well. Nonetheless, the most widely used measure of effectiveness for banner advertisements is still the click-through rate. The purpose of this article is to review the measures currently used to measure effectiveness in web advertising and to empirically determine the factors that might contribute to observed variations in click-through rates based on an actual sample of advertising campaigns. The study examined the complete set of all advertising insertions of 77 customers of a large advertising agency over a one-year period. A resulting sample of 1,258 placements was used to study the effect of banner formats and exposure levels on click-through rates using analysis of variance. Results suggest that the strongest effect on click-through rates comes from the use of trick banners ( eta(2) = 0.25) and that other factors such as size of the advertisement, motion, use of "click here," and "online only" type of announcers all have a significant impact of click-through rates. Implications of these findings as well as limitations of the current study are discussed and directions for future research agendas proposed.