Advertising investment risk has become high due to the explosive information increase and the commoditization evolution. In order to reduce such risk, it is important to understand the relation between advertising elements and consumer responses, and apply the knowledge to decision making process of advertising production. There have been lots of efforts to develop a set of scales to measure the advertising perceptions and relate it to the consumer responses. However, that is not adequate when we assume to support the actual practices of the advertising production. Although the advertising perceptions are helpful to understand the reasons of the consumer responses, they don't directly lead the way to the advertising production.

In this study, we analyze the effects of 28 elements of TV advertisements on the advertising perceptions and the consumer responses. The elements adopted in this study are controllable by advertisers so that they can apply the research findings to the advertising production. Empirical analyses using a dataset of the advertising perceptions and the consumer responses to more than 800 TV advertisements clarified not only the important advertising elements which had high impacts on the consumer responses but also how and why they affect consumer responses. Simultaneously, simulation of the consumer responses using random forests was found to be helpful to explore the effective combination of the advertising elements.