An Approach to Model,Estimate, and Control the Effect of Ambient Temperature on Advertising Experiments

Abstract

The impact of ambient weather may not be obvious and weather may play a bigger part than previously attributed to variations in experimental and test market results including advertising results. While some researchers have passed off this issue as just another “seasonal effect,” ambient weather conditions have played a major impact on various test market studies. The purpose of this study is to demonstrate the following: (1) ambient temperatures that differ from seasonal mean temperatures may affect sales in certain cities; moreover, the effects differ among sales of certain varieties both within and between cities, and (2) these effects may be estimated and controlled. The results show that while seasonal effects explain most variation, there is a significant amount of variation explained by ambient temperature. This variation could be the difference between which a number of alternative commercials or promotions are selected for implementation.