Does TV Advertising Really Affect Sales? The Role of Measures, Models, and Data Aggregation

Abstract
Traditional econometric models suggest that advertising has a clear and significant positive effect on sales in the current period. However, recent studies using scanner data indicate that the estimated effects of TV advertising on households' brand choices are weak and rarely significant. Do those findings mean that TV advertising does not really have an impact on current brand choices and sales? Or are the discrepant findings due to differences among the measures, models, and aggregation levels used by different researchers? The authors address these issues. Their analysis indicates that aggregating data over time and households may create a false impression of advertising having a statistically significant effect on sales.