The Impact of Advertising Along the Conversion Funnel

The Impact of Advertising Along the Conversion Funnel

By Stephan Seiler, Song Yao
2017Working Paper No. 3508

We assemble a unique dataset which combines information on supermarket feature advertising with path-tracking data on consumers’ movement within the store as well as purchase information. Using this novel data, we trace out how advertising impacts consumer behavior along the path-to-purchase. We find that advertising has no significant effect on the number of consumers visiting the category being advertised. The null effect is precisely estimated and even at the upper bound of the confidence interval, a one standard deviation shift in advertising increases category traffic by only 1.3%. On the other hand, we do find a
significant effect at the lower end of the conversion funnel. A one standard deviation change in advertising (evaluated at the point estimate) increases category-level sales by 10%. We further decompose the impact on sales and find that the increase is driven by consumers buying a larger number of different products within the category. The number of consumers purchasing in the category is not affected. We find no evidence of spillover effects of advertising between categories which are stocked in proximity of each other in the store and find positive spillovers between products in the same category. Based on these patterns, we propose a simple model that explores potential mechanisms.