Current Issue Summary
March 2021 (Vol. 61, Issue 1)
Accounting for Causality When Measuring Sales Lift from Television Advertising: TV Campaigns Are Shown to Be More Effective for Lighter Brand Users
In their statistical analysis of single-source data, Henry Assael and Masakazu Ishihara (both at New York University), and Baek Jung Kim (University of British Columbia) cite the feasibility of using observational data, rather than controlled experimentation, in accounting for causality in advertising. Randomized trials are difficult to implement when measuring sales lift in TV advertising, because of the need to control for a wide variety of demographic characteristics and customer purchasing behavior, before the campaign begins. The authors account for the lack of such randomized trials by using observational data to control for endogeneity (i.e. factors other than advertising related to sales), while also accounting for heterogeneity (i.e. variations in consumer characteristics affecting responses to advertising).
Assael, Ishihara and Kim suggest that the biggest problem with not properly controlling for variances of this type is that without such controls “there is little effect of advertising exposure on brand purchase.” Indeed, “households that were exposed to advertisements during the campaign period are statistically indifferent from the households that were not exposed to advertisements during the campaign period.”
With controls for endogeneity and heterogeneity, the results showed greater ad effectiveness among lighter users. The study tested sales lift for a brand in a chocolate-candy category for a TV campaign that ran from March 28 to June 16, 2016. The single-source data collected by Nielsen Catalina Solutions combined TV advertising exposure and purchase data from the same household at the individual level, “so that the effects of television advertising on purchase behaviors can be studied.” A full 435,327 households were involved.
Among the takeaways:
- The advertising examined did increase brand purchase for both purchasers and nonpurchasers. But these results were even more pronounced for lighter brand purchasers and non-purchasers when controlling for endogeneity and accounting for heterogeneity.
- Failure to control for endogeneity resulted in the opposite effect—the campaign being most effective in influencing heavy users.
- With proper matching controls in place, the authors suggest: “Targeting heavy brand users would yield no advantage.
Read the full article here.