Do Superbowl Ads Affect Brand Share?

To read the full-text of this research, you can request a copy directly from the authors.


Measuring the effects of brand advertising is challenging because of complicated dynamics and a lack of exogenous variation. We consider brand advertising in the setting of the NFL’s Super Bowl telecast. The Super Bowl is advantageous for measuring advertising for a number of reasons. First, the event has the potential to create significant increases in “brand capital” because ratings average over 40 percent and ads are a focal point of the broadcast. Second, variation in exposures is exogenous because a brand cannot choose how many impressions it receives in each market. Viewership is determined based on how much the local population wants to watch the Super Bowl, which also varies within market over time as well as across markets. We document much of this within-market variation in Super Bowl viewership to be attributable to NFL team performance and local preferences for teams. With this significant and exogenous variation in Super Bowl advertising exposures we test whether brand shares are affected accordingly. We run our analysis using Nielsen ratings and the brand shares in the beer and carbonated beverage categories. We find null and/or insignificant effects. We suspect the lack of positive effects arise because current industry practice employs methodologies that overestimate ad effectiveness. This leads brands to over-invest such that the marginal benefit of an exposure is driven toward zero.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

Full-text available
The IRI Marketing Data Set was introduced in 2008 (Bronnenberg, Kruger and Mela, 2008) with five years of data. Since then, the data set has been expanded to include more than twice as much data (2001-2011). This paper provides an update on the data set, and a bibliography with abstracts of 70 publications, theses, conference papers and dissertations which have used the data set. Access to the 130 gigabyte data set requires a nondisclosure agreement and a handling fee of $1000.
ResearchGate has not been able to resolve any references for this publication.