Advertising and Brand Attitudes: Evidence from 575 Brands over Five Years

Rex Yuxing Du, Mingyu Joo, and Kenneth C. Wilbur, 2018, 18-138

The ultimate goal of advertising is to influence sales. However, recent literature shows that it is very difficult to estimate advertising effectiveness on sales. Extremely large sample sizes are required to adequately power advertising experiments and observational methods may be a poor substitute for experimental estimates. Sales data are sparse and highly variable, ad effects tend to be small and short-lived, and advertising exposures are nonrandomly assigned.

Rex Yuxing Du, Mingyu Joo, and Kenneth Wilbur investigate whether brand attitudes can be reliable measures of advertising effectiveness. In contrast with sales data, brand attitude data are non-sparse and relatively stable, possibly more influenced by advertising (though most of the effect signs and magnitudes remain unknown), and measurable in large samples of consumers whose selection is unrelated to advertising treatment.

They examine a unique dataset of 575 established brands from 37 industries over a five-year observation window, merging weekly brand attitude survey data with weekly advertising expenditure data. In totality, the data include $264 billion spent on advertising, 37% of all ad spend measured during the observation window, and approximately ten million brand attitude surveys.

They address the following three questions for mature brands that advertise regularly:

  • How do brand attitudes change with advertising by the same brand and its competitors?
  • How do these relationships vary across attitude measures and types of advertising media?
  • How do various strategies to control for time-varying unobservables change effect sizes and precision?

The brand attitude metrics they consider are the percentages of survey respondents indicating favorable perceived quality, perceived value, and recent satisfaction for each brand in each week. The three types of advertising media they consider are national traditional media, local traditional media, and digital media. The models they estimate all include lagged brand attitudes, contemporaneous and lagged ad spending by type of media, brand fixed effects, week fixed effects, and weighted standard errors to reflect exogenous variation in the number of survey respondents each week.

They investigate two sets of control variables as partial remedies to this advertising timing endogeneity problem: brand/quarter fixed effects; and industry/week fixed effects. The data indicate that brand/quarter and industry/week fixed effects are individually and jointly important determinants of brand attitude data.

The key findings are that

  • brand attitude metrics all rise with multiple lags of the brand’s own national traditional advertising;
  •  local traditional ads increase perceived quality and perceived value;
  • digital ads increase perceived value;
  • the effects of competitors’ ads are generally negative.

The results indicate that inclusion of proper control variables brings advertising parameter estimates closer to expectations without major reductions in estimation precision. However, the brand attitude data analyzed in this study provide insufficient power to estimate precise industry-specific effects, suggesting further investigation of other intermediate metrics such as store traffic, consideration, or information acquisition via online search as candidate advertising response measures.

Put into Practice

These findings and empirical strategies may aid marketers and their agencies in using data to guide important practical questions such as whether to advertise, how much to spend, and how to allocate ad budgets. Such empirical guidance may be especially needed in industries where available data complicate the estimation of causal effects of ads on sales, such as markets with long purchase cycles or long interpurchase times.

Rex Yuxing Du is Bauer Professor of Marketing, Bauer College of Business, University of Houston. Mingyu Joo is Assistant Professor of Marketing, School of Business, University of California, Riverside. Kenneth C. Wilbur is Associate Professor of Marketing, Rady School of Management, University of California, San Diego.


  • Corporate: FREE
  • Academic: FREE
  • Subscribers: FREE
  • Public: $18.00



Employees of MSI Member Companies enjoy the benefits of complete online access to content, member conferences and networking with the MSI community.



Qualified academics benefit from a relationship with MSI through access to, conferences and research opportunities.



The public is invited to enjoy partial access to content, a free e-newsletter, selected reports and more.




Learn more about becoming a member of the institute

Read More

Stay Informed

The MSI Mailing List

Subscribe to our email list to stay informed about upcoming events, news, etc.

Social Media