Modeling the Determinants and Effects of Creativity in Advertising

Back to the Peer Reviewed Journals List

Smith, Robert E, MacKenzie, Scott B., Yang, Xiaojing, Buchholz, Laura M., Darley, William K. (2007). Modeling the Determinants and Effects of Creativity in Advertising. Marketing Science, Vol. 26, No. 6. pp. 819-833, 15 pages.

Reviewed by Christopher Berry, October 2009

Executive Summary

Smith et al get straight to the point in their opening line: “Marketing researchers and practitioners agree that creativity is one of the essential elements for advertising success in a cluttered marketplace”.

Their study is aimed at defining ad creativity, measuring it, and examining its determinants.

The article examines previous methodologies of defining ad creativity and produces a better one. They identify two concepts in their literature review, divergence and relevance, as being the primary variables in defining creativity. They define divergence being ‘originality’ – how much a particular ad stands out from the pack. They define ‘relevance’ as “the extent to which at least some ad/brand elements are meaningful, useful, or valuable to the consumer”.

They infer that ads with low divergence and high relevance are easy to miss and ignore in a cluttered landscape, while ads with high divergence and high relevance are the most effective.

The authors do not (refreshingly) start from scratch. Instead, they go about refining previous definitions of both divergence and relevance using a survey methodology to gradually reduce the number of questions pertaining to each group and their relative impacts on ‘creativity’. These include questions like “The product or brand was meaningful to me”, “The ad demanded my attention”, and “The production elements of the ad were of high quality”.

Then they showed students a number of ads which included award winning ones – making the assumption that award winning ads would be judged more creative. Students were asked the shortlist of questions based on the authors’ operational definition of creativity and evaluated the effectiveness of their model. Their model is a series of questions, grouped thematically into groups and indexed, then treated like independent variables to explain variation in a dependent variable: creativity. They found that award winning ads indeed tended to be more creative.

They found that divergence is the leading indicator of creativity. Interestingly, they found that creativity was very strongly correlated with attention to the ad, but not to purchase intent.


The article will be of particular interest to web analytics practitioners and managers in very brand oriented companies.

A practitioner can borrow from this methodology to add detail to their analysis of why different pieces of creative worked, and why some were less than successful. It might also inform a range of informed hypotheses that could be progressively A/B tested. They might do this by using the survey questions (helpfully listed in Appendix A) and evaluate the ads themselves, or, use the questions to survey customers directly themselves. The answers could be put into a spreadsheet or statistical software and compared with the actual performance of specific ads, be it display, video, social, search, or on-site promotion. An analyst could then correlate creativity to various metrics, including attention, clickthrough, and ultimately conversion. Not that the analysis would be completely fair or necessarily perfect in assigning ROI, but it would yield actionable insight.

The authors go one step further, where I think web analytics should go: “In addition, it seems time for marketing research to move beyond the attentional effects of ad creativity and investigate its role in the persuasion process.”

Practitioners steeped in the direct response world might question the real role of creativity in ads. Indeed, much of the dialogue in Search Engine Marketing (SEM) world (possibly the closest thing to ‘direct’ marketers we have in digital) focuses on the questions of rank, price, and keyword relevancy. There are 95 precious characters that go into a Google Adword Unit. How important is divergence? How important is relevance? There’s an opportunity, using this framework, to explore the effects of creativity on conversion. The same could be said for landing page optimization from a purely web analyst point of view.

I would recommend this article to web analytics practitioners and managers who are particularly concerned with the role of ad creativity and its correlation to various metrics: including ROI.

A single copy of the full journal reviewed above is available to members of the Web Analytics Association. To request a copy, email Shannon Taylor.