Advertising Principles - Evidence-based principles

Persuasive Advertising Review

A review of: Persuasive Advertising: Evidence-based Principles, J. Scott Armstrong (Ed.), Palgrave Macmillan, 2010. Philip Gendall

Persuasive Advertising is the result of 16 years of reading and research by Scott Armstrong and his collaborators, Gerry Lukeman and Sandep Patnaik. Their aim was to evaluate what is known about persuasion in advertising and then translate their findings into evidence-based principles that would guide development of more effective advertisements. The result is a comprehensive summary of more than 3,000 empirical studies and 50 books that is well- written and thoughtfully presented. It is hard not to be impressed by the scale of the enterprise and the clarity of its exposition.

Armstrong nails his colours to the mast in the Preface, which is headed by a quote from Bill Bernbach: “Advertising is fundamentally persuasion,” and an underlying assumption throughout the book is that advertising is a powerful force that persuades people to act. If you share this belief, you will find much evidence in the book to support your view. Personally, I am not so convinced, for reasons that I will explain.

Armstrong believes that advertising is a powerful, persuasive force, so powerful in fact that it can determine the success or failure of a product, or even a whole enterprise. Thus, Song Airlines’ failure is attributed to the fact that it “ignored the fact that their new airline was a utilitarian service and instead treated it as hedonic in their advertising”. Similarly, the demise of Digital Equipment Corporation is ascribed to its advertising, which listed features rather than benefits. There are many reasons why products and businesses fail or succeed – management competence, cash flow, timing, luck – advertising may be a contributing factor, but the assumption that advertising is primarily responsible is, in my opinion, overrated (or, at least, not substantiated in the examples provided).

Armstrong’s review of published work is very extensive; most of it is quite recent – published in the last 20 years - but some of the empirical studies are more dated and were conducted in the 1950s or 1960s. Their results may still be relevant, but advertising is a social phenomenon and, like all social phenomena, is subject to change over time. Some motivations such as fear, greed, anxiety, a need for reassurance, are probably just as relevant now as they were 50 years ago. But today’s consumers are not the same as consumers of the 19650 and 1960s; they are more sophisticated and better readers of advertising than their parents or grandparents. Consequently, the current relevance of some of the early research quoted by Armstrong is debatable. The problem is knowing which findings still apply and which do not, a point the authors could have explored further.

With some qualifications about the value of different types of evidence, the evidence of a particular type is given similar weight. Thus, the results of a laboratory study of 100 US university students’ perceptions often receive the same attention as a large, field experiment involving actual behaviour. Consequently, the narrative gives little or no sense of the quality of the different studies reported on; the reader has to assume that Armstrong’s assessment of the relative merits of the various findings accurately reflects this. More importantly, most of the evidence presented is based on Glossary Link recall, intentions or other cognitive measures, rather than measures of behaviour. Armstrong argues there is good evidence that these cognitive measures are strongly correlated with actual behaviour, but this view is not universally shared, nor is there unanimous agreement on the direction of causation.

Another problem with many of the studies Armstrong reviews is how the constructs tested were executed. This is an issue for any research that involves the execution of an abstract construct such as fear or humour. Some experimental stimuli are unequivocal: if one treatment includes the price of the product and the other does not, there is no doubt about the cause of any observed effect. But if one advertisement is a ‘humorous’ one and another is not, what can we say if the first advertisement is less effective than the second? Can we say that humour does not work? Perhaps. But how do we know if the first advertisement was actually funnier than the first, and even if it was, how do we know that a more humorous advertisement (a better executed ad) would not have been more effective?    Furthermore, if the first advertisement was more effective, how do we know what, specifically, will be effective in a different situation?

In fact, Armstrong emphasises the importance of context, or conditions, in the application of the principles he outlines. However, many of the principles require subjective judgement of the conditions (‘Do not violate taste or standards’), some of the principles are tautological (‘Use power words if they fit the product’) and some appear contradictory (‘Use high prices to connote high quality’ and ‘When quality is high, do not emphasise price’). Unfortunately, for those faced with the task of developing an advertising Glossary Link campaign, context is the enemy of advertising principles. If ‘principles’ are heavily context dependent, are they really principles? This is not a criticism of what Armstrong’s book sets out to do, but rather a caution that likening its principles to the ubiquity of Newton’s law of gravity, as Armstrong does, is drawing too long a bow.

However, these are all relatively minor issues compared to what seems to me the book’s most glaring oversight. This is the complete absence of any reference to the work of Andrew Ehrenberg or, apart from one mention in an appendix, the work of John Philip Jones, or of any suggestion that there may be an alternative view of the way in which advertising works. Ehrenberg’s ‘weak’ theory of advertising proposes that advertising ‘works’ by reinforcing and maintaining behaviour rather than by persuasion, the key element of the ‘strong’ theory of advertising. Jones challenges some of Ehrenberg’s claims, but also argues that effective advertising for fast moving consumer goods is better explained by the weak theory than by the strong one, and that persuasion happens less often than many advertisers believe. These views are supported by a substantial body of empirical evidence that Armstrong does not mention.

Jones suggests that differences in the style of American and British advertising and the attitudes of practitioners in the two countries may explain why advertising works differently in the two countries. But if he is right, and advertising does work differently in the United States and Great Britain, then principles of persuasive advertising may be limited in their jurisdiction. However, this does not explain why none of Jones’s single source research on advertising effectiveness is considered in Armstrong’s book, particularly as Jones does not argue that persuasion is unimportant in advertising.

I accept that this is a book on persuasive advertising, aimed primarily at an American audience, but even so it is hard to imagine any serious review of advertising effects that ignores the work of Ehrenberg and Jones and the weak theory of advertising, and this seems an odd decision for such a reflective task. Nevertheless, this is a very interesting book; wide ranging, scholarly, but nicely written, and absorbing for anyone with an interest in advertising. However, as a summary of evidence-based principles of advertising it is incomplete. Ignoring the substantial body of evidence that challenges the view of advertising as strongly persuasive does a disservice to those who have never considered the alternative. Even if you reject the views of Ehrenberg and Jones, a comprehensive review of advertising studies should at least present these alternative views and allow readers to make up their own minds.

Philip Gendall is a Professor in the Department of Communication, Journalism and Marketing, Massey University.

Persuasive Advertising is only the end of the beginning: A rejoinder to the review. J. Scott Armstrong

Professor Gendall and I agree about Persuasive Advertising (Armstrong 2010) on some points and differ on others. I appreciate his kind comments about the book, but in this reply I will focus on our differences. I follow the sequence of topics in his review. The italicized sub-headings represent my viewpoints.

Advertising is an investment. I am steadfast that the purpose of advertising is to persuade people to act-- to buy, sell, vote, provide help, save souls, and so on. In other words, it should produce some gain or return on the investment. The gains are not only for stockholders but also for all stakeholders, such as customers, employees, and retailers. I accept that Professor Gendall is not convinced on the need for results, but I expect that many advertisers will agree with me. As for the power of advertising, sometimes it is profitable –in the sense of benefits are derived– and sometimes not. The purpose of my book is to achieve greater benefits from investments in advertising.

Examples are not evidence. Examples are designed to illustrate, not to provide evidence. Thus, I was pleased that my examples did not convince Professor Gendall, as they were not intended to. In my book, I try to dispel the common belief that examples constitute evidence and put them in separate sections so as to avoid this error.

Early evidence is relevant. As I state in the conclusions section of the book, the principles are expected to change over time as we learn more about how they are best applied and the conditions under which they apply. For example, contrary to early research, I found that prospect theory offers no benefit to advertisers. So the principles are based on the cumulative evidence to date. I was impressed by the value of the early studies and their contributions to the development of the principles, so I do not recommend discounting research merely because it was done many years ago.

Different types of evidence are advantageous and they can be weighted differently. I provided much heavier weight on experimental evidence. Beyond that, I aimed for full disclosure, and even pointed out some studies that deserved little weight. I consider the fact that there were many approaches and many criteria as an advantage rather than a disadvantage. In any event, I provide full disclosure of the evidence so that users of the book can weight the evidence as they wish.

Evidence-based advertising requires good judgment. I expect that this will always be the case for advertising – as it is in medicine and other complex fields. Of course, we know a lot about how to improve judgment. Thus, for example, I recommend that independent judgments be combined to improve judgments. I also recommend the use of checklists as an aid to judgment.

Evidence-based advertising supplements rather than replaces knowledge about advertising. I was surprised that Professor Gendall views Persuasive Advertising as “a comprehensive review of advertising” and as being similar to work by Ehrenberg and Jones. While I am honored to be included in such good company, I view evidence-based advertising as a completely different approach. It does not aim to refute or replace other approaches to advertising – and it is not the final word. It is narrowly focused on evidence-based knowledge about persuasion.

Is that really new? As I describe in a forthcoming paper (Armstrong 2011), I have been unable to find any of the 195 principles in existing advertising textbooks. In addition, the principles go beyond “common sense.” When given true-false questions about the principles, on average, people do no better than guessing.

Hopefully, the principles will supplement advertisers’ knowledge and skills and lead to more effective advertising. In addition, I hope that researchers will continue their valuable work of determining when the principles are most effective and how they can be applied effectively. While I am the messenger, the book is a tribute to the useful findings produced by thousands of researchers, and to the painstaking efforts of those who helped ensure that the principles reflect the knowledge to date.

Armstrong, J. S. (2010), Persuasive Advertising. Basingstoke, Hampshire: Palgrave Macmillan. Armstrong, J. S. (2011), “Evidence-based advertising: An application to persuasion”
Forthcoming in International Journal of Advertising.
J. Scott Armstrong is a Professor at The Wharton School, University of Pennsylvania, Philadelphia, PA


© Copyright J. Scott Armstrong and Kesten C. Green. All rights are reserved.
Web Design by Zoe Communications Ltd.