Measuring Advertising Effectiveness

William D. Wells

• Publisher: Psychology Press
• Year: 2014
• ISBN:
  9781317779506
• Edition: 1
• Topic: Business
• Subtopic: Consumer Behaviour
• Index: Business

This volume includes edited and revised versions of the papers delivered and discussed at the recent Advertising and Consumer Psychology Conference. Following the theme of the conference -- "Measuring Advertising Effectiveness" -- the book blends academic psychology, marketing theory, survey methodology, and practical experience, while simultaneously addressing the problems and limitations of advertising. Acknowledging that advertisements are subtle, diverse, complex phenomena that require detailed investigation, this compilation explores the multidimensional nature of advertising's diverse effects from both academic and applied perspectives. Updates on theories and methods -- along with expert commentaries -- help to make this a valuable collection that will be of interest to advertising and marketing specialists and communications experts alike.

VI

Copy Testing

We now turn to overtly applied considerations. In chapter 15, two veterans present a practical cookbook that summarizes many years of copy testing experience. As the discussant of this chapter notes, it contains “many helpful hints for pretesting short-term effectiveness.” As the discussant also notes, not everyone will agree with all these prescriptions. The authors are Larry Percy, formerly Research Director at Lintas and now proprietor of his own consulting firm, and John Rossiter, coauthor with Percy of a well-known advertising textbook, and faculty member at the Australian Graduate School of Management. The discussant is Esther Thorson of the University of Missouri-Columbia.

In chapter 16, two academicians challenge many popular copy testing methods by proposing alternatives to key assumptions. In commenting on this challenge, two practitioners acknowledge that, in principle, the academicians might be right. The authors, Richard W. Olshavsky and Anand Kumar, are based at the University of Indiana. The practitioners are Larry Percy, coauthor of the preceding chapter, and Abhilasha Mehta of Gallup & Robinson, Inc.

Chapter 17 summarizes a large collection of real evaluations of real advertisements. It presents substantial evidence that the difference between preexposure intention and postexposure intention is the single most valid predictor of sales effectiveness, and that the relationship between day-after recall and sales effectiveness is at best extremely weak. It also provides substantial evidence that whether consumers like an advertisement has no bearing on its effectiveness, that bare-bones renditions of sales propositions predict the sales effectiveness of finished ads, that executional differences have measurable effects on sales, and that wear-out with repeated exposure follows a predictable course. Allen Kuse, the author of this chapter, is an executive at research systems corporation, a firm that specializes in pretests of rough and finished advertisements. The discussant of this chapter, Brian Wansink of the University of Illinois, Urbana-Champaign, identifies implications of Kuse’s methods and conclusions for academic advertising research.

Chapter 18 presents a case history in which an “informational” advertisement and a “transformational” advertisement reached their communications objective through different routes, and it details the process through which that outcome came about. Abhilasha Mehta and Scott C. Purvis, the authors, are executives at Gallup & Robinson, Inc. Brian Wansink commented on this chapter as well. In his comments, Wansink links Mehta and Purvis’ real-world concepts and findings to concepts and findings in more theoretical academic work.

Chapter 19 describes a quite different case. Here, the contrast is between more traditional methods of measuring advertising effectiveness and a method that examines a chain of relationships that leads from perceptions of product attributes to beliefs that the product will contribute to important aspects of the perceiver’s life. The authors, like the content, blend academic and industry affiliations. Thomas J. Reynolds, a former academician, is an executive at Richmond Partners, Inc. Jerry C. Olson is an academic researcher at Pennsylvania State University. John R Rochon is an executive at Richmont Partners, Inc. The discussant, Christine Wright-Isak of Young & Rubicam, Advertising, fits means-end chain analysis into the more general context of measuring advertising’s broader, medium to long-range effects.

Chapter 20 examines an important exception to the common assumption that advertisements are intended to elicit purchase rather than use. The products featured in chapter 20 are already in inventory in many of their users’ homes, and the object of the advertising is to move the product off the pantry shelf. As chapter 20 shows, that situation requires a unique approach to measuring effect. The authors of chapter 20 are both academicians—Brian Wansink at University of Illinois, Urbana-Champaign and Michael Ray at Stanford.

Chapter 21 focuses on another important exception to standard copy testing tasks. It reminds us that an important segment of marketing communication is not distributed through electronic and print media, but rather reaches the purchaser at the point of sale. It describes the assets of “place-based” media, and outlines a procedure through which the effectiveness of such media can be assessed. Like the authors of chapter 19, the authors of this chapter blend industry and academic posts. James Lucas is a research executive at Frankel & Company, a prominent promotions firm. David Prensky, formerly an advertising agency researcher, is on the faculty at Trenton State University.

15

A Theory-Based Approach to Pretesting Advertising

Larry Percy

Marketing Communications Consultant

John R. Rossiter

Australian Graduate School of Management

The job of evaluating the likely success of any particular advertising execution requires original primary research custom tailored to each particular circumstance, but executed within strict theoretical guidelines. Accordingly, we caution against sole reliance on the standardized procedures offered by any one particular supplier or syndicated service. Flexible procedures based on the brand’s advertising communication objectives provide a better fit and more valid results.

The purpose of pretesting is to improve the chances that advertising will work as planned when placed in media. Whether it will work as planned, however, depends on three factors:

● The creative content of the executions.

● The correct media placement and scheduling.

● Competitive advertising activity.

Pretesting deals with the first factor only. Campaign tracking over time evaluates all three factors. We pretest advertising to ensure it is “on strat-egy”—that it is capable of achieving the communication objectives desired for the brand, to be able to correct and revise the executions if necessary, and then to predict how it will work in the market.

METHODS UNSUITED FOR PRETESTING ADVERTISING

Before we begin to examine the best ways to pretest advertising, however, we want to point out three ways not to go about it, even though the first two are often used. The methods about which we hold strong reservations are focus groups, recall, and physiological measures.

Focus Groups

Whereas focus groups are essential for formulating the advertising strategy prior to the campaign, they are inappropriate for testing ads (Rossiter & Donovan, 1983). There are at least three compelling reasons. Focus groups vastly overexpose advertisements. In a group setting, advertisements are thoroughly discussed, a far cry from the 30 seconds, more or less, that broadcast advertising has available to communicate in the real world and the 1 to 2 seconds that a print or outdoor ad has to gain attention. This overexposure leads respondents in focus groups to exaggerate positive and negative aspects of the advertisement, as well as to focus on elements in the execution that would never get processed in the real world, generating a serious validity problem.

This validity problem follows from the group setting itself, which produces group interactions that largely prevent reactions from occurring as they would normally. Normally, people process advertising as individuals, even if they are watching TV with others.

The third problem is the lack of reliable projection. Even if the focus group methodology itself were valid, there would remain a serious reliability problem, owing to the small overall sample size. It would be a rare (as well as inefficient and expensive) test in which 100 respondents were exposed to the test advertising.

Recall

The most that can be said about advertising recall measures, especially day-af-ter-recall (DAR) testing, is that they may be a rough measure of attention to the advertising. However, the fundamental flaw with recall testing, especially DAR, is that it is ad recall based, not brand recall based. Most DAR measures are initially and prejudicially tied to a particular media vehicle as a memory cue, and to a particular insertion, at a particular time. Such a measure cannot be generalized (Gibson, 1983; Percy, 1978). A similar problem applies to the “recall” measures used by leading pretesting services, such as RSC’s ARS Recall^SM and MSW’s Clutter Awareness^SM, which have not been shown to be valid measures of advertising effectiveness even in these firm’s own studies (Blair, 1988; Klien & Tainiter, 1983; see also Young & Robinson, 1992). No pretest measure of advertising recall has ever been shown to predict advertising effectiveness. This is because media vehicle (as in DAR) and the test situation (as in the case of most syndicated recall measures) are cues that are irrelevant to the consumer’s decision process.

Physiological Measures

Physiological measures provide a measure of attention only. EEG or branvwave measures record electrical response to executions, but the interpretation of this response is not clear (Olson & Ray, 1989; Rothschild, Hyun, Reeves, Thorson, & Goldstein, 1988; Stewart & Furse, 1982). Eye-movement measures (Kroeber-Riel, 1984; Weinblatt, 1985, 1987) exhibit some relationship with brand recognition, but not with brand recall or brand attitude.

THEORETICAL BACKGROUND

We believe that advertising pretesting must be based on the communication objectives for the advertised brand (or company or service) and that these, in turn, depend on the buyer decision process that the advertising is intended to influence. The five primary communication effects for advertising are outlined in the following, and it is from this set of effects that the objectives for a particular execution will ...

Table of contents

• Cover Page
• Half Title page
• Half Title page
• Title Page
• Copyright Page
• Contents
• Preface
• I Effects and Effectiveness
• II Subtle Processing
• III The ELM Model
• IV Cognitive Elaboration
• V Context
• VI Copy Testing
• VII Afterword
• Author Index
• Subject Index