Playing the angles: a managerial opportunity in ad placement.

Larsen, Val ; Wright, Newell D.

ABSTRACT

This study explores the effects of an ad form variable-camera angle-on ads in an ad pod. It found that while camera angles do not affect recall/recognition or brand attitudes for the ad in which the camera angles are abnormal, it does affect these measures for proximate ads in the pod. The results suggest that advertising managers can gain a business advantage by placing ads in pods where other ads have abnormal executions. Doing so should magnify positive ad responses.

INTRODUCTION

According to James Marra (1982), the first and essential rule in advertising is "Gain Attention." But that first rule is qualified by the old adage "it's not creative unless it sells." Thus, in advertising, the objective of a creative team is attention without distraction, to break through boredom barriers and, amid a clutter of ads, win attention with ad concepts and ad executions that do not distract from the product and product claims. Of the tools available to the creative team as they seek to achieve this objective, ad form is the one most likely to be used across different ads.

Unfortunately, as a conference of the Marketing Science Institute (1988a) has made clear, practitioners are often forced to make decisions about ad form without having any clear idea what impact those decisions may have on viewers. MSI has suggested that the relationship between ad executional variables and viewer information processing should be a research priority (1988b). In the spirit of that priority, this study focuses on an important ad form variable, specifically, the ways in which camera angle can be manipulated to maximize attention gains while minimizing distraction costs.

WHAT IS FORM?

The nature of form has been a perennial question in philosophy, a question that has been answered in a variety of ways. After surveying various responses to the question, Steven Pepper (1970) argued that "formists" generally agree on one thing: form is rooted in the intuition of similarity. Thus, two objects have the same form to the extent that they resemble each other on some dimension. Insofar as they do not resemble each other, they are particular.

Given those definitions of form and particularity, it is obvious that everything has form in some degree, for everything resembles something else on some dimension. But dimensions differ in their range of applicability. Size and collar cleaning capacity are both dimensions on which products may be compared; however, the size dimension is relevant to many product classes whereas collar cleaning capacity is relevant only to a few. Thus, on the continuum anchored by form and particularity, size is a more purely formal, collar cleaning capacity a more particularistic dimension. In this study of ad form, the focus will be a dimension that is applicable to all ads that contain images.

THE MULTIPLICATION OF FORMS

The main pitfall of formal analysis is its tendency to produce new forms ad infinitum. In an inductive analysis of ad forms, a new variable emerges whenever two commercials are perceived to be similar on some dimension. Since commercials can be similar on an infinite number of dimensions, any researcher who looks for similarities between ads will find variable after variable after variable. These variables will be identified on heterogenous dimensions at all different levels of abstraction. They will, consequently, result in an arbitrary chaos of autonomous forms even if only the more generalizable and, therefore, more purely formal dimensions are retained.

It is possible that the multiplication of forms cannot be entirely avoided by any approach to formal analysis; however, the systematic approach proposed in this study minimizes the problem by embedding different forms in an underlying matrix. In the context of the matrix, relationships can be established between apparently discrete executional variables (Metz 1974).

PREVIOUS RESEARCH ON AD FORM

As the discussion in the previous section indicates, if the term form is used in its broadest sense, research on ad form includes all reseach that identifies a shared attribute of ads and explores that attribute's effect on viewers and listeners. Thus, with this broad definition, the body of previous research on ad form is very large.

There is, however, one study which may be taken as a compendium of previous ad form research. In the mid 1980's, MSI funded and published a massive study by Stewart and Furse (1986) of 193 form variables coded for 1,000 television commercials with each commercial being watched and evaluated in a realistic setting by 300-450 respondents. The variables in the study included essentially all aspects of ad form that had previously been identified by either academic researchers or practitioners as having an important effect on ad responses. This project confirmed the utility of studies focused on form, for it found that 13-26 percent of variance in recall and 9-11 percent of variance in persuasion could be accounted for by form variables.

While it has provided and will continue to provide useful guidance to other researchers who focus on ad form, this study, precisely because it is so impressively massive, also illustrates especially well the principal hazard of research focused on form-the multiplication of overlapping and inconsistent forms. Since the coding of variables was guided by no overarching theory, the categorizing scheme used in the study suffered, as Stuart and Furse freely admit, from indeterminate looseness. Citing Kuhn (1970), they acknowledge that, along with all previous studies, their study is pre-theoretical and, therefore, merely descriptive. They suggest that it needs to be superceded by research grounded in a theory that can provide a set of well-defined measures, testable propositions about specific relationships, and a comprehensive statement about what effects are important (Stewart and Furse 1986, p. 9). In focusing on the missing theory of ad form, Stuart and Furse have identified a gap in advertising research that this study will partially fill with semiotics, the science of signs.

SEMIOTICS AND FORM

Semiotics is the general science of signs (Peirce 1931-1958; Saussure 1958). It emerged as a new discipline when the insights and methods of linguistics were applied in new non-linguistic domains. A semiotician's objective is to uncover the hidden foundations on which the meaningfulness of signs rests (Culler 1975; Saint-Martin 1987). These foundations are typically some relatively stable deep structure that grounds and organizes the apparently chaotic surface manifestation of a phenomenon, e.g., the relationships between Saussure's (1959) langue and parole or Chomsky's (1957) deep and surface structures.

The specific class of signs treated in this study is images, in Peirce's (1931-1958) technical terminology, icons. A sign is iconic to the extent that it resembles its referent on some dimension. So all pictures are fundamentally iconic. But despite their close resemblance to the things they signify, pictures also have the properties of a surface structure, i.e., they tend to be more varied and fragmented than the reality they signify. In pictures, sudden shifts in time and space are possible, shifts that don't occur in reality. The fact that ordinary experience of a thing is more stable than audiovisual experience suggests that everyday experience may function as a kind of deep structure against which changes in the surface iconic depiction can be measured.

But on what dimension are direct experience and images in ads to be compared? Ideally, the dimensions used will be applicable to most or all iconic ads and, therefore, will be truly formal, not particularistic. One feature of an image that has the kind of variability that makes it a surface structure attribute but the universality that makes is a critical element of form is point of view. The same object can be viewed from a number of different points of view and can be shown successively from different points of view. Some of those points of view will be normative in that they reflect normal spatial relations relative to that object. Other points of view will be anomalous in that they are rarely encountered in our ordinary interactions with the thing. Marr's (1982) work suggests that the information processing demands posed by normal points of view (e.g., looking at the profile of a car at eye level from a distance of twelve feet) will be lower than those posed by anomalous points of view (e.g., looking at the underside of a car from a distance of twelve feet).

PAST RESEARCH ON ANGLE EFFECTS

When viewing an object, some points of view may be more affectively charged than others. It follows that some camera angles may be more affectively charged, for camera angles reflect spatial position relative to a focal object. Film makers have long believed that camera angles influence affective responses (Giannetti 1982), and Kraft (1987) has provided empirical support for this longstanding belief. He photographed six brief slide show narratives from eye level and 40 degrees above, 40 degrees below eye level. After viewing the narratives, subjects rated actors on various dimensions. Across all stories, actors viewed from 40 degrees below eye level were judged to be taller, stronger, bolder, and more aggressive than those viewed from 40 degrees above. Those viewed from eye level were in the middle on each response dimension. Meyers-Levy and Peracchio (1992) replicated Kraft's (1987) study in an advertising context finding that camera angle influenced responses to computers and bicycles. As in Kraft's study, attitudes were most favorable from the low level angle, and least favorable for the high level point of view. While Kraft and Meyers-Levy and Peracchio demonstrate the importance of camera angles, the domain of their studies is quite restricted. They varied camera angle on only one of three possible dimensions, the vertical axis. In this study, the angle was varied on two dimensions, the vertical (y) and the sagital (z) axes.

RESPONSE THEORIES

Two competing theories that offer opposite predictions on the effects of normal versus abnormal camera angles are tested in this study. The first theory, distraction theory (Festinger and Maccoby 1964, Nelson Duncan, and Kiecker 1993), suggests that abnormal ad executions (unusual camera angles) will be less effective than normal executions (customary angles) because they will distract attention from the ad's product claims and reduce product claim recall. The second theory, the dual-mediation hypothesis (Homer 1990; McKenzie and Lutz 1989; Brown and Stayman 1992) suggests that abnormal executions will be more effective than normal ones, through a two-stage process. First, the greater arousal elicited by the abnormal execution will elevate attitude toward the ad; then that positive AAd will spill over onto product claim processing, producing greater product claim recall.

While distraction theory and the dual-mediation hypothesis lead to diametrically opposed predictions on brand learning and brand attitudes, it is possible that the two theories are not incompatible. Abnormal executions may produce both a distraction effect and an attention spillover affect. Net effects would then depend on the relative magnitude of these negatively and positively valenced effects. If the negative distraction effect were larger, increasing the salience of the execution would reduce learning and brand attitudes. If the positive spillover effect were larger, a more salient execution would lead to enhanced learning and attitude. So which effect is likely to dominate in what circumstances? A third set of competing hypotheses will be offered in this study suggesting that the dominance of the negative distraction and positive attention effects will be moderated by the amount of ad clutter. In low clutter contexts where much attention is likely to be focused on the ad, abnormal executions may be expected to draw attention away from the ad, in other words, produce the negative distraction effect predicted by distraction theory. In high clutter contexts, on the other hand, where attention may already be directed to other ambient stimuli, an abnormal execution may draw attention back to the ad and produce the positive attention effect predicted in the dual-mediation hypothesis. In other words, this theory predicts an interaction between execution normality/abnormality and ad clutter.

INTERNAL AND EXTERNAL EFFECTS

Scott (1994) has pointed out that researchers often ignore the complexity of responses to ads. If in the context of a given ad, viewers see that an unusual camera angle has been used, they may draw unpredictable conclusions about what the angle is meant to suggest in the ad. This means that the internal effects of an ad execution may be confounded by the semantic meanings inferred by viewers. This semantic confound is not likely to exist for proximate ads in the same ad pod. Viewers are unlikely to think that odd angles used in one ad have any significance in a proximate ad. And yet, more basic distraction/attention effects should spill over onto proximate ads since at a less reflective level, the borders between ads are arbitrary. Since semantic confounds can occur only internally in responses to the ad that contains the execution, external effects of the execution on proximate ads may be the better test of classical conditioning and other unconscious attention and attitude effects on ad responses. Both internal and external effects are tested in this study.

HYPOTHESES

The hypotheses presented here follow from distraction theory (first hypothesis), the dual-mediation hypothesis (second hypothesis), and the composite moderator theory proposed in this study (third hypothesis).

H1: Compared with normal camera angles, abnormal angles will produce lower brand learning (and brand attitudes).

H2: Compared with normal camera angles, abnormal angles will produce higher brand learning (and brand attitudes).

H3: The effects of normal/abnormal angles on brand learning (and brand attitudes) will be moderated by the level of ad clutter, such that abnormal executions produce higher brand learning when clutter is high and lower brand learning when clutter is low.

METHOD

These hypotheses were tested in an experiment that had a 2 (normal/abnormal angle) by 2 (high/low ad clutter) design. A convenience sample of 91 university students were used for this theory test. Camera angle was manipulated in an insurance ad featuring three people who had supposedly been helped by a fictitious company called Southwest Mutual Insurance. The actors were shown sitting on a chair not saying anything. A gender and age consistent voice recounted in a voiceover a positive experience each person had had with Southwest Mutual. (The same voiceover narratives were used in all ad versions to avoid a narration confound.) In the normal version of the ad, the camera was at eyelevel about six feet away from the actor. In the abnormal version, it was only a foot away and three feet above the actors head looking down. The normal and abnormal commercials were embedded in ad pods of 5 (low clutter) and 15 (high clutter) commercials.

Dependent measures were taken for the insurance ad (internal effects) and a proximate ad (external effects). Dependent measures included recognition and recall of the product class and brand name and recall of the product claims. All recognition and recall items were reported as 0 where absent, 1 where present, so mean scores (reported in Table 1) represent the percentage of subjects remembering an ad. Ad claim recall was measured by asking subjects to list all ad claims they could remember. Claims were scored by two judges (Cohen's Kappa .87). Mean number of claims per ad version ranged from .74 to 2.98 (Table 1). Attitude toward the brand was measured with a four item scale ranging from 1, very favorable, to 5, very unfavorable) (Cronbach's alpha .89).

RESULTS

Since there were a number of dependent recall and recognition variables, hypotheses were tested using MANOVA. The ad recall/recognition and brand attitude parts of the hypothesis were tested separately. It is possible to conclude that H2 (the attention effect) is more valid than H1 (the distraction effect) based purely on an inspection of the recall/recognition means reported in Table 1 and brand attitude means reported in Table 2. While the internal effects of angle are inconclusive, the external effects are clearly consistent with H2, inconsistent with H1. Nine of the ten external measures of recall/recognition were consistent with H2. And for both internal and external measures, brand attitudes were more favorable (on the 1 favorable, 5 unfavorable scale) for ads with abnormal angles. Still, across all internal and external dependent memory variables, H2 was not supported by the MANOVA test. The main effect was not significant (F(10,164) = .71, p = .72). The lack of an internal effect where semantic confounding is possible may explain this lack of significance, for H2 would seem to be supported by the external effect of the angle manipulation. The probability that nine out of ten external recognition/recall means would be in the direction hypothesized by H2 is low (z = 4.22, p < .001). And while the external effect on brand attitude was not significant that the .05 level, it was significant at the .1 level (t = 1.52, p = .07). H3, the hypothesis that angle and clutter would interact, supporting the distraction hypothesis in low clutter contexts, the dual-mediation hypothesis in high clutter contexts, was not supported.

DISCUSSION

The angle manipulation used in this study was relatively subtle. While the camera was repositioned on more dimensions than in previous research, only one camera angle was used in each version of the ad. Presumably, the effect of the angle form element would be stronger in ads where multiple abnormal camera angles were used than in ads such as these where only a single angle was used. The striking pattern of directional results consistent with an attention effect for external ads suggests that a small effect of form probably does occur as camera angles become more unusual. This effect would seem to be dominated by interpretations of the semantics of the angle in internal ads.

These findings have practical implications for advertising managers. Ads with unusual executions would seem to enhance responses to proximate ads. Positions in an ad pod next to unusual ads would seem, therefore, to have extra impact and extra value. This finding represents an opportunity for informed managers because advertising vendors do not charge higher prices based on which pod of ads a particular ad is shown in or on where in the pod the ad is placed. Vendors should be willing to accommodate requests for a particular position. Thus, until the effects identified in this study become widely known and pod position pricing is changed to reflect the differential impact of different pod positions, managers can enhance advertising effectiveness at no cost to their firm.

REFERENCES

Brown, S. P. & D. M. Stayman. (1992). Antecedents and consequences of attitude toward the ad: A meta-analysis, Journal of Consumer Research, 19 (June), 34-51.

Chomsky, N. (1957). Syntactic Structures, The Hague: Mouton.

Culler, J. (1975). Structuralist Poetics, Ithaca: Cornell University Press.

Festinger, L. & N. Maccoby. (1964). On resistance to persuasive communications, Journal of Abnormal and Social Psychology, 68 (4), 359-366.

Giannetti, L. D. (1982). Understanding Movies, Englewood Cliffs, NJ: Prentice-Hall.

Homer, P. M. (1990). The mediating role of attitude toward the ad: Some additional evidence, Journal of Marketing Research, 27 (February), 78-86.

Kraft, R. N. (1987). The influence of camera angle on comprehension and retention of pictorial events, Memory and Cognition, 15 (July), 291-307.

Kuhn, T. S. (1969). The Structure of Scientific Revolutions, 2nd ed., Chicago: University of Chicago Press.

Marr, D. (1982). Vision: A Computational Investigation into the Human Represenation and Processing of Visual Information, San Francisco: W.H. Freeman and Company.

Marra, J. (1982). Principles of gaining attention, Advertising Techniques, 17 (5), 24-26.

Marketing Sciences Institute. (1988a). Evaluating the effects of consumer advertising on market position over time: How to tell whether advertising ever works, Conference Sponsored by MSI Advertising Steering Group, June 8-10, Center for Executive Education, Wellesley, MA.

Marketing Sciences Instititute. (1988). Research Priorities 1988-1989: A Guide to MSI Research Programs and Procedures, Boston: Marketing Science Institute.

MacKenzie, S. & R. J. Lutz. (1989). An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context, Journal of Marketing, 53 (April), 48-65.

Metz, C. (1974). Film Language: A Semiotics of the Cinema, Chicago: University of Chicago Press.

Meyers-Levy, J. & L. A. Peracchio. (1992). Getting an angle in advertising: The effect of camera angle on product evaluation, Journal of Consumer Research, 13 (September), 196-213.

Nelson, J. E., C. P. Duncan & P. L. Kiecker. (1993). Toward an understanding of the distraction construct in marketing, Journal of Business Research, 26, 201-221.

Pepper, S. (1970). World Hypotheses, Berkley: University of California Press.

Pierce, C. S. (1931-58). Collected Papers, (Eds. Charles Hartshorne, Paul Weiss &Arthur W. Burks, Cambridge, MA: Harvard University Press.

Saint-Martin, F. (1990). Semiotics of Visual Language, Bloomington: Indiana University Press.

Saussure, F. de. (1958). Course in General Linguistics, New York: McGraw-Hill.

Scott, L. M. (1994). Images in advertising: The need for a theory of visual rhetoric, Journal of Consumer Research, 21 (September), 252-273.

Stuart, D. W. & D. H. Furse. (1986). Effective Television Advertising: A Study of 1000 Commercials, Lexington, MA: Lexington Books.

Val Larsen, James Madison University

Newell D. Wright, James Madison University

TABLE 1

Effects of Angle on Brand Information Recall/Recognition

INTERNAL Clutter Normal Abnormal

RECALL

Ad Claims Low .84 [44] .74 [47]
 High .67 [47] .60 [42]
 Total .79 [91] .67 [89]
Product Class Low .50 [44] .57 [47]
 High .40 [45] .33 [42]
 Total .45 [89] .46 [89]
Brand Name Low .27 [44] .34 [47]
 High .22 [45] .19 [42]
 Total .25 [89] .27 [89]

RECOGNITION

Product Class Low .73 [44] .83 [47]
 High .78 [45] .67 [42]
 Total .75 [89] .75 [89]
Brand Name Low .68 [44] .81 [47]
 High .60 [45] .62 [42]
 Total .64 [89] .72 [89]

EXTERNAL Clutter Normal Abnormal

RECALL

Ad Claims Low 2.89 [44] 2.98 [47]
 High 1.96 [45] 1.90 [41]
 Total 2.42 [89] 2.48 [88]
Product Class Low .68 [44] .83 [47]
 High .33 [45] .38 [42]
 Total .51 [89] .62 [89]
Brand Name Low .14 [44] .17 [47]
 High .02 [45] .05 [42]
 Total .08 [89] .11 [89]

RECOGNITION

Product Class Low .98 [44] 1.00 [47]
 High .96 [45] 1.00 [42]
 Total .97 [89] 1.00 [89]
Brand Name Low .25 [44] .30 [47]
 High .27 [45] .29 [42]
 Total .26 [89] .29 [89]

TABLE 2

Effects of Angle on Brand Attitude

INTERNAL Clutter Normal Abnormal

Ad Claims Low 4.41 [44] 4.33 [47]
 High 4.68 [44] 4.41 [42]
 Total 4.54 [88] 4.37 [89]

EXTERNAL Clutter Normal Abnormal

Ad Claims Low 4.48 [44] 4.03 [47]
 High 4.53 [44] 4.46 [41]
 Total 4.51 [88] 4.23 [88]