The Impact of Comparative Advertising on Perception Formation in New Product Introductions

Beth A. Walker, The Pennsylvania State University
John L. Swasy, The Pennsylvania State University
Arno J. Rethans, The Pennsylvania State University
ABSTRACT - The applicability of categorization theory for assessing the effects of comparative advertising is discussed in light of Gorn and Weinberg's (1984) results. Findings of an exploratory study are presented to illustrate the associative and differentiating properties of comparative advertising.
[ to cite ]:
Beth A. Walker, John L. Swasy, and Arno J. Rethans (1986) ,"The Impact of Comparative Advertising on Perception Formation in New Product Introductions", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 121-125.

Advances in Consumer Research Volume 13, 1986      Pages 121-125

THE IMPACT OF COMPARATIVE ADVERTISING ON PERCEPTION FORMATION IN NEW PRODUCT INTRODUCTIONS

Beth A. Walker, The Pennsylvania State University

John L. Swasy, The Pennsylvania State University

Arno J. Rethans, The Pennsylvania State University

ABSTRACT -

The applicability of categorization theory for assessing the effects of comparative advertising is discussed in light of Gorn and Weinberg's (1984) results. Findings of an exploratory study are presented to illustrate the associative and differentiating properties of comparative advertising.

INTRODUCTION

Advertisers are making increasing use of comparative advertising in which reference is made to a named or recognizably presented competitor. Concomitant with this use is an increasing number of empirical studies attempting to assess the effectiveness of comparative advertising, particularly in terms of the differential effects of comparative versus noncomparative communications

The majority of these studies suggests that comparative advertising is no more effective than traditional advertising when it comes to believability (Boddewyn and Marton 1978; Prasad 1976; Wilson 1976; Wilson and Mudderrisoglu 1979), recall (Jain and Hackleman 1978), attitude toward the product (Belch 1981; Goodwin and Etgar 1980), purchase intentions (Golden 1976), and actual behavior (Swinyard 1981). In some instances, in fact, comparative advertising was less effective than traditional advertising (Levine 1976: Shimp and Dyer 1978).

A recent study by Gorn and Weinberg (1984, 1983), however, does provide some support for the continued use of comparative advertising. They suggest that the purpose of most comparative advertising is a reduction in the psychological distance between a challenger and the brand leader. Accordingly, the authors focused on perception formation as the critical dependent variable in their relative effectiveness test. It was found that comparative magazine advertising did significantly increase perceived brand leader-challenger similarity as well as the perceived similarity between all brands in a product class. Given these initial and encouraging findings, the authors called for more research focusing on the effectiveness of comparative advertising on perception formation.

In this paper we extend the Gorn and Weinberg effort by suggesting a theoretical framework for their approach and by illustrating this framework within the context of a study on consumer reaction to comparative television launching commercials for a regional premium beer. Specifically, we propose the use of categorization theory as a framework for analyzing the differential effects of comparative advertising.

REVIEW OF LITERATURE

The communication effectiveness of comparative advertising has been an issue ever since the Federal Trade Commission encouraged advertisers to name competing brands as an alternative to the then more prevalent "brand X" euphemism (Cohen 1976). Wilkie and Farris (1975) proposed that such comparative advertising should be more effective than more traditional advertising in generating increased attention and recall, increased comprehension of claims, and greater yielding to claims. Of the studies investigating the relative effectiveness of comparative advertising, most were not supportive of these initial propositions. Comparative advertisements were often perceived as more confusing, less believable, and no more persuasive than noncomparative advertisements (Belch 1981; Shimp and Dyer 1978; Swinyard (1981).

More positive findings, however, were recently reported by Gorn and Weinberg (1984, 1983), who tested the relative effectiveness of comparative advertising by focusing on perception-related dependent variables. These authors argue that, in practice, comparative advertising is usually started by a challenger with the objective of placing the challenger in the same league as the brand leader in the mind of the consumer. A favorable result, then, for the challenger's comparative ads would be a reduction in the perceived psychological distance between the two brands.

In their study, Gorn and Weinberg focused on perceived challenger-leader brand similarity as the dominant dependent variable. They manipulated type of ad by the challenger (comparative, noncomparative), context (leader ad present, leader ad absent) and product category (cigarettes, golf balls, toothpastes). They found that greater perceived challenger-leader similarity was obtained with comparative advertising than with noncomparative advertising across the three product categories. Type of ad, however, did not affect attitude toward the product, a finding similar to previous research results. This latter result led these researchers to speculate that attitude may not be the most sensitive dependent measure and that longer term exposure might be necessary to affect attitude since attitude is a higher order response. Cognitive response data did not illuminate possible underlying processes which might have shaped and determined these results. As a result, Gorn and Weinberg called for more theoretical and empirical research on perceived challenger-leader similarity as an indicator of comparative advertising's effectiveness.

One promising theoretical framework for the analysis of perceived brand similarity focuses on the cognitive processes of categorization and prototyping (Rosch 1975, 1978). Casual observation as well as recent research attest to the pervasive human tendency to categorize both objects and persons into groups, types and other categories so that non-identical stimuli can be treated as if they were equivalent (Cantor and Mischel 1979; Rosch, Mervin, Gray, Johnson and Boyes-Braem 1976). Categorization simplifies the individual's processing task of the potentially overwhelming number of stimuli in the environment. Research further indicates that knowledge in a given category is internally structured around a typical (i.e., exemplar) or ideal instance (i.e., prototype) which captures the meaning of the category (Cantor, Mischel and Schwartz 1982). Categorization of a new stimulus under this perspective is accomplished by a prototype-matching process and is a function of the degree of similarity between the new stimulus and the category exemplar or prototype. In this way, categorization influences the processes of attention, integration, storage and retrieval of information. [The issue of whether individuals perceive categories based on exemplars or based on prototypes has been debated in the categorization literature (Cantor and Mischel 1979; Medin and Smith 1984). For our purposes, no strong distinction is made between these two approaches. Rather we follow Elio and Anderson (1981) and assume that either the exemplar or the prototype or both provide the consumer with a set of expectations (cf. Sujan 1985).]

The potential usefulness of the categorization perspective in examining consumer reaction to marketing stimuli was recently demonstrated by Sujan (1985). The results of her study suggest that category-based expectations guide consumer evaluation processes. In particular, her research suggests that consumers have well-defined categories stored in memory which enable them to perceive and evaluate new products. When applied to an analysis of comparative advertising, Sujan's findings suggest that by making reference to particular comparison brands, the consumer may categorize the new brand as an example of this category and retrieve the affect associated with the activated category. As a result, the categorization perspective is viewed as interesting paradigm for understanding comparative advertising effects.

STUDY HYPOTHESES

The primary objective of the current study was to explore the impact of comparative television advertising on consumer perceptions of a new entry into a product category. Specifically, the goal of the advertising campaign to be reviewed here was to associate a new brand of beer with the leading prototypes in the premium beer segment. [The comparison brands contained desirable or positive features. Therefore, the objective of the comparative advertisement was association rather than differentiation.] Cast within the categorization framework this campaign goal translates into the following hypotheses:

H1a: Perceived similarity between the new entry and prototype brands will be greater with exposure to a comparative (vs. noncomparative) television commercial.

H1b: Perceived dissimilarity between the new entry and non-prototype brands will be greater with exposure to a comparative (vs. noncomparative) television commercial

STUDY METHODOLOGY

Experimental Design and Procedure

The part of a larger experimental study examined here can best be represented as a simple two-group comparison of comparative versus noncomparative advertising. The larger study was designed to measure the effectiveness of various types of launching commercials for a new regional premium beer. Since the study was conducted before the actual introduction of the new brand, it permitted a "clean" assessment of comparative advertising's impact, free of the confounding effect of prior brand knowledge.

Upon arrival at the test site, pre-screened subjects were led to a television viewing room ostensibly to evaluate different types of news programming. Subjects were told that they would watch some news programming for about thirty minutes and would be asked some questions following the presentation. In addition, they were informed that the programming would be interrupted periodically so that they could report their thoughts, ideas, and opinions on the segment preceding the interruptions. To overcome the potential inadequacies of the single message exposure noted by Gorn and Weinberg (1984), the experimental commercial was embedded three times within the total news programming effort. Following the television programming subjects responded to the dependent variable measures contained in a questionnaire booklet and then were debriefed and paid.

Experimental Stimuli

Professionally produced 30-second television launching commercials for a new regional premium beer were used as the experimental stimuli in this study. The comparative version of the ad featured implicit comparisons between the new brand and three well known premium brands. In the opening and closing scenes of the ad, the three competitor brand bottles were shown turning and transforming into three bottles of the new beer. These visual comparative cues were accompanied by an audio track stating "Now there is a beer with a taste that will turn you around." To provide a reference point for judging the effects of the comparative at. a second introductory ad was used. This "noncomparative" version of the ad, opened with a scene of a glass blower creating a new bottle in his 19th century shop conveying a sense of tradition and quality, followed by close-up shots of the new beer's bottle and label.

The use of actual, professionally developed television commercials precluded the manipulation of the comparative cues exclusively. We recognize that the inability to manipulate the presence or absence of the comparative cues reduces the internal validity of the study results. Additional data collection measures were therefore taken to address alternative explanations.

Dependent Variables

The categorization approach to an association and differentiation strategy of comparative advertising suggested the need for a careful definition of the prototypes against which the new product was to be compared. Discussion with the client's advertising agency personnel revealed that Budweiser, Strohs, and Miller were perceived to be the relevant exemplars for the new premium beer. A focus group conducted with members of current research population confirmed this belief. Further, the three most frequently mentioned brands constituting the super premium beer category were Michelob, Heineken and Lowenbrau. Accordingly, semantic differential scales employing the anchor points of "low similarity" and "high similarity" were used to measure similarities to each of the premium and super-premium beers. A similarity and dissimilarity score for each category was then derived by averaging responses on the three respective scales.

The cognitive responses reported during the programming and after the test commercials were coded so as to assess the frequency of responses that were comparative in nature.

Subjects

To ensure a minimum ability to categorize beer-related stimuli, study participants were pre-screened for their level of knowledge and experience with the product category of interest. Based on responses to a pre-screening questionnaire covering usage patterns for a variety of product and media habits, non-beer drinkers were excluded from the study. A total of forty student volunteers at a major Northeastern university participated in the study and were paid $5.00 for their participation.

RESULTS

Perceived Brand Similarity

Subjects exposed to the comparative advertisement perceived the new beer brand to be more similar to premium beers than did subjects exposed to the noncomparative advertisement (4.25 vs. 3.78). Although these means were in the predicted direction, their difference was not statistically significant, (F(1,38) = 1.13, p=.29). As predicted, however, similarity to super-premium beers was significantly affected (F(1,38) = 4.63, p < .03). Subjects in the comparative advertisement condition perceived the new brand to be less similar to the super-premium brands (2.90 vs. 4.10).

One alternative explanation to the above interpretation of the findings is based on the nonequivalence of the information in the two ads. Because the visual and audio contents of the treatment advertisements were not identical (i.e., beyond the presence or absence of the comparative cue), it is not possible to definitively attribute the brand perception findings exclusively to the brand comparison cues. Hence, to provide more evidence that the visual comparative cues affected consumer's ad processing, additional analyses were performed.

First, to determine if the perception results were influenced by the general visual portion of the advertisement, a control group receiving the audio portion of the ad was conducted. The "radio commercial" was embedded in a news programming environment similar to that experienced by our treatment groups. Thus, the comparative television commercial (comparative visual cues, other visual cues and audio cues) was compared to a radio ad (audio only, in which the brand comparison was not mentioned). It was hypothesized that exposure to the television (vs. radio) commercial would result in greater perceived similarity between the new brand and premium brands, and less similarity to super-premium brands. As predicted, a significant effort on similarity to premium beer brands (F(1,38) = 4.75, p < .04) was observed. Subjects viewing the television advertisement perceived the new beer brand to be more similar to premium beer brands than subjects in the radio condition (4.25 vs. 3.37) (l=low, 7-high similarity). For the super premium category index, the means were in the expected direction, although the difference was not statistically significant (F(1,38) = 1.65, p=.21). An examination of the three super premium brands separately revealed significant differences in similarity with Heineken (2.45 (tv) vs. 3.65 (radio), F(1,38) = 3.76, p=.06) and Lowenbrau (2.55 (tv) vs. 3.75 (radio), F(1,38) = 3.24, p=.08), but not for Michelob (3.80 (tv) vs. 3.60 (radio), F(1,38) = 0.10, p= .75). These results suggest that the effects on perceived similarity are attributable to some visual cue(s). Unfortunately, we cannot discern from this analysis if the effects are due specifically to the comparative visual cues.

Secondly, to more clearly identify the role of the comparative visual cues, cognitive responses were analyzed to assess whether the comparative cues captured the viewers' attention. The responses were examined for explicit references to the comparative nature of the premium beer brands. Seventy-five percent of the subjects who viewed the comparative television advertisement reported at least one comparative cognitive response. Only 10% of the subjects in the noncomparative treatment condition recorded one or more comparative thoughts. In combination, these results increase our confidence that the visual comparative cues affected the categorization of the new brand. We recognize that tue to the confounding visual cues, however, the effect of the comparative cues cannot be unambiguously determined. Further research is needed to determine specifically whether the surrounding visual cues hindered or enhanced the impact of the comparative visual cues on perception formation.

DISCUSSION AND FUTURE RESEARCH

In this paper we proposed the potential usefulness of categorization perspective for the analysis of the impact of comparative advertising on perception formation in new product introductions. In such situations, the objective of the attending advertising campaign frequently is to position the new entry in the same league as a named or recognizably presented competitor in the mind of the consumer (Prasad 1976). Under these conditions, the prototype-matching processes suggested by the categorization approach forms an appealing explanation for consumer reactions to this type of advertising.

A study using launching commercials for a new regional premium beer was used as an initial exploration of the proposed categorization framework. In contrast to many earlier studies, the comparisons of the new brand in this study were implicit rather than explicit and involved multiple comparison brands rather than a single market leader brand. The data were found to be generally consistent with the proposed framework and suggest that further tests of the theory are warranted.

Several interesting research areas are highlighted. Comparative advertising has been shown to reduce the psychological distance between the sponsored brand and the brand to which it is compared. What psychological distance are we talking about? What is the nature of this distance? That is, what specific beliefs may be affected? In a challenger/leader comparison, two different psychological processes may result depending on the extent of information evoked by the market leader. If the brand leader is processed as an individual brand not associated with a category, the reduction of psychological distance may be with respect to the challenger brand and the particular brand leader. If the brand leader is processed as a category exemplar or prototype, however, the reduction in psychological distance is more likely to be between the challenger brand and a product category. The specific beliefs that are generalized to the challenger brand may depend on which type of processing is evoked.

If the psychological distance is reduced between the challenger brand and the leader as processed on an individual brand basis, the features and attributes associated with that particular brand will be generalized to the challenger. If the psychological distance is reduced between the challenger and a product category, the beliefs associated with the product category will be associated with the new brand. The extent to which specific brand beliefs differ from beliefs associated with a product category will determine how differently the new challenger brand will be perceived by the consumer depending on the resulting process. For example, a new car may be compared to a Mercedes Benz. If Mercedes Benz is processed as an individual brand and not as a representative of a "luxury car" product class, the new car may be perceived as an import with high quality, handling, and durability. If the Mercedes Benz is processed as a prototype of the product class luxury cars, however, the new car may be perceived as big, comfortable, smooth ride, and as a gas guzzler. The beliefs associated with the new car would depend on how the comparison brand is processed. The set of beliefs that the advertiser wants to create for the new brand would determine the most appropriate prototype or exemplar selected for comparison.

A second issue concerns the use of multiple brand comparison ads. What is the effect when a challenger is compared to more than one other brand? The consistency/inconsistency within the group of comparison brands may affect the consumer's processing and resulting perceptions of the challenger or sponsored brand. A consistent comparison set is defined here as one that is internally homogeneous and congruent with a consumer's ideas regarding category membership. An inconsistent comparison set is less homogeneous among themselves, but is still representative of the product class category. If the category set of comparison brands is consistent, category based processing is likely to result. The beliefs associated with the new brand will be that of the category. In contrast, if the set of comparison brands is inconsistent, piecemeal processing may result. For example, if a multiple brand comparison group included Heineken, Lowenbrau, and Budweiser, the beliefs about the sponsored brand are not likely to be the result of category based processing but are determined on a piecemeal basis. The reduction of psychological distance that occurs may only be belief specific depending on the degree of inconsistency. If the brands overlap on one attribute, that attribute may be transferred to the challenger brand.

The consistency/inconsistency of the multiple brand category may also mediate the influence of the comparative ad on attitude towards the new brand. If the category is consistent, category based processing may result and the affect associated with that category may be transferred to the challenger brand (cf. Sujan 1985). If the category is inconsistent, piecemeal processing may result. Therefore, attitude towards the new brand may be specific to the processed features or attributes of the comparison brands. The impact of ad type on attitude may also depend on the valence of the attitude toward the activated category. If the attitude towards the category is strongly positive or negative, we may be more likely to see a comparative ad effect on brand attitude than if the subject had a neutral initial attitude toward the category. The effects of consistent/inconsistent categories assumes a certain level of consumer knowledgeability about the advertised product class. It is the consumer who judges the heterogeneity of a product class as well as the consistency of the subcategory is within it. Consumer knowledgeability, then, also may play a role in mediating the impact of comparative advertising.

The third issue concerns several of Gorn and Weinberg's findings regarding psychological distance reduction and the categorization framework. Specifically, Gorn and Weinberg (1984) asked subjects to rate the similarity of the challenger/leader brands. Higher similarity responses were reported by subjects in the comparative (vs. noncomparative) ad conditions. This reduction in psychological distance is consistent with the categorization framework. Subjects were also asked to rate the similarity between two fictitious brands of the same product class (i.e., golf balls, cigarettes, toothpaste) as well as between two brands for unadvertised product categories (i.e., cola and deodorants). One issue is whether exposure to a series of comparative ads leads consumers to view the entire product class as more homogeneous. Secondly, whether this comparative advertisement exposure effect generalizes to perceptions of other unadvertised product classes. Higher similarity responses were reported between the two fictitious brands by subjects in the comparative (vs. noncomparative) ad conditions. It is difficult to explain why two new hypothetical brands would be seen as more similar following comparative ad exposures. Perhaps the comparative advertisement focused attention on the homogeneity of the entire product class rather than on distinctive subcategories. Also, the advertised product classes may be perceived as homogeneous instead of as a heterogeneous set of sub-categories. Gorn and Weinberg also demonstrated that this effect generalized for "colas" (i.e., cola brand "A" and cola brand "B") but not for deodorants (i.e., brand "A" roll-on and brand "B" aerosol spray deodorant). Perhaps the question's cue context affected these results by priming or directing the level of categorization. That is, colas were perceived as more similar because they were categorized at a more general or inclusive level (Cantor and Mischel 1979). In contrast, the deodorants were perceived as different because the question distinguished the form of the deodorants, forcing categorization at a more specific or exclusive level. If a comparison between diet cola A and caffeine-free cola B, and deodorant "A" and deodorant "B" had been-used instead, we suspect that Gorn and Weinberg may have found different results.

An interesting pattern of perceived similarity results presented in table two in Gorn and Weinberg's paper may also be explained by the level of categorization at which the brands were processed. In each case, the two fictitious brands were perceived as being more similar to each other than the challenger/leader brands presented in the advertisement. Respondents had no information about the two fictitious brands besides very general product category information. At the product category level, then, these brands were perceived as similar to each other. Subjects had much more specific information about the challenger and leader brands. The two brands are likely to be similar on some aspects but different on others. This more specific information probably served to differentiate the brands Consequently, categorization occurred at a more specific level and these brands were perceived as less similar to each other than the fictitious brands.

One important issue underlying this discussion and research is that the accurate determination of the exemplars or prototype that represent a category is essential to assess the impact of comparative advertising in association and differentiation between brand categories. This issue became clear to us when analyzing the data for the super premium brands of beer. Identifying the bases that consumer's use for representing categories (i.e., prototype matching, exemplars, mixed models) is difficult and has not been resolved in the categorization literature. The particular bases used to represent a category, however, may influence the selection of prototypes or exemplars used in the similarity/dissimilarity measures. In this research we used focus group interviews to determine the representative brands in a category. In future research multidimensional scaling techniques and/or clustering (Cantor and Mischel 1979) may be used to more accurately determine consumers' category structure.

REFERENCES

Belch, George E. (1981), "An Examination of Comparative and Noncomparative Television Commercials: The Effects of Claim Variation and Repetition on Cognitive Response and Message Acceptance," Journal of Marketing Research, 18 (August), 333-349.

Boddewyn, J. S. and K. Marton (1978), "Comparison Advertising and Consumers," Journal of Contemporary Business,

Cantor, N. and W. Mischel (1979), "Prototypes in Person Perception," in L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 12, New York: Academic Press.

Cantor, N., M. Mischel and J. C. Schwartz (1982), "A Prototype Analysis of Psychological Situations," Cognitive Psychology, 14, 45-77.

Cohen, S. E. (1976), "Widespread FTC Probe Will Seek Codes that Hinder Comparative Ads," Advertising Age, 47 (February 23), 1, 144.

Golden, Linda (1979), "Consumer Reactions to Explicit Brand Comparisons," Journal of Marketing Research, 16 (November). 516-532.

Goodwin, Stephen and Michael Etgar (1980), "An Experimental Investigation of Comparative Advertising: Impact of Message Appeal, Information Load and Utility Product Class," Journal of Marketing Research, 17 (May), 187-202.

Gorn, Gerald J. and Charles B. Weinberg (1983), "Comparative Advertising: Some Positive Results," in Advances in Consumer Research, Vol. 10, Richard P. Bagozzi and Alice H. Tybout (eds.), Ann Arbor, MI: Association for Consumer Research. 377-380.

Gorn, Gerald J. and Charles B. Weinberg (1984), "The Impact of Comparative Advertising on Perception and Attitude: Some Positive Findings," Journal of Consumer Research, 11 (September), 719-727.

Jain, Subhash C. and Edwin C. Hackleman (1978), "How Effective is Comparison Advertising for Stimulating Brand Recall," Journal of Advertising, 7, 20-25.

Levine, Philip (1976), "Commercials that Name Competing Brands," Journal of Advertising Research, 16 (December),

Prasad, V. Kanti (1976), "Communications-Effectiveness of Comparative Advertising: A Laboratory Analysis," Journal of Marketing Research, 13 (Hay), 128-137.

Rosch, E. (1975), "Cognitive Representations of Semantic Categories," Journal of Experimental Psychology: General, 104, 192-233.

Rosch, E., C. B. Mervis, W. D. Gray, D. M. Johnson and P. Boyes-Braem (1976), "Basic Objects in Natural Categories," Cognitive Psychology, 8, 382-439.

Rosch, E. (1978), "Principles of Categorization," in E. Rosch and B. B. Lloyd (eds.), Cognition and Categorization, Hillsdale, NJ: Erlbaum, 1978.

Shimp, Terence and David Dyer (1978), "The Effects of Comparative Advertising Mediated by Market Position of the Sponsoring Brand," Journal of Advertising Research, 3 (Summer), 13-19.

Smith, Robert E. and William R. Swinyard (1983), "Attitude Behavior Consistency: The Impact of Product Trial Versus Advertising," Journal of Marketing Research, 20 (August), 257-267.

Sujan, Mita (1985), "Consumer Knowledge: Effects of Evaluation Strategies Mediating Consumer Judgments," Journal of Consumer Research, 12 (June), 31-46.

Swinyard, William R. (1981), "The Interaction Between Comparative Advertising and Copy Claim Variation," Journal of Marketing Research, 18 (May), 175-186.

Wilkie, William L. and Paul W. Farris (1975), "Comparison Advertising: Problems and Potential," Journal of Marketing, 39 (October), 7-15.

Wilson, R. Dale (1976), "An Empirical Evaluation of Comparative Advertising Messages: Subjects' Responses on Perceptual Dimensions," in Jerry C. Olson (ed.), Advances in Consumer Research, Vol. 7, Ann Arbor, MI: Association for Consumer Research.

Wilson, R. Dale and Aydin Muderrisoglu (1980), "An Analysis of Cognitive Responses to Comparative Advertising," in Advances in Consumer Research, Vol. 7, Jerry C. Olson (ed.). Ann Arbor. MI: Association for Consumer Research.

----------------------------------------