Subject/Referent Positioning in Comparative Advertising: a Pilot Study

Michael D. Johnson, The University of Michigan
David A. Horne, The University of Michigan
ABSTRACT - This paper presents a pilot study that examines the use of comparative adverting in accomplishing certain product positioning objectives. It is hypothesized that using a product with a large number of distinctive features, such as a market leader, as a subject or a referent in comparative ads may systematically affect perceptions of product similarity. The results support the hypothesis and suggest the need for-further research.
[ to cite ]:
Michael D. Johnson and David A. Horne (1987) ,"Subject/Referent Positioning in Comparative Advertising: a Pilot Study", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 164-167.

Advances in Consumer Research Volume 14, 1987      Pages 164-167


Michael D. Johnson, The University of Michigan

David A. Horne, The University of Michigan


This paper presents a pilot study that examines the use of comparative adverting in accomplishing certain product positioning objectives. It is hypothesized that using a product with a large number of distinctive features, such as a market leader, as a subject or a referent in comparative ads may systematically affect perceptions of product similarity. The results support the hypothesis and suggest the need for-further research.


Even a casual observation of current promotional practices reveals an extensive use of explicit comparative advertising. The practice is readily apparent in all media for a number of both products and services. the popularity of comparative advertising is particularly interesting in light of the often conflicting and generally negative empirical evidence regarding its effectiveness (for reviews see Lincoln and Samli 1979, Ash and Wee 1983, and Shimp and Gresham 1983). This discrepancy may be due, at least in part, to the lack of a unified conceptual framework regarding the goals of comparative advertising. Initially, the use of explicit comparisons in advertising was driven by a vague, regulatory posture (Federal Trade Commission circa 1971) rather than theoretical developments. As a result, a very wide range of variables have been used across studies to measure the effectiveness of comparative ads. Most importantly, variables relating to the promotional goals of the advertiser have taken a back seat.

This paper reports the results of a pilot study designed to test a conceptual approach to comparative advertising that emphasizes specific promotional objectives. the approach views comparative advertising as a means to accomplish certain product positioning goals (Wilkie and Farris 1975; Shimp and Dyer 1978; Gorn and Weinberg 1984). With these goals in mint, a related psychological framework based on consumer perceptions of proximity (Tversky 1977) is used in an attempt to construct effective comparative ads. After reviewing previous approaches, we outline our own approach and its implications for ad development. the pilot study and results are then described.


From Cognitive Responses to Positioning

Although a number of empirical studies of comparative advertising have been conducted in the last ten years (see reviews cited above), few offer useful theoretical explanations of existing research results. It is important to re ember that conceptual or theoretical developments were not the driving force behind this research. Certain authors' have, however, provided important insights into how comparative ads are processed by consumers.

Belch (1981), for example, applied a cognitive response approach (Wright 1973) to the use of one-sided versus two-sided comparative advertising messages. He argued that two-sided messages should be sore effective than one-sided arguments because counter arguments are less readily available. Moreover, the relatively positive effect of two-sided claims should increase with exposure as the positive thoughts generated have more opportunity to sink in or affect attitudes. Murphy and Amundsen (1981) focused on the issue of new brand utilization of comparative claims using a cognitive dissonance approach. They suggested that consumers ignore or town play attitude disrupting, or dissonant, information. Because a new brand may claim equality or dissonance approach. They suggested that consumers ignore or down play attitude disrupting, or dissonant, information. Because a new brand may claim equality or superiority over more familiar brands, this may be met with resistance or skepticism. Alternatively, however, these authors mention the possibility that new brands may, in fact, encounter fever negative cognitive responses (using Wright's framework). Because opinions regarding nev products have yet to be formed, there is little basis for counterarguing. This idea was suggested originally by Shimp and Dyer (1978). Also in the context of the cognitive response framework, Swinyard (1981) linked active counterarguing to attribution theory. Accordingly, whether or not an advertiser is seen as either truthful or self-serving determines the strength of the counter arguments.

While these studies provide some basis for understanding, the generally negative research results on comparative advertising remain inconsistent with generally positive practitioner attitudes reported elsewhere (Hisrich 1983). This suggests that the approaches described above fall short. Consistent with more recent theoretical approaches (Gorn and Weinberg 1984), we feel it is important to focus on the promotional and related marketing strategy goals that comparative ads may be used to accomplish.

As originally discussed by Wilkie and Farris (1975), the important marketing goals of comparative advertising concern product positioning. Two particular positioning goals may be facilitated through the use of comparative ads: association and differentiation. Direct comparisons may be used to show consumers that particular products are either very similar, very different, or similar in some ways and different in others. Unfortunately, relatively few empirical studies of comparative advertising have emphasized positioning (Shimp and Dyer 1978; Pride, Lamb, and Pletcher 1977; Gorn and Weinberg 1984).

Viewing comparative advertising as a positioning tool focuses attention on a very direct measure of comparative advertising effectiveness. Whether the goal is association or differentiation, effectiveness can be measured by the perceived proximity or similarity of the products involved. Gorn and Weinberg (1984) focused specifically on the psychological distance between brands as a result of comparative advertising ant, in fact, obtained positive results. In their study, comparative ads by a challenger brand resulted in an increase in perceived similarity between the challenger and a market leader when contrasted with noncomparative ads. Very simply then, the important question is whether or not a comparative advertisement affects the perceived similarity or dissimilarity of the sponsor's product in the desired direction.

Gorn and Weinberg's results suggest that comparative ads facilitate association. They to not, however, discuss types of comparative ads that may either be particularly effective at increasing association or may, in fact, differentiate products. If perceived similarity is the dependent measure of interest, it is important to understand how consumers judge similarity in order to explore this question.

The Contrast Model

Tversky (1977) provides a psychological model of similarity, called the contrast model, that has important implications for marketers interested in comparative advertising (Johnson 1986). According to Tversky, we judge similarity by contrasting common and distinctive features. While common features add to similarity, distinctive features detract. (For a detailed discussion see Tversky [19771 or Johnson [1986].) An important aspect of Tversky's model is that the task environment affects the relative influence or weight of common and distinctive features toward a judgment.

A particularly important task factor is whether a product is used as a subject or a referent in comparisons of the form -how similar is a to b? (where a is the subject and b is the referent). Subject/referent position can systematically influence perceived proximity. Johnson (1981), for example, found consumers' judgments of the similarity of Shasta Cola to Coke to be larger than judgments of the similarity of Coke to Shasta Cola. The contrast model explains such asymmetries in judgment by the weights placed on the different feature sets. Judgments made in a subject/referent format emphasize the features of the subject. When two stimuli differ in the size of their distinctive feature sets, subject/referent asymmetries may result. In the Shasta/Coke example, because Coke has more distinctive features than Shasta, Shasta is judged as more similar to Coke than Coke is to Shasta. Using Coke as a referent rather than a subject in comparison reduces the impact of Coke's distinctive features on perception. Subject/referent asymmetries of this type are more likely the more the products being compared differ in the size (or the salience) of their distinctive feature sets (Johnson 1986).

In light of the contrast model, the positioning goals of comparative advertising can be restated as an attempt to control the weight consumers place on common and distinctive features when judging proximity. Advertisers pursuing an association strategy should emphasize common features. Those pursuing a differentiation strategy should emphasize distinctive features. One very direct way to influence these weights is by taking advantage of subject/referent asymmetries. The use of subjects and referents in comparative advertising is a controllable advertising variable. Consider a new or existing product that has relatively few distinctive features in memory compared to an established product. If the goal is association, the distinctive aspects of the established brand may be de-emphasized by using the established brand strictly as a referent rather than a subject in the ad. If one is attempting to differentiate the new product, it may be wise to focus attention on or use the established product as a subject rather than a referent. This will effectively emphasize the established product's distinctive features in order to differentiate the two products .

Two concrete examples will help illustrate the two strategies. In an attempt to reaffirm the differences between Seven-Up and soft-drinks such as Coke and Pepsi, Seven-Up recently used ads that focused initially on Coke and Pepsi before making any reference to Seven-Up (to emphasize that Coke and Pepsi have caffeine while Seven-Up never had it, never will!"). Alternatively, in an attempt to associate and compete directly with Budweiser, Miller used an introductory campaign for Meister Brau beer in which Budweiser was used strictly as a referent in comparative ads ("Meister Brau tastes as good as Budweiser, at a better price"). It is also interesting that the positive results for comparative ads in the Gorn and Weinberg study were obtained using a relatively subject/referent format. In each of these authors' ads, the challenger brand was used more as a subject and the leader was used more as a referent. Consistent with the predictions of the contrast model, the comparative ads produced a high level of perceived proximity. > e only empirical comparison, however, was with noncomparative ads. Systematic differences in layout, particularly the subject/referent position of the products in the ads, were not studied.

Again, comparative ads should be viewed as a positioning tool. Using the contrast model as a guide, our concern is specifically with how the construction of comparative ads affects perceived position. Stated in research form, the hypothesis tested in the pilot study below is the following:

[H1 ]: When the products in a comparative ad differ in associated distinctive features, using the product with more distinctive features as a referent in comparison facilitates a strategy of association while using the sane product as a subject using the same product as a subject facilitates differentiation.

To summarize, comparative ads may be an effective way to achieve product positioning goals. A direct measure of effectiveness in this case is the perceived proximity of the products in question. Moreover, control over the subject/referent position of products in a comparative ad may facilitate an increase or decrease in perceived proximity.


Method and Design

As an initial test of the hypothesis, a pilot study was conducted involving the advertising of a new (hypothetical) cola soft-drink named "Sparkle". The market leader, Coke, was chosen as the comparison product to assure a difference in associated features. (The study was conducted prior to the introduction of New Coke). The subject/referent position of Sparkle and Coke was the only independent variable manipulated in four test ads. Two ads used Coke as the subject and Sparkle as the referent while two others used Sparkle as the subject and Coke as the referent. Two ads were used in each subject/referent condition to control at least somewhat for the effect of any particular execution of the subject/referent operationalization. In all four ads, three characteristics or aspects of Sparkle were advertised (Sparkle has half the calories, half the caffeine. and more cola taste than Coke).

In the first version of the Sparkle-subject/Coke-referent ads (Ad 1), a can of Sparkle appeared on the left with a copy line that read Sparkle, the Cola for You (at the top of the ad) ... when Coke is just too such" (at the bottom of the ad). The three attributes were listed just to the right of the can (in the middle of the ad). In a second version (Ad 2), a can of Sparkle - was presented along with the attribute information to the right of the can with a short copy line at the top of the ad reading, "Less....Yet More". In the first version of the Coke-subject/Sparkle-referent ads (Ad 3), two cans were shown, Coke on the left and Sparkle on the right. The copy line read Is Coke too much? (at the top) and Sparkle ... the Cola for You (at the bottom). The attribute information was listed in between the two cans in the middle of the page. In the second version of the ad (Ad 4), only a can of Coke was presented to the left. The copy line at the top read Too little with too much?" followed by What has half the calories, half the caffeine, and more Cola taste?" to the right of the can, with "Sparkle. . .Less but More" at the bottom of the ad.

Objectively, in all four ads Sparklest attributes place it somewhere in between regular colas (Coke as well as Pepsi and RC) and diet, caffeine free colas (e.g. Diet Caffeine Free Coke, Diet Pepsi Free). If the existence of asymmetric subject/referent perceptions is meaningful in a comparative advertising context, whether Sparkle or Coke are subjects or referents in an ad should result in very different positions. According to the contrast model, because Coke has so many distinctive features, using Coke as a subject in comparison with Sparkle should position Sparkle farther away from Coke (and, perhaps, the other regular colas) than when Coke is used only as a referent. Using Coke as a subject should bring to mind many of Coke's distinctive features while using Coke as a referent should lower the weight consumers place on Coke's features.


To test the hypothesis, a story board version of each ad (done professionally and in color) was shown to a separate group of student subjects in a classroom situation (ns 20, 17, 18, and 22 for the four ads respectively). the ads were shown using a slide projector for one minute. As discussed above, the relevant dependent measure for the marketing manager is the effect of the different ads on the perceived proximity (relative position) of the new soft-drink. After exposure to the ads, subjects in each condition were asked to rate the overall similarity among a group of twelve different cola soft-drinks that included both Coke and Sparkle on a seven po int scale from 1 (very dissimilar) to 7 (very similar). These judgments were averaged across subjects within groups for analysis.

Analysis and Results

The main hypothesis [H1] is that the similarity of Sparkle and Coke will be higher, or Sparkle will be positioned more closely to Coke, when Sparkle is used as a subject rather than a referent in comparison. The average judgments across the ads support this hypothesis. When Sparkle was the subject, the mean similarities for Sparkle and Coke were 3.63 and 3.80 respectively for Ads 1 and 2. The corresponding figures for Ads 3 and 4, where Coke was the subject and Sparkle the referent, were 3.17 and 3.31. However, as similarity judgments are inherently relative, it is problematic to interpret these results out of context. To order to judge the significance of the results these differences must be viewed relative to the judgments for all the brands in the market.

To examine the perceived position of Sparkle that resulted from the ads, the similarity judgments for all possible pairs of colas were analyzed using an additive tree procedure (ADDTREE; Sattath and Tversky 1977). the length of the branches in an additive tree that separate different brands represent the perceived distance between the brands. Such an analysis helps determine the particular competitive structure of a market as perceived by consumers (i.e. brands on the same branches are more competitive than brands on different branches; Urban, Johnson, and Hauser 1984). Analysis of the resulting trees for each ad condition revealed differences in perceived market structure across the subject referent conditions but not across ads within each condition. As a result, similarity judgment i were averaged across the two ads in each condition for analysis. The resulting trees for the Sparkle-Subject/Coke-Referent and Coke-Subject/Sparkle-Referent ads are shown in Figures A and B respectively (Kruskal's stress values of .061 and .058).

The results reveal an important difference in the perceived position of Sparkle in the two conditions which supports the hypothesis. When Sparkle is the subject of the comparison (Figure A), it is perceptually more competitive with Coke and the other regular colas than when Coke is the subject of the ad (Figure 8). Apparently, the Sparkle-Subject ads were more successful in lowering the weight on the distinctive features of Coke than the Coke-Subject ads. The relative positions of Coke and Pepsi in the two solutions provide additional support for this interpretation. In the Coke-Subject ads, Coke and Pepsi are on different branches of the tree while in the Sparkle-Subject ads Coke and Pepsi are on the same branch. Put differently, when Coke is the subject of the ad, Coke has a more unique position in the tree.


Overall, this research-presents a potentially useful approach for practitioners and researchers interested in comparative advertising. the contrast model provides guidelines for the construction of comparative ads to meet particular promotional goals. In particular, the subject/referent position of brands in an advertising comparison may have some effect on the perceived position of brands in a market. At minimum, the pilot study results reported here suggest that the hypothesis deserves further research attention. Within the context of a controlled experiment, the results were as predicted. Naturally, one can not make conclusions on the basis of one relatively artificial, small scale study. In a larger scale study, the authors are currently exploring the effects of subject/referent positioning in comparative advertising using more realistic methods and exposure conditions.






Ash, S. and Wee, C. H. (1983). "Comparative Advertising: A Review With Implications for Further Research," in R. Bagozzi and A. Tybout (eds.), Advances in Consumer Research, Association for Consumer Research, 10, 370-376.

Belch, George (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.

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

Hisrich, Robert D. (1983). Executive Advertisers' Views of Comparative Advertising," Sloan Management Review, (Fall), 39-50.

Johnson, Michael D. (1981). Context Effects in Product Perception, in [. Monroe (et.), Advances in Consumer Research, Association for Consumer Research, 8, 112-115.

Johnson, Michael D. (1986). Consumer Similarity Judgements: A Test of the Contrast Model, Psychology and Marketing, 3 (1), 47-60.

Lincoln, D. and Salmi, A. C. (1979). Empirical Evidence of Comparative Advertising's Effects: A Review and Synthesis," Proceedings: Educators' Conference, American Marketing Association, 367-372.

Murphy, J. and Amundsen, M. (1981). "The Communications - Effectiveness of Comparative Advertising For A New Brand on Users of the Dominant Brand," Journal of Advertising, 10 (1), 14-20 & 48.

Pride, W., Lamb, C., and Pletcher, B. (1979). Are Comparative Advertisements More Informative for Owners of the Mentioned Competing Brand Than for Nonowners?" Proceedings: Educators' Conference American Marketing Association. (August), 298-301.

Sattath, S. and Tversky, AX (1977). "Additive Similarity Trees," Psychometrika, 42, (3), 319-345.

Shimp, T. and Dyer, D. (1978). The Effects of Comparative Advertising Mediated by Market Position of Sponsoring Brand," Journal of Advertising, 3, 13-19

Shimp, T. and Gresham, L. (1983). "An Information Processing Perspective on Recent Advertising Literature, Current Issues and Research in Advertising, 39-75.

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

Tversky, A. (1977). Features of Similarity, Psychological Review, 84 (4), 327-352.

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

Wright, P. L. (1973). The Cognitive Processes Mediating Acceptance of Advertising," Journal of Marketing Research, 10, 53-67.