A Test of Prescriptive Advice From the Rossiter-Percy Advertising Planning Grid Using Radio Commercials

Robin Higie Coulter, University of Connecticut
Murphy A. Sewall, University of Connecticut
[ to cite ]:
Robin Higie Coulter and Murphy A. Sewall (1994) ,"A Test of Prescriptive Advice From the Rossiter-Percy Advertising Planning Grid Using Radio Commercials", in NA - Advances in Consumer Research Volume 21, eds. Chris T. Allen and Deborah Roedder John, Provo, UT : Association for Consumer Research, Pages: 276-281.

Advances in Consumer Research Volume 21, 1994      Pages 276-281

A TEST OF PRESCRIPTIVE ADVICE FROM THE ROSSITER-PERCY ADVERTISING PLANNING GRID USING RADIO COMMERCIALS

Robin Higie Coulter, University of Connecticut

Murphy A. Sewall, University of Connecticut

An archive of 570 radio commercial tests of low involvement products is used to evaluate prescriptive advice offered by the Rossiter-Percy Advertising Planning Grid. The Rossiter-Percy Grid is an extension of the established Foote Cone and Belding Planning Grid that managers have long used as a guide to advertising copy strategy. Analyses of the radio commercial tests indicate meaningful differences in the ability of execution formats and copy variables to affect brand recall and preference. These findings suggest that the theoretical underpinnings of the Rossiter-Percy model provide an important basis for understanding the effectiveness of advertising copy strategy decisions.

INTRODUCTION

When deciding the composition of an advertisement, advertising and marketing managers must address a variety of issues, including: the communications objectives, the type of product, the characteristics of the target audience and their purchase motivation. Over the years, many managers have relied upon the Foote Cone and Belding (FCB) Planning Grid to help guide advertising decisions and plans-(Vaughn 1980, 1986; Ratchford 1987; Ratchford and Vaughn 1989). Recently, Rossiter, Percy and Donovan (1991; Rossiter and Percy 1987) introduced the Rossiter-Percy Advertising Planning Grid, which builds upon the established FCB Grid. Briefly, the Rossiter-Percy Grid suggests the need to consider two communications objectives: brand awareness (including brand recall and brand recognition) and brand attitude before deciding on the copy variable and execution strategies to be used. The Rossiter-Percy Grid prescriptive copy variable and execution strategies for brand awareness are not product category dependent (i.e., the same advice applies regardless of the product category.) However, according to the Rossiter-Percy model, the execution strategy for the brand attitude communications objective is dependent on two variables: 1) consumer involvement; i.e., perceived risk in choosing the test brand on the next purchase and 2) consumer motivation (informational versus transformational).

The purpose of this article is to determine the extent to which some of the prescriptive advice offered by the Rossiter-Percy Grid is effective in establishing better brand recall and more favorable brand attitude for low involvement products advertised on the radio. We review the prescriptive advice offered by the Rossiter-Percy Grid for establishing better brand recall and the advice for establishing more favorable brand attitude for low involvement informational and transformational motivations. Then, we use an archive of 570 radio commercial tests to examine the extent to which following the advice yields favorable brand recall and brand attitude.

BACKGROUND

Brand Recall and Prescriptive Advice

Studies of television, radio, and magazine advertisements have provided evidence of the importance of copy variables and execution style in achieving favorable brand recall (Duncan and Nelson 1985; Haller 1972; Ogilvy and Raphaelson 1982; Sewall and Sarel 1986; Stewart and Furse 1986). Copy variables that have been hypothesized, or have been shown, to have a positive effect on brand recall include: short main copy line, repetition of the message, use of personal references, and celebrity presenters (Belch 1981; Kahle and Homer 1985; MacLachlan 1984; Rossiter, Percy and Donovan 1991). Other research has shown both positive and negative relationships between humor and brand recall and music and brand recall (Cantor and Venus 1980; Gelb and Zinkhan 1985; Stewart and Furse 1986; Walker and von Gonten 1989; Weinberger and Campbell 1991).

Rossiter and Percy suggest that the attention-getting abilities of some execution factors will facilitate achieving higher brand recall, and consequently they offer the following prescriptive advice for establishing brand recall:

! Associate the category need and the brand in the main copy line.

! Keep the main copy line short.

! Use repetition of the main line copy.

! Include a personal reference.

! Use a bizarre execution.

! Use a jingle (for broadcast ads).

Brand Attitude and Prescriptive Advice

Research has also studied the effects of execution format on brand attitude, brand preference and other measures of persuasion (Diamond 1968; Ogilvy and Raphaelson 1982; Rossiter, Percy and Donovan 1991; Sandage, Fryburger and Rotzoll 1988; Wells, Burnett and Moriarity 1989). As noted, the prescriptive advice offered by the Rossiter-Percy Grid for establishing favorable brand attitude is dependent on two variables: consumer involvement and consumer motivation. The majority of the advertisements in the archive of radio commercials were for low involvement products (as might be expected because of the fleeting nature of the message and the difficulty in providing detailed, specific information via this medium). Thus, our empirical study examines only low involvement products, and we selectively review the low involvement informational and transformational motivations of the Rossiter-Percy Grid. (See Rossiter, Percy and Donovan 1991 for a detailed explanation of high involvement informational and transformational motivations.)

Involvement and Motivations. The involvement dimension of the Rossiter-Percy Grid is defined on the basis of the typical target audience's perceived risk in choosing the test brand on the next purchase. Brand choices in a product category that are "sufficiently low in perceived risk to simply 'try the brand and see'" are referred to as low involvement decisions. Products typically purchased in this manner include aspirin and candy (Rossiter, Percy and Donovan 1991, p.15). In contrast, brand choices in product categories that are "risky enough to be worth processing advertising information at a more detailed level" are referred to as high involvement decisions. Products typically purchased in this manner include insurance and automobiles (Rossiter, Percy and Donovan 1991, p.15).

The motivation dimension of the Rossiter-Percy Grid is based on the work of Katz (1960), Fennell (1978), Wells (1981) and Ratchford and Vaughn (1989) and distinguishes between informational and transformational motives. Rossiter, Percy and Donovan (1991) suggest that informational purchase motives are those motives that can be satisfied by providing information about the product or brand. The authors suggest that these are negatively-reinforcing and include: problem removal, problem avoidance, incomplete satisfaction, mixed approach-avoidance and normal depletion. On the other hand, transformational purchase motives are positively reinforcing and include sensory gratification, intellectual stimulation and social approval.

TABLE 1

HYPOTHESIZED RELATIONSHIPS BETWEEN COPY VARIABLES AND EXECUTION FORMAT AND DEPENDENT VARIABLES

For products that are purchased as a result of low involvement/informational brand processes, the Rossiter-Percy Grid prescribes the use of a simple problem-solution format. Here, the negative purchase motivations are addressed by supplying pertinent reasons for brand usage (eg., demonstrating that the consumer "problem" is alleviated through use of the product). For products that are purchased as a result of low involvement decisions and transformational motivations, the Rossiter-Percy Grid prescribes a format that incorporates emotional authenticity; here, the slice-of-life format is recommended because it helps consumers to "put themselves emotionally into the role of using the advertised brand" (Rossiter and Percy 1987, p. 239). To summarize, a problem-solution format is expected to facilitate positive attitude for low involvement/informational products, whereas the slice-of-life format is expected to be more effective in creating favorable brand attitude for low involvement/transformational products.

THE DATA AND HYPOTHESES

The present study examines an archive of 570 radio commercials for which both brand recall and purchase intention change were measured. The commercials had never been aired, and included 28 low involvement product categories. These commercials were drawn from 1,664 radio commercials tested by an advertising research firm between July 1982 and December 1989.

This archive of radio commercials provides a good opportunity to investigate the viability of several of the prescriptive advertising tactics set forth by the Rossiter-Percy Grid. Because the data were not collected specifically for the purpose of evaluating the Grid, we used surrogate measures when testing the propositions. Table 1 presents the Rossiter, Percy and Donovan variables and our substitutes; the details of our operationalizations are provided in the Classification and Measurement section. Table 1 also reports the hypothesized relationships between the execution elements and formats and brand recall and brand attitude.

PROCEDURE

For each data collection over the seven year period, approximately 100 people were recruited at shopping malls in each of three metropolitan areas (typically, one eastern, one central and one western). The participants were asked to come to a specific location to answer a short survey about their entertainment interests and brand purchase intentions. While subjects were responding to the questions, a "radio" (actually a disguised recorder) was played in the room. Three commercial messages were included in the musical program. The order of the commercials was rotated for different groups of subjects. The procedure is similar to the "semiforced exposure" for television commercials described by Dunn and Ziff (1974). At the end of the session, subjects again reported their brand purchase intentions. The following day, subjects were contacted by telephone and asked if they recalled hearing the commercials played while they were completing the questionnaire. Measures of unaided brand recall, category aided brand recall, claimed recall and proven recall were obtained.

TABLE 2

PRODUCT CATEGORIES AND INFORMATIONAL AND TRANSFORMATIONAL MOTIVATIONS

CLASSIFICATION AND MEASUREMENT

Product Category Classification and Purchase Motivations

As noted, the Rossiter-Percy Grid distinguishes between low involvement and high involvement product categories. Using the Rossiter, Percy and Donovan (1991) definition of low involvement (trial experience is sufficient) and a listing of products previously identified as low involvement in similar research (Ratchford 1987), two judges independently evaluated the product categories, assigning 28 as low involvement. There was 100 percent agreement between the two judges. Next, using the guidelines based upon the informational and transformational motives, the two judges independently classified the motivations as informational (19 product categories, including 355 commercials) and transformational (9 product categories, including 215 commercials). Again, there was 100 percent agreement between the two judges. Table 2 lists the products included in the analyses and their respective motivations.

Dependent Variables

Brand Recall. The day-after telephone interview was used to establish subjects' level of brand recall. During the interview subjects were asked "Do you recall a commercial for (product category of the test product, for example a pain reliever or a soft drink)?" (which is based upon Burke Marketing Research's definition of category-aided brand recall). For a particular commercial, brand recall is expressed as the percentage of the individuals who correctly named the brand.

Brand Attitude. The data set we used did not include a direct measure of brand attitude. We used a change in brand purchase intention for the product category as a surrogate to assess persuasion and brand attitude. Our reasoning for this is based on the expectation of a positive correlation between purchase intention and brand attitude (Haley and Baldinger, 1991; Ogilvy and Raphaelson, 1982). Subjects completed a measure of purchase intention both prior to and after the exposure to the commercials. At both times, subjects responded to the prompt: "Name the brand that you plan to select when you next make a purchase in each of these five product categories" (the advertised product category and four other product categories which varied over the data collection). The change in purchase intention (post-intention minus pre-intention) for a commercial was used as the measure of brand attitude.

Independent Variables

Two professionals at the research company coded all of the commercials in the archives for copy variables and execution format (discussed subsequently). Any discrepancies in codings were resolved by subsequent discussion between the coders.

Copy Variables. As noted, the Rossiter-Percy Grid suggests that a short main copy line, repetition of the main copy line and the use of a bizarre execution will favorably affect brand recall. The archive did not provide the number of words in the main copy line. Hence, message repetition (i.e., the number of times the main message of the commercial was repeated) was used to compute two surrogate measures C the number of words per message repetition and the number of ideas per message repetition. The assumption herein is that fewer words and ideas per message repetition would have the same positive effect on brand recall as a short main copyline. As shown in Table 1, we used number of words and number of ideas per message repetition as surrogates for the number of words in the main copy line. Additionally, as measures of Rossiter, Percy and Donovan's "repetition of the main copy line", we used number of message repetitions and number of brand name mentions in the commercial.

Execution Format. The radio commercials in the archive were coded by the research professionals as: sing-and-sell playlets (32%), slice-of-life (31%), announcer only (26%), testimonial (8%), and problem-solution (3%) ads. Sing-and-sell playlets use jingles to help gain attention, and slice-of-life commercials employ a common situation related to the product category with "everyday people" in the ad. Announcer only ads are rational, information-intensive advertisements, with virtually no humor or gimmicks. The problem-solution format acknowledges a problem that the consumer might face, and then offers the advertised brand as the appropriate solution to the problem; testimonials employ a spokesperson to discuss positive personal experiences with the brand.

TABLE 3

ANCOVA RESULTS FOR BRAND RECALL

For all product categories, Rossiter, Percy and Donovan (1991) predicted that a bizarre execution format would be effective in producing favorable recall. For our purposes, we define "bizarre" as different from the "typical" or "expected" format. Hence, we classified testimonial and problem-solution format ads (11 percent of all ads; i.e., infrequently used) as atypical, and announcer only, slice-of-life and sing-and-sell ads (89 percent of all ads; i.e., frequently used) as typical.

For products that fall into the low involvement/informational category, Rossiter, Percy and Donovan (1991) posit that the problem solution format would be most effective in achieving favorable brand attitude. Hence, we classified problem-solution and testimonial execution formats as prescribed; other formats were classified as not as prescribed. The slice-of-life (representing emotional authenticity) format is predicted to be the most effective low involvement/transformational brand attitude strategy. As a consequence, we classified slice-of-life as prescribed, and other formats were classified not as prescribed.

RESULTS

Brand Recall

To assess whether or not the Rossiter-Percy Grid advice for low involvement product commercials is effective in generating brand recall for radio commercials, we included execution format (atypical/typical) as an independent variable and the number of words, ideas, message repetitions, brand mentions and words and ideas per message repetition as covariates in an ANCOVA. Recall that the commercial is the unit of analysis. Table 3 reports our findings.

The results indicate that the typicality of the execution format, the number of ideas, the number of words and ideas per message repetition, and the number of message repetitions significantly affect brand recall. Specifically, as predicted by the Rossiter-Percy Grid, atypical execution formats achieve a greater brand recall than typical execution formats; the respective adjusted means are 16.2 and 12.0. Also as hypothesized, the fewer the number of ideas per message repetition and the more the message was repeated, the better the brand recall. In contrast to the Rossiter-Percy Grid advice, the number of words per message repetition has a positive effect on brand recall. Although not hypothesized (but necessarily included in the analysis because of the ideas per message repetition variable), the number of ideas in the ad has a positive effect on brand recall. In total, 8.3 percent of the variance in brand recall is explained by the aforementioned variables.

Brand Attitude

To assess the Rossiter-Percy Grid advice regarding brand attitude, we included two factors, product category purchase motivation (informational/transformational) and execution format (as prescribed/not as prescribed) in the ANOVA. The results indicate that the purchase motivation by execution format interaction is significant (F1/566=24.9; p<.001) and the execution format main effect is significant (F1/566=13.1; p<.001). The Figure illustrates the interaction effect. As predicted, for low involvement/transformational products, the slice-of-life format outperforms other formats in achieving favorable brand attitude (slice=3.88, other=.31. For low involvement/informational products, however, the recommended problem solution format was less effective than other formats (problem+testimonial =1.53, other=2.39). The effect of execution format indicates that across all of the commercials, following the Rossiter-Percy Grid advice results in more favorable brand attitude (as prescribed=3.13, not as prescribed=1.75). In total, 8.6 percent of the variance in brand attitude is explained by the informational/transformational dimension and the execution format.

Much research has distinguished between recall and attitude change measures (Gibson 1983; Higie and Sewall 1991; Ross 1982; Stewart 1986, 1989). Thus, it is not surprising that the Rossiter, Percy and Donovan Grid did not include the copy variables expected to effect brand recall in their brand attitude prescriptive advice. To empirically determine the validity of this distinction, we included the covariates expected to affect brand recall as covariates in an ANCOVA with product category purchase motivation and execution format (as prescribed/not as prescribed) as independent variables. Consistent with others' findings, this analysis indicates a significant motivation by execution format (F1/549=30.50, p<.001) and a significant execution format main effect (F1/549=12.59, p<.001). Among the covariates, only the number of brand mentions significantly affected brand attitude (F1/549=10.35, p<.001).

DISCUSSION

Researchers have offered guidelines for developing advertisements using a variety of conceptualizations. The effectiveness of the prescriptive advice, however, has been infrequently tested. Our examination of the low involvement product radio commercial archive provides insights as to the extent to which following the Rossiter-Percy Grid advice produces improved brand recall and brand attitude. Overall, our empirical findings support Rossiter, Percy and Donovan's conceptual formulation. However, for radio commercials, there may be potential refinements to the prescriptive advice.

FIGURE 1

INFORMATIONAL/TRANSFORMATIONAL MOTIVES BY EXECUTION FORMAT INTERACTION ON BRAND ATTITUDE

Brand Recall

Brand recall was enhanced, as expected, by more message repetitions, fewer ideas per message repetition and an atypical execution format. We found that brand recall improved with the number of words per message repetition and the number of ideas in the commercial. In other words, brand recall was facilitated by providing more information. Contrary to expectations, the number of brand name mentions did not promote greater brand recall.

These results have important implications for copy writers. In part, the findings suggest the importance of including information in commercials for low involvement products. In addition, our results indicate that infrequently used execution formats are more effective than frequently used formats. Thus, as suggested by the Rossiter-Percy Grid, unique executions (eg., atypical formats, copy variables or story lines) may be more likely to attract attention and promote brand recall. Finally, the Rossiter-Percy Grid hypothesis that a simple and frequent communication of the main message in an advertisement facilitates brand recall is corroborated by our findings that more message repetitions and fewer ideas per message repetition are effective in establishing greater brand recall.

Brand Attitude

The archive of radio commercials enabled us to examine the extent to which following a recommended execution format for low involvement/informational and low involvement/ transformational products would increase brand attitude. As expected, we found that the slice-of-life execution format was a more effective strategy for low involvement/ transformational products. However, for low involvement/information products, the prescribed problem-solution (and testimonial) format was a less effective method for generating favorable brand attitude. One possible explanation for these contrary findings is that an atypical format (similar to the brand recall hypothesis) may result in more favorable brand attitude for low involvement/informational motives. It may be that the problem-solution format is less effective for low involvement products than for high involvement products because consumers pay less attention than is necessary to process the message and have an impact on brand attitude.

Summary

Both brand recall and brand attitude models explained approximately eight percent of the variance in the dependent measures. Thus, although the Rossiter, Percy and Donovan prescriptive advice contributes to understanding the communications objectives of a particular commercial, the creativity and originality of the advertising creative staffs may be of significantly greater value. Moreover, our results further support the need to identify and pursue different communications objectives. The empirical evidence indicates that the inclusion of variables expected to promote brand recall (with the exception of number of brand mentions) does not significantly affect brand attitude.

LIMITATIONS AND FUTURE DIRECTIONS

This investigation provides useful insights about the types of commercial strategies that may be effective for low involvement product radio commercials. The theoretical underpinnings of the Rossiter-Percy Grid appear to provide an important basis for understanding the effectiveness of execution format and copy variables. The analyses of this archive of radio commercial tests indicate meaningful differences among the ability of various formats and copy variables to affect brand recall and brand attitude. Nonetheless, other research might seek to propose alternative product classification schemes and related conceptually-based prescriptive advice to investigate commercial strategy effectiveness.

The currently available field data enabled us to assess some of the hypotheses set forth by Rossiter, Percy and Donovan (1991), an important step in determining the efficacy of their advice. Because the data did not have a one-to-one correspondence of the measures detailed by Rossiter, Percy and Donovan, we attempted to meet the spirit of their intentions with the most appropriate surrogate measures. Our preliminary examination suggests that it may be worthwhile to invest in an experimental research design that more rigorously tests the stated variables and operationalizations included in the Rossiter-Percy Grid.

Finally, radio commercials were used to examine the effects of various copy variables and execution formats on brand recall and brand attitude for low involvement products. An obvious extension of this study would be to test the Rossiter-Percy Grid prescriptive advice for high involvement products. Further, a comparison across media might provide additional insights in terms of the most effective execution and copy strategies for specific media and advertising vehicles.

The authors thank Radio Recall Research, Holmdel, NJ for supporting this research.

REFERENCES

Belch, George E. "An Examination of Comparative and Noncomparative Television Commercials: The Effects of Claim Variation and the Repetition on Cognitive Response and Message Acceptance." Journal of Marketing Research 18, 3 (1981): 333-349.

Cantor, J. and P. Venus. "The Effect of Humor on Recall of a Radio Advertisement." Journal of Broadcasting 24, 1 (1980): 13-22.

Diamond, Daniel S. "A Quantitative Approach to Magazine Advertisement Format Selection." Journal of Marketing Research 4, 3 (1968): 276-287.

Duncan, Calvin P. and James E. Nelson. "The Effects of Humor in a Radio Advertising Experiment." Journal of Advertising 14, 2 (1985): 33-40, 64.

Dunn, Theodore F. and Ruth Ziff. "PREF: A New Copy Testing System." Journal of Advertising Research 14, 5 (1974): 53-59.

Fennell, G. "Consumers' Perceptions of the Product-Use Situation." Journal of Marketing 42, 2 (1978): 38-47.

Gelb, Betsy D. and George M. Zinkhan. "The Effect of Repetition on Humor in a Radio Advertising Study." Journal of Advertising 14, 2 (1985): 13-20, 68.

Haley, Russell I. Haley and Allan L. Baldinger. "The ARF Copy Research Validity Project." Journal of Advertising Research 31, 2 (1991): 7-32.

Haller, T. B. "Predicting Recall of TV Commercials." Journal of Advertising Research 12, 5 (1972): 43-45.

Higie, Robin A. and Murphy A. Sewall. "Using Recall and Brand Preference to Evaluate Advertising Effectiveness." Journal of Advertising Research 31, 2 (1991), 56-63.

Kahle, Lynn R. and Pamela M. Homer. "Physical Attractiveness of the Celebrity Endorser: A Social Adaptation Perspective." Journal of Consumer Research 11, 4 1985: 954-961.

Katz, D. "The Functional Approach to Studying Attitudes." Public Opinion Quarterly 24, 2 (1960): 163-204.

MacLachlan, James. "Making a Message Memorable and Persuasive." Journal of Advertising Research 23, 6 (1983/1984): 51-59.

Ogilvy D. and J. Raphaelson. "Research on Advertising Techniques that Work - and Don't Work." Harvard Business Review 60, 4 (1982): 14-16.

Ratchford, Brian T. "New Insights About the FCB Grid." Journal of Advertising Research, 27, 4 (1987): 24-38.

Ratchford, Brian T. and Richard Vaughn. "On the Relationships Between Motives and Purchase Decisions: Some Empirical Approaches." In Advances in Consumer Research, Vol. 16, T.K. Srull, ed. Provo, UT: Association for Consumer Research, 1989.

Ross, Harold L., Jr. "Recall versus Persuasion." Journal of Advertising 22, 1 (1982), 13-16.

Rossiter, John R. and Larry Percy, Advertising and Promotion Management. New York: McGraw-Hill Book Company, 1987.

Rossiter, John R., Larry Percy and Robert J. Donovan. "A Better Advertising Planning Grid." Journal of Advertising Research 31, 5 (1991): 11-21.

Sandage, C.H., Vernon Fryburger and Kim Rotzoll. Advertising: Theory and Practice. New York: Longman (1988).

Sewall, Murphy A. and Dan Sarel. "Characteristics of Radio Commercials and Their Recall Effectiveness." Journal of Marketing 50, 1 (1986), 52-60.

Stewart, David W. "Measures, Methods and Models in Advertising Research." Journal of Advertising Research 29, 3 (1989) 54-60.

Stewart, David W., "The Moderating Role of Recall, Comprehension and Brand Differentiation on the Persuasiveness of Television Advertising." Journal of Advertising Research 26, 2 (1986), 43-47.

Stewart, David W. and David H. Furse. Effective Television Advertising. Lexington, MA: D.C. Heath and Company 1986.

Vaughn, Richard. "How Advertising Works: A Planning Model." Journal of Advertising 20, 5 (1980): 27-33.

Vaughn, Richard. "How Advertising Works: A Planning Model Revisited." Journal of Advertising, 26, 1 (1980): 27-33.

Walker, David and Michael F. von Gonten. "Explaining Related Recall Outcomes: New Answers from a Better Model." Journal of Advertising Research 29, 3 (1989), 11-21.

Wells, William, John Burnett and Sandra Moriarity. Advertising: Principles and Practice. Englewood Cliffs, NJ: Prentice Hall, 1989.

Weinberger, Marc G. and Leland Campbell. "The Use and Impact of Humor in Radio Advertising." Journal of Advertising Research 30, 6 (1991), 44-52.

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