ARTHUR J. KOVER STEPHEN M. GOLDBERG AND WILLIAM L. JAMES CREATIVITY VS. EFFECTIVENESS? AN INTEGRATING CLASSIFICATION FOR ADVERTISING In many agencies, advertising creativity and effectiveness seem almost antipathetic. This research explores consumers’ emotional reactions to help define advertising perceived as both creative and effective. In doing this, the article also raises questions about some standard individual measures of advertising response, opting in addition for measures of emotional response. ARTHUR J. KOVER Professor Ford ham University W STEPHEN M.
G O L D B E R G Associate Professor Fordham University ithin the advertising industry, there seems to be a never-ending struggle between those who create the advertising (“creatives”) and those advertising managers who insist that it be “effective. ” Advertising agencies exist, sometimes precariously, in unstable environments (Comanor et a l , 1981; Hirschman, 1989). Therefore, stability in both organization and output generally are preferred whenever possible by agency management. Management’s goal is to have stable output which is predictable and “effective. Effectiveness, however, has many different dimensions by which it can be measured (Cook and Kover, forthcoming); the key element is some reliable measurement on We thank the comments and suggestions of Douglas Stayman, Joseph Priester, and particularly Bill Wells. Joseph Sirgy generously suggested a number of sources for persona! enhancement. This research was supported by a grant from the Research Committee, Graduate School of Business, Fordham University and a supplementary grant from Young & Rubicam, New York.
Dom Rossi, then President of Ayer, New York, gave permission for the Agency to prepare edited and labeled videotapes without cost. Our thanks to all. WILLIAM L, JAMES Professor Hofsira University which agency management and clients can concur. Usually, that measurement is an aspect of persuasion or (ideally) marketplace sales. Creatives, on the other hand, typically deride the criteria used by management and clients and allege they have Httle to do with the way advertising really works. Within a typical agency structure, the advertising product is formed by people in the “creative department. Creative people (copywriters and art directors) beheve that creativity is necessary for effectiveness, that the creative element pushes the message into viewers’ minds. In fact, some even feel that creativity is effectiveness (Kover, 1995). This belief seems general despite a few creative people who believe that creativity is merely a front for self-indulgent “artistic” attempts (Bensman and Gerver, 1958; Ordahl, 1993). Therefore, as might be expected, many agency managers mistrust creative advertising.
Creativity as defined by “the creatives” can be bothersome, costly, and timeconsuming. Creative advertising may win awards but may have little to do with advertising effectiveness (Gaylord, 1994). Creativity, after all, is unpredictable Jojrnal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 199S 29 CREATIVITY VS. EFFECTIVENESS? and can upset the appearance of stability and predictability that smoothes the lives of bureaucracies (Martin, 1994). The distinction between “effectiveness” and “creativity” is structural, reflecting different goals and needs of different departments (Ibarra, 1992).
The motivation behind this research was to see if this distinction, an artifact of advertising agency structure, would continue to exist when consumers were exposed to the advertising. Can, in fact, the different “language games” (Wittgenstein, 1953) of creativity or effectiveness be joined by a new language that transcends both? How do consumers react to different kinds of advertising executions? Do they respond as those in the advertising business believe they will? Or is there another dimension of response that transcends and might include creativity and effectiveness as defined by advertising people?
This paper describes the study we performed to investigate this issue. First, we provide a definition of terms we used and then outline the methodology employed. We next define the measures of interest followed by the results from a consumer survey. A detailed discussion of those results is followed by the creation of an integrative classification of advertising. Our conclusion is that the contention between “creatives” and agency management is specious—some “creative” advertising is effective, some “effective” advertising is creative, and some advertising (purportedly one or the other) can be neither. Definitions
This research examined advertising that had been classified as creative or effective by conventional criteria, accepted in the advertising business. “Creative” 30 advertising was that which had won a major award for creativity; in this case, recipients of One Show awards. The One Show is one of the most highly respected competitions for advertising creativity. Those who judge the One Show are successive panels of senior creative people at advertising agencies. While there is a certain circularity of judgments by people designated as “creative” to judge advertising creativity, this is the accepted practice in the industry.
One Show awards are eagerly sought by copywriters and are recognized for the care taken by those involved in the judging process. For “effective” advertising we used winners of EFFIE awards. Judges for EFFIEs are successive panels of marketing and research executives, mainly from the New York City area. They evaluate advertising campaigns based upon specific marketing goals and on market evidence that those goals had been met. In the case of EFFIEs, some objective evidence (sales or other market data) is used as a criterion to validate the success of winning campaigns. Both awards are respected in the field.
Both, however, quite properly limit themselves to their respective criteria. That is, it makes little difference to One Show judges if the advertising is effective; it should make equally little difference to EFFIE judges if the advertising is “creative. ” Submissions of advertising for each award are self-selecting. Advertising agencies submit advertising to a competition with the expectation that the advertising is creative (or effective). According to an initial analysis of submissions to either the One Show or EFFIE, there is little overlap. Perhaps a caveat is in order here.
Panels of expert judges may not be the ultimate measuring devices for either creativity or advertising effectiveness. But, at this stage, they are about all we have and their acceptance gives them some degree of face validity in an industry that seems to accept little else at face value. Method Overall. A general sample of consumers was asked to view a reel of television advertising while at home. Each reel contained some advertising that had won a One Show creative award, some that had won an EFFIE effectiveness award, and other advertising that was deemed neither particularly creative nor effective.
After viewing each commercial, participants were asked to respond in view of the commercial’s ability to elicit interest in the brand/product advertised, its likability, internal congruency, and its creativity. In addition, the participants responded to a battery of emotional descriptors to help determine the affect of each commercial. Specifics: Test Advertising. Each reel used by participants contained six commercials, three for one type of product, three for another.
Within each product type, one commercial was creative (winner of a One Show award in 1992), one was effective (winner of an Effie in 1992), and a third commercial was neither. To provide this “benchmark” advertising, four advertising executives in two agencies (two creative, two account management) supplied and selected advertising in those four general categories which they considered decent, ordinary, or “plain old advertising. ” With one exception, none of this advertising had been submitted for either the creative or the effectiveness Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995
CREATIVITY VS. E F F E C T I V E N E S S ? Figure 1 Advertising Selected: Categories and Specific Commercials Middle-of-the-road (“plain, old commercial”) Family restaurant B Insurance M Automotive F Sneaker E Category Restaurants Creative commercial Family restaurant Sf Effective commercial Family restaurant C” Bank credit card V* Automotive L’ Sneaker R# Financial services Bank credit card K* Automobiles Sneakers Automotive T# Sneaker N* NOTES: “Creative” commercials are winners of awards In the 1992 One Shovi/ competition, “Effective” commercials are winners of awards in the 1992 EFFIE competition.
Middle-of-the-road commmercials were selected by a panel of advertising executives to represent “decent, ordinary advertising,” Key to awards; ” Gold Award # Silver Award t Bronze Award award. Serious attempts were made to keep the advertising in precisely the same categories. In order to do so, a few commercials which had won other than the top (Gold) awards were selected. The selection of the “plain old advertising” was done to overcome in part what could have been a flaw in the design. Winners of the One Show awards are creative but are they effective?
Likewise, winners of EFFIEs are effective but are they creative? Unless a sponsor submitted advertising to both competitions, we will not know. The addition of the “plain old advertising” group allows comparison of both creative and effective advertising to similar advertising judged neither particularly creative nor particularly effective. Television advertising was selected in four categories: restaurants, financial services, automotive, and sneakers. This is advertising with wide appeal. Commercials in these categories are often submitted for awards.
The commercials selected are shown in Figure 1. Specifics: Test Stimuli (Reels of Advertising). The design was balanced so that each participant was exposed to six commercials. All participants were shown the three restaurant commercials and three commercials from one of the other three categories. Two sets of three reels were prepared so that the order of presentation could be reversed. Therefore, all participants evaluated the three restaurant commercials while subsamples of about one-third of the total sample evaluated commercials in the other categories.
Figure 2 shows the setup for each reel and the number of responses received for each. Sample: Participants. An independent research firm, The Data Group, used its national random telephone sample to select consumers who had VCR units in their homes. One hundred twenty-five heads of such households were recruited to watch “a reel of TV commercials” and to “fill out a questionnaire. ” As an incentive, participants were told they could keep the reel of test advertising. After two postcard follow-ups and up to four telephone callbacks, 69 usable questionnaires were returned (55 percent agreeing to articipate). Analysis of the areas and zip codes of participants and nonparticipants did not indicate any significant differences between the two groups. The characteristics of participants in this research are given in Table A-1 in the Appendix. The sample size places the researcher in a difficult position. Overall, it is large enough to permit some statistical analysis; however, individual cells are too small to allow for similar treatment. Consequently, we are constrained to report quantitative results where appropriate only for the complete participant data set.
We are similarly constrained to treat subsamples only in a qualitative way; we can examine trends and indications rather than look for statistically significant differences. Measures Participants were asked a set of questions after viewing each commercial. (See Appendix Figure A-2 for the questions used in this research. ) The measures of interest to us included liking the commercial, the effect the commercial had on purchase interest, and perceived creativity (was the commercial seen as old or new and was it perceived as exciting or dull).
In addition, measures of congruency were obtained; congruency is defined here as the extent to which an advertisement is perceived as “fitting” its category (Heckler and Childers, 1992; Kover and James, 1993). The idea also resembles that of “typicality” used in other research (Goodstein, 1993). Note that this definition differs from that of much previous research in which context congruency (the advertisement in its setting) was examined (Celuch and Slama, 1993; Isen, 1984; Lambert, 1980). Finally, we collected information on the affect 31
Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 CREATIVITY VS. EFFECTIVENESS? Figure 2 Test Reel Rotations and Consumer Exposures Reel Rotation Restaurants/Financial services S V B K C M Restaurants/Financial services M C K B V S Restaurants/Automotive S L B T C F Restaurants/Automotive F C T B L S Restaurants/Sneakers S R B N C E Restaurants/Sneakers E Consumer exposures fell into that cluster. Two of the commercials classified a priori as “plain old advertising” did not produce a definitive majority; for analytical purposes, we allowed them membership in two clusters.
Table 1 details the cluster membership of each commercial and the percentage of times a commercial was classified as belonging to a particular cluster. Note that Cluster I has one commercial from each of the original classifications; Cluster II is composed of two creative and two “plain old” commercials; Cluster III has one effective and two “plain old” advertisements; Cluster IV contains one creative, two effective, and one “plain old” commercial. Table 2 profiles each of the clusters in terms of the response measures. Table 2 shows that the clusters formed a continuum of response to advertising ranging from Cluster I to Cluster IV.
Clearly, the participants’ responses indicated that there must be something operating within this range of measures which generated groupings different from the original selection criteria for the research. Creativity. If creativity consists of two elements, excitement and perceived novelty, we can classify commercials in the four clusters as follows: • Creative: Viewers rate commercial as new and exciting. • Intermediate: Viewers rate commercial as fairly new and not dull. • Not creative: Viewers rate commercial as old and dull. Given this definition.
Cluster I is the one containing advertising that consumers saw as creative. Effectiveness. Effectiveness is defined in this research as the ability of a commercial to elicit interest in purchase or use. 1 2 3 4 5 9 11 12 17 8 12 Total: 69 usabie from 125 sent (55% return) 6 C NBR S each commercial generated using an emotional response battery developed by the N. W. Ayer advertising agency. The battery consists of 18 clusters of words, each cluster describing a set of emotions. A description of the battery, its validation and use by N. W. Ayer can be found in an earlier article (Kover and Abruzzo, 1993).
There is no lack of research in the field using measurements of liking and purchase intent. Other measures we used are more novel, particularly the measurements of congruency. Heckler and Childers (1992) posited two elements of congruency, expectancy, and relevancy. Goodstein, on the other hand, was concerned about typicality, an idea roughly equivalent to the idea of “fit. ” He noted that “[t]ypical advertisements evoked more processing, atypical advertisements evoked less processing” (Goldstein, 1993). Our three measures of congruency followed the leads of these authors.
For expectancy, we asked questions about how easy it was to identify the advertising for the type of product advertised; for relevancy, we asked how well the advertising message “hung together” for the viewers; for typicality, we asked directly how well the advertising fits its category. Analysis and Results Using the measures of purchase interest, liking, congruency, and creativity described above, we performed a cluster analysis on participants’ responses. A vector of respondent data served as input to the K-means procedure in the SPSS for Windows program (V 6. ). Based on the decline in variance accounted for, the four-cluster solution was determined to be most appropriate. (It is well known that determination of a clustering solution is subject to the interpretation of the researcher. We used the standard method of stopping at the point where additional variance accounted for is small relative to the immediately preceding solution. ) A commercial was assigned as belonging to a particular cluster if the majority of the participants’ response vectors 32 Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995
CREATIVITY VS. E F F E C T I V E N E S S ? Table 1 Commercial Clusters Cluster 1 Commercial CreaffVe Restaurants S Financial K Automotive T Sneakers N 18 25 66 30 (%) Cluster II Cluster 11 1 Cluster IV 25 40 13 0 3 10 44 35 3 25 28 35 Effective Restaurants C Financiai V Automotive L Sneakers R Plain Old Restaurants B Financial M Automotive F Sneakers E 35 13 50 3 40 38 14 23 25 58 24 15 7 20 3 60 28 25 40 0 7 50 5 10 21 15 30 59 5 10 17 40 congruency is a mark of creativity. It jounces the viewer into attention and presents new ideas and insights. Creatives perceive that the typical consumer sits in front of a television with dazed and dulled senses. The only way to break through is literally to shock or seduce the viewer into some attention (Kover, 1995). This research would dispute those creatives. Congruency. If anything, congruency in this research is related to effectiveness more closely than creativity. Congruency is how well the advertising “fits” the category it represents. For instance, does an automobile commercial fit what people expect when they first see such a commercial?
Conventionally, a congruent automobile commercial will show a “beauty shot” of the car, and probably another of the car driving down an endless (and otherwise empty) road. The conventional copywriters’ view of congruency (or “fit”) is: ‘ Interview with Glen Slater, copywriter, J. Walter Thompson Co. , 1993. Again, the commercials in Cluster I evoked the greatest interest. Clusters II and III evoked moderate interest. This finding parallels the feeling of advertising copywriters that creative advertising also is effective advertising (Kover, 1995).
While commercials in Cluster IV elicited the least purchase interest, they were not the least creative. Viewers saw these commercials as fairly new and not dull—at the least not uncreative. In addition, the viewers thought that commercials in Cluster IV stretched the boundaries of congruency: they did not hang together, could fit another category fairly well, and were hard to identify as commercials for their product or service. To some advertising creatives, non- Table 2 Response Profiles of Clusters Cluster Measurement Effectiveness
Question Purchase interest Like commercial Story hangs together (relevancy) Fits this or another kind of advertising (expectancy] Identifiability or fit (with sound or visual removed] 1 Very interested Like a lot Very well II III IV Not interested Dislike Not well Fits another fairly well Hard Somewhat Like little Fairly well Somewhat Neutral Fairly well Fits this fairly well Easy Liking Congruence Fits this Fits this fairly well very well Fairly easy Fairly hard Creativity New/old: Novelty Dull/exciting New New Not dull Old Fairly new Not Dull
Exciting Dull Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 33 CREATIVITY VS. EFFECTIVENESS? “If we can do something that is just different enough to grab that viewer, something that will touch . . . that person, then maybe we can slip in our name, our message. “^ That is, an incongruent advertisement can get by the guard that viewers have against the advertising message. “Stretching the envelope” of congruency too far, however, may produce advertising that is neither perceived as creative nor effective at provoking purchase interest.
An often-quoted example is introductory advertising for Infiniti automobiles, so poetic but diffuse that viewers did not know that an automobile was being advertised. Advertising in the most effective cluster (I) is congruent while advertising in Cluster IV (least effective) is not congruent. If congruency is expectedness, then expected advertising in this research is the most effective; advertising that is not expected is least effective. This, of course, goes against copywriters’ expectations. Perhaps we should allow for the possibility that this result is an artifact of the research design itself.
Participants were asked to view each commercial only once before responding. If commercials were unexpected (not congruent), then Krugman’s threeexposure hypothesis (Krugman, 1965) may come into play; time and energy spent viewing the first exposure may be just to figure out what it says. Consequently, we split the sample in Cluster IV using the criteria of previous exposure versus nonexposure to that commercial. The numbers are small, but even so there was virtually no difference in response between those claiming first-time versus past exposure to each ^ Interview with a senior copywriter, DMB&B, 1993. ommercial. “‘ This is true for the responses on which the commercials were clustered (purchase interest, novelty, and the others) and for emotional responses to the advertising. Perhaps Krugman’s hypothesis, accepted as gospel for so many years, needs reexamination. Recent work on the effectiveness of single exposures (Gibson, 1994) suggests that single exposures may have significant effect in the marketplace, by inference, perhaps as much as for multiple exposures. Approaching a New Classification In general, these responses by viewers go against the conventional wisdom about advertising.
Advertising that was more-orless expected was liked and was also seen as persuasive and creative by viewers. Commercials that jolt the viewer by their unexpectedness (using the congruency measures) were neither persuasive nor perceived as particularly creative. Creativity and effectiveness appear to join in consumers’ minds, rather than being separate, as the industry folklore would have it. Conventional classifications of creativity and effectiveness accepted in the industry seemed not to hold in this research. The question, of course, is why? Copywriters may hold the answer.
Recent work among copywriters (Kover, 1995) uncovered an “implicit theory” centering on emotional response to the advertising. Copywriters believe that the “connection” with advertising is always emotional (Isen, 1984); the issues of rational versus emotional responses apparently do not arise. Any emotion elicited toward the advertising may then be transferred to the product or service advertised. If those copywriters’ views are correct, then better understanding the effectiveness of advertising (and therefore a new classification of advertising) is possible.
Alternate Analyses Clustering of individuals’ responses in relatively small groups can be affected by minor differences in responses. We make no claim that the cluster solution we derived captures the total reality of advertising responses. It is, however, a stable solution and a sensible, if partial, representation of the world of consumer responses to advertising. “* The other question about the clustering is, if the kinds of results achieved would differ should the clustering be performed among the sample of 12 commercials rather than the individual sample of 69?
To answer this, a Johnson Hierarchical Clustering of the 12 commercials was carried out on the average linkages between groups on the variables used in the previous clustering. The linking extracted was as likely to be among commercials from disparate groups (e. g. , creative and “plain old”) as between commercials from the original groups. This lends some credence to the groupings found in the original individualbased clustering. ^ Data are available for those interested. Send request to the first author. ” “Adequate” to be used as a represen* tation of the real {deCerteau, 1988), pp. 34^35. Viewers’ Emotional Responses to Advertising
In addition to the measures described previously, participants’ emotional responses to the advertising were obtained immediately after exposure to each commercial. Timing of the questioning about emotion was 34 Journal ot ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 CREATIVITY VS. EFFECTIVENESS? Commercials that jolt the viewer by their unexpectedness . . . were neither persuasive nor perceived as particularly creative. designed to elicit this ephemeral response while it was still fresh. The battery consists of a number of sets of words; these sets of words are factors representing a longer list of emotional response descriptions.
Viewers were instructed to check as many of these word clusters as described their feelings while viewing the advertising. ^ Table 3 Emotional Responses to Clusters of Advertising (Percentages based on participants) Clusters II Emotion Not bored/attentive Active/aroused Wishful/desirous Benefited/enlightened Affectionate/loving Competent/confident Virtuous/rove rent Serene/restful Angry/powerless Amazed/astonished Tense/conflictfut Ashamed/embarrassed Suspicious/skeptical 59 59 45 17 13 18 3 3 49 11 17 4 2 4 1 2 IV 45 40 24 15 14 5 15 0 9 4 1 24 21 17 9 5 1 13 1 2 11 3 9 0 2 2 9 5 3 51 38 3 6 1 Table 3 gives the percentages of participants selecting different emotional responses for a commercial. The emotions are presented in three groups. The group of seven items at the top of the table indicate positive affect from advertising, the middle three items are neutral, and the bottom five items represent negative affect. The positive and negative emotional clusters are ordered from greatest to least positive affect, least to greatest negative affect. ^ Note that the two top items pertain to arousal and attention. Previous research at N. W.
Ayer suggested that the first item (“Not bored/attentive”) may be a precursor for any other response; that attention is the first step, necessary but not sufficient, in a two-step emotional response. ^ The emotional battery was developed, tested, and used at N. W. Ayer, an advertising agency in New York City. Details are outlined in Kover and Abruzzo (1993) and in Yoon (1991). ^ The actual beta weights are proprietary to N. W. Ayer. This ranking Is based on separate multiple regressions based on responses to 42 commercials with purchase interest and liking as dependent variables. 3 1 5 15 9 2 Confused/uninformed Disgusted/irritated enough by itself to have major Commercials in Clusters I effect. Cluster IV elicited mainly through III elicited attention strong negative responses: conamong many viewers; only I and fused and disgusted. Those reII elicited some form of arousal as sponses are “not me” (Sullivan, well as some degree of affection. 1953); they are kept away from In addition. Cluster I elicited a the persona. These four commergamut of positive, human feelcials were the fastest paced, inings: wishfulness, being benevolved situations that were outfited, and competence.
In other words. Cluster I aroused a complex side normal experience, and (to one author) condescended to the of emotions, both informational viewer. For instance, sneaker and transformational (Percy, 1991; commercial E showed a model Rossiter and Percy, 1991). That cavorting with Santa Claus (who complex of emotions seems to rose from the waves like an overmark the response of an ideal huweight male Venus in a red union man being: competent and caring, suit) on a California beach. aroused and active but full of desire. These affectual phrases Earlier research at N.
W. Ayer closely parallel the core patterns determined that consumers are of positive characteristics of the somewhat unwilling to record developing infant as described by negative responses. In order to Stern (1985). control for the actual levels of response and examine relative Cluster 11 elicited only one levels among the four clusters. other emotional cluster (“affecTable 4 normalizes the cells of the tionate”); apparently this is not Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 35 CREATIVITY VS. EFFECTIVENESS?
Table 4 Emotional Responses to Clusters of Advertising (Percentages based on responses) Clusters Ill Emotion Not bored/attentive Active/aroused Wishful/desirous Benefited/enlightened Affectionate/loving Competent/confident Virtuous/reverent Serene/restful Angry/powerless Amazed/astonished Tense/conflictful Ashamed/em ba rrassed Suspicious/skeptical Confused/uninformed Disgusted/irritated ‘ Too few responses to include 36 45 54 48 37 54 50 46 • 59 14 10 12 31 45 19 21 25 24 22 28 17 20 8 22 * 17 21 IV 6 17 5 3 9 4 17 0 • 12 31 37 12 35 32 • 12 34 somewhat creative. But it is not affecting. • Ho-Hum (Cluster III).
May attract some attention initially but is perceived as old and dull (even though two of the three commercials were quite new when this research was done). • Not-Me (Cluster IV). This advertising is seen as a negation of whatever fits the person’s needs for him- or herself. It is a negation of personhood. If this classification scheme is correct, advertising that provides for personal enhancement will 30 19 18 0 10 31 14 11 50 38 67 69 1 0 previous table v^’ith percentages calculated on total responses to an item rather than on respondents. The result is to separate the four clusters even more strongly.
Cluster I again emerges as eliciting the widest range of complex positive response; Cluster IV, conversely, elicits the widest range of negative, “not-me” responses. Cluster II, using this approach, evoked conflicting emotions: both affectionate and conflictful while, again. Cluster III evoked low levels of response except for suspicion, “what are they trying to pull off on me? ” must be regarded as both qualitative and suggestive. With this caveat in mind, we can categorize the advertising in this research as: • Person Enhancing (Cluster I). This advertising evokes a wide range of emotions associated with an ideal human.
This advertising is seen as congruent, creative, and effective. • Attention-Getting but Not Touching My Life (Cluster II). Although this advertising succeeded in raising attention and some arousal, emotional response was limited and conflicting. This advertising, relying heavily on fast action, unreal situations and arcane humor is seen as a host of “somewhats”: somewhat persuasive, somewhat congruent. work. In some way, this kind of advertising aids, if only momentarily, in each person’s quest to create a fuller persona, an ego that is continuously evolving through the Self (Stevens, 1982).
That advertising will be creative whether it is clever or not. Insight and empathy makes effective advertising that people also see as creative. Humor, fast cuts, fancy camera angles—all the triumphs of technique so common in current advertising— may have no relation to impact and effectiveness. Indeed, the opposite might be true. Discussion The key to understanding advertising is how people respond to it. Traditional definitions of creativity and effectiveness reflect agency and client organizational structures as much as anything else. This research attempted to bring the consumer back into the definition of advertising.
The answer to the question, “What kind of advertising works? ” was shown by participants to be personal enhancement. Personal enhancement can be defined as that which reflects the ideal for himor herself that each person carries as a sort of internal blueprint. Just as people carry within Toward an Integrative Classification of Advertising Remember that these results 36 Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 CREATIVITY VS. EFFECTIVENESS? themselves their picture of their ideal bodies (Harre, 1994), we believe that they carry a template for themselves as humans, an ideal self-image (Johar and Sirgy, 1991).
To the extent that advertising enhances this, it will have impact; to the extent that it ignores or denigrates, it will have less impact/ The specific content is almost immaterial. The commercials that clustered here were quite different. The approach also may be less important than imagined. Factual appeals, fantasy and play, nostalgia (Holak and Havlena, 1995)—all can lead to impact if they are self enhancing and reinforce the ideal self. The conclusion is quite simple: think of the viewer first. Offer her or him something with which an enhancing emotional connection can be made.
The result will be impactful, and it is likely to be viewed as creative advertising as well. This finding, suggestive as it is, opens questions about some current approaches to understanding advertising effectiveness. For instance, one would expect that advertising that reinforces ideal self-image would be that which is processed peripherally (using the ELM concept) (Johar and Sirgy, 1991; Sirgy, 1982). To use the language of Rossiter and Percy, we would expect that advertising that reinforces that self-image would be transformational rather than informational (Percy, 1991; Rossiter and Percy, 1991).
But this was not the case. Of the three Cluster I commercials, only one (Automotive commercial T) could be classified as transformational; the remaining two (Automotive commercial L and ^ This finding may not apply to shock advertising (such as anti-drug campaigns) in which a negative blueprint of the ideal self can be presented. Restaurant commercial B) were informational, telling quite plainly the functional benefits offered. On the other hand, the three commercials in Cluster IV were all transformational, implying some change in one’s being as a result of using the product.
Some recent approaches argue, using different language, that attitude toward the advertisement is perhaps even more important in effectiveness than previously thought. Brown and Stayman (1992) indicate that “results support the [attitude toward the ad] model . . . as [having] a more important role for the indirect influence of ad attitudes on brand attitudes (via brand cognitions) than that found in previous model tests. ” Blackston (1995) holds that the “feel good” effect of the advertising is an important aspect in . . we propose that the affect flows from the advertising toward the person . . . that the advertisement elicits certain affect within the person. increasing brand value (especially for packaged goods). Blackston, however, equated this “feel good” variable with liking the advertising, quite different from the complexity of personal enhancement. In fact, the key to the difference may be the direction of the affect. In theories covering attitudes toward the advertisement (including liking the advertisement), affect flows toward the advertisement.
In approaches such as Rossiter and Percy’s, it is a question of congruity of affect and goal of the advertising. In this case, we propose that the affect flows from the advertising toward the person (more precisely, that the advertisement elicits certain affect within the person). To the extent that the affectual pattern resembles one’s ideal self-image (Johar and Sirgy, 1991) the advertising is likely to work. Hence, advertising which triggers whatever negative feelings people have about themselves is likely to be actively resisted.
To some extent this resembles the hypothesis contained in copywriters’ implicit theory of advertising communication (Kover, 1995) in which positive affect is elicited by the advertising; this positive affect is then transferred to the brand or product advertised. Further, it resembles the finding of Murry (1992) in which attitude toward the brand is mediated by affect raised by the advertising (and by the programming). This article, therefore, opens a somewhat different way to look at advertising affect. Central or peripheral processing, informational or transformational motivations, may have their effects.
This project suggests, however, that the key pathway is; (1) to reinforce the individual’s idea! self, and (2) in doing so, increase favorable attitude toward both advertising and the brand or product advertised. In a sense, enhancing the individual through the advertising message means that the product can also enhance the self. Whether that enhancement comes from feeling different (transformational) or in mastery over some part of the rational world (informational) may be less important than the fact that the person is enhanced by the advertising message. References Bensman, Joseph, and Israel Gerver. “Art and the Mass Soci37 Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 CKHATIVITY VS. E F F E C T I V E N E S S ? ety. ” Social Problems 6, 2 (1958): 4-10. Blackston, Max. “Copy Testing and Brand Equity: What’s the Connection? ” Journal of Advertising Research 35, 1 (1995): RC2RC7. gory-based Applications and Extensions in Advertising: Motivating More Extensive Ad Processing. ” Journal of Consumer Research 2, 1 (1993): 87-99. Haley, Russell I, and A. L. Baldinger. “The ARF Copy Research Validity Project. Journal of Advertising Research 31, 2 (1991): 11-32. Kover, Arthur J. “Copywriters’ Implicit Theories of Communication: An Exploration. ” Journal of Consumer Research 21, 4 (1995): 30-45. , and Joseph Abruzzo. “The Rossiter-Percy Grid and Emotional Response to Advertising: An Initial Evaluation. ” Journal of Advertising Research 33, 6 (1993): 21-27 , and William L. James. “When Do Advertising ‘Power Words’ Work? An Examination of Congruence and Satiation. ” Journal of Advertising Research 33, 4 (1993): 32-38. Krugman, Herbert E. The Impact of Television Advertising: Learning Without Involvement. ” Public Opinion Quarterly 9, 2 (1965): 21-38. Lambert, David R. “Transactional Analysis as a Congruity Paradigm for Advertising Recall. ” Journal of Advertising 9, 2 (1980): 37-41, 44-45. Martin, Barry. Personal communication, 1994. Murry, Jr. , John P. ; John L. Lastovicka; and Surendra N. Singh. “Feeling and Liking Responses to Television Programs: An Examination of Two Explanations for Media-Context Effects. ” Journal of Consumer Research 18, 4 (1992): 441-51. Ordahl, Stafford. Personal communication, 1993.
Percy, Larry. “Understanding the Mediating Effect of Motivation and Emotion in Advertising Measurement. ” In Transcript Proceedings of the Eighth Annual Advertising Research Foundation Copy Research Workshop. New York, NY: Advertising Research FoundaHon, 1991. Brown, Steven P. , and Douglas M. Stayman. “Antecedents and Consequences of Attitude toward the Ad: A Meta-analysis. ” Journal of Consumer Research 19, 1 Harre, Rom. Physical Being: A Theory for a Corporeal Psychology. (1992): 34-51. Oxford: 1994. Celuch, Kevin G. , and Mark Slama. “Program Content and Heckler, Susan E. and Terry L. Advertising Effectiveness: A Test Childers. “The Role of Expectanof the Congruity Hypothesis for cy and Relevancy in Memory for Cognitive and Affective Sources Verbal and Visual Information: Involvement. ” Psychology and What Is Incongruency? ” Journal Marketing 10, 4 (1993): 285-99. of Consumer Research 18, 4 (1992): 475-92. Comanor, Williams S,; Arthur J. Kover; and Robert S. Smiley. Hirschman, Elizabeth C. “Role”Advertising and Its ConseBased Models of Advertising quences. ” In Handbook of Organi- Creation and Production. ” Jourzational Design, P. G.
Nystrom nal of Advertising 18, 4 (1989): and W. Starbuck, eds. London: 42-53. Oxford University Press, 1981. Holak, Susan L, and William J. Cook, William A. , and Arthur J. Havlena. Feelings, Fantasies, and Kover. “Research and the MeanMemories: An Examination of the ing of Advertising Effectiveness: Emotional Components of Nostalgia, Mutual Misunderstandings. ” In Working Paper, 1995. Measuring Advertising Effectiveness, W. Wells, ed. Lawrence Ibarra, Herminia. “Homophily Erlbaum, forthcoming. and Differential Returns: Sex Differences in Network Structure deCerteau, Michel.
The Writing of and Access in an Advertising History. Tom Conley, trans.. Firm. ” Administrative Science New York, NY: 1988. Quarterly 37, 3 (1992): 422-i7. Gaylord, Murray. “How Account Management Interacts Isren, Alice M. “The Influence of with the Creative Function. ” PaPositive Affect on Decision-Makper presented at an Advertising ing and Cognitive OrganizaResearch Foundation Special tion. ” In Advances in Consumer Seminar, 1994. Research, T. C. Kinnear, ed. Provo, UT: Association for ConGibson, Lawrence D. “What Can sumer Research, 1984. One TV Exposure Do? In Transcript Proceedings of the Eleventh Annual Advertising Research Foun- Johar, J. S. , and M. Joseph Sirgy. “Value-Expressive Versus dation Copy Research Workshop. Utilitarian Advertising Appeals: New York, NY: Advertising ReWhy and When to Use Each Apsearch Foundation, 1994. peal. ” Journal of Advertising 20, 3 (1991): 23-33. Goodstein, Ronald C. “Cate36 Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 CREATIVITY VS. EFFECTIVENESS? Rossiter, John R. , and Geoff Eagleson. “Conclusions from the ARF’s Copy Research Validity Project. ” Journal of Advertising Research 34, 3 (1994): 19-32. and Larry Percy. “A Better Advertising Planning Grid. ” Journal of Advertising Research 31, Psychology. 1985. New York, NY: ARTHUR J. KOVER is professor of marketing and chairperson of the marketing area at the Graduate School of Business Administration at Fordham University He has heid research positions at N. W. Ayer, Cunningham & Walsh, J. Waller Thompson, and other advertising agencies Dr Kover is past president of the Market Research Council. STEPHEN M. GOLDBERG is associate professor of marketing at Fordham Business Schooi He received both his M B A. and Ph. D. from Wharton.
He has published m the Journal of Marketing Research, the Journal oi Marketing, and other marketing publications. His current interests are in the areas of marketing strategv. competitive pricing, and new-product development. WILLIAM L. JAMES is professor of marketing and international business at Hotstra University He received his Ph D from Purdue University’s Krannert Graduate School of Management. He has published extensiveiy in marketing and related business fields. Stevens, Anthony. Archetype: A Natural History of the Self. Lon- don: 1982. Sullivan, Harry Stack. The Interpersonal Theory of Psychiatry.
New York, NY: 1953. Wittgenstein, Ludwig. Philosophical Investigations. G. E. M. 5(1991): 11-21. Sirgy, M. Joseph. “Self-Concept in Consumer Behavior: A Critical Review. ” Journal of Consumer Research 9, 3 (1982): 287-300. Stern, Daniel L. The Interpersonal World of the Infayil; A View from Psychoanalysis and Developmental Anscombe, trans. , Oxford: 1953. Yoon, Carolyn. “Tears, Cheers, and Fears: The Role of Emotions in Advertising. ” In the Proceedings of the Marketing Science Institute Conference. Durham, NC: Marketing Science Institute, 1991. Appendix 1 Technica! Note
A natural question one could ask is that if the variables used in the cluster analysis were interrelated, were they in fact separate dimensions useful in generating the clusters of commercials. This is not a regression question, however, and a concern about multicollinearity does not generate the same type of reasons to reject the cluster results. In this case, truly colUnear variables would simply be redundant and have little effecl on the cluster solution. As could have been expected with these variables, there are some high levels of correlation found among the eight items used to cluster the commercials.
Ratings on liking the commercials are correlated highly with acquisition interest (0. 77) (Haley and Baldinger, 1991; Rossiter and Eagleson, 1994); how the commercial hangs together (0. 71), and excitement level (0. 74). All other correlations were well below 0. 4. Acquisition interest ratings are highly correlated with how well the commercial hangs together (0. 69) and excitement level (0. 68) but no higher than 0. 4 with the remaining items (except, obviously, ratings on liking). How well the commercial hangs together correlates highly with excitement level (0. 61).
That four of the eight measures show high levels of intercorrelation might be a cause for concern. Consequently, a stepwise discriminant analysis was run on these data with cluster membership used as the grouping variable. High f to enter and remove criteria (of 2. 5 and 2. 49, respectively) were used to determine if the items used to develop clusters were either not useful or were redundant. In addition, the discriminant analysis provides statistics to enable determination of how different the clusters really are. All eight variables entered the discriminant function and the lowest F to remove statistic after the final step was 2. ^ (for excitement level). Therefore, we can conclude that all eight contribute effectively in defining conmercial clusters; high levels of redundancy are nonexistent despite some high correlations. Box’s Ki statistic was used to test whether the within-groups’ covariance matrices were different; the null hypothesis was rejected at;? = 0. 0000. Interestingly, the “confusion” matrix results (predicted versus actual group membership) showed that 96 percent of the commercials were correctly classified into their clusters. Cluster I was the only one to result in less than 90 percent correct classification (89. %). We conclude that the commercial clusters are correctly separated by the measures used in this research. Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995 39 CREATIVITY VS. EFFECTIVENESS? Appendix 2 Figure A-1 Demographic Profiles of Participants (Totai = 69) Demographics Gender Male Female Education High school or less Trade school Appendix 3 Figure A-2 Consumer Questionnaire [Participants were instructed to watch a specific commercial before answering these questions] Now, thinking just of ttils commercial- What emotions did this commercial arouse in you?
Check as many or as few items as you think appropriate. Aroused, playful ( ) Embarrassed, guilty ( ) Benefited, enlightened ( ) Affectionate, innocent ( ) Attentive, not bored ( ) Virtuous, moral { ) Tense, anxious ( ) Ama2ed, astonished { ) Disgusted, irritated ( ) Confident, proud ( ) Suspicious, skeptical ( ) Angry, powerless, sad ( ) Wishful, hopeful, desirous ( ) Sorrowful, afraid ( ) Serene, soothed ( ) Confused, uninformed { ) How much did you tike this commercial?
Liked it very much ( ) Liked it a little ( ) Neither liked nor disliked it ( ) Disliked it a little ( ) Disliked it a lot ( ) Put an X on the scale below to tell us how much this commercial interested you in purchasing/ 44 56 23 9 34 Some college College graduate Post graduate 15 19 63 12 13 12 35 53 12 Relational status Married now Formerly married Single/live with other Single/live alone Age Under 35 35 to 49 50 and older Employment Full time Part time Unpaid, full time Not employed Retired 70 18 7 2 2 27 Occupation Professional, manager using [product or service], 8 12 10 10 Service Technical Office worker Blue collar (including farmer) Sales Homemaker Owner/proprietor None given Income Under $20,000 $20,000-29,999 $30,000-39,999 $40,000-49,999 $50,000 and above Residential Area^ City or large town Suburbs Rural Region of US New England Mid-Atlantic Southeast North Central South Central Mountain Northwest Southwest ^ Self-description. 9 7 3 4 12 28 15 16 29 29 53 18 9 12 13 27 16 6 10 7 Not interested Somewhat Very at all interested interested In your opinion, how well did the story of this commercial hang together?
Very well { ) Fairly well ( ) Not too well ( ) Not well at all ( ) Have you seen this commercial before? Yes ( ) No ( ) Have you seen any advertising for [company] before? Yes ( ) No ( } Does this commercial fit the product or service it advertises or would it fit another kind of product or service better? Fits this product/service extremely well ( ) Fits this product/service fairly well ( ) Would fit another product/service fairly well ( ) Would fit another producfservice extremely well { ) Imagine this commercial without the sound: no music, no voices.
Would It be hard or easy to identify this as a commercial for [product or service]? Very hard ( ) Fairly hard ( ) Fairly easy ( ) Easy { ) Now imagine that you had your eyes closed and you heard only the sound. Would it be hard or easy to identify this as a commercial for [product or service]? Very hard ( ) Fairly hard ( ) Fairly easy ( ) Easy ( ) How would you classify this commercial? This commercial is: Old ( ) New ( ) This commercial IS: Dull ( ) Not dull ( ) Exciting ( ) 40 Journal of ADVERTISING RESEARCH—NOVEMBER/DECEMBER 1995
