Printer Friendly

Examining effects of advertising campaign publicity in a field study.

Previous experimental research found that the pre-exposure of publicity about advertisements has two distinct but related effects in advertised brand recall: (a) a facilitative effect on publicized brands and (b) an inhibitive effect on nonpublicized brands. We speculate that publicity effects exist beyond the controlled experiments. In this article, we used a field study to investigate the effects of publicity messages related to the commercials aired during three Super Bowl games. We found that publicity had a positive impact on the memory of subsequent advertisements for both recall and recognition, but publicity effects were more evident in recall than in recognition.


IN AN ARTICLE ENTITLED "Commercials Become News and the Airtime is Free," Philip Dusenberry, a former chairman of the BBDO advertising agency, told The New York Times (Rothenberg, 1990):
  The advertising in news takes on more value because it's being
  mentioned in a non-advertising context.... Anytime you release a new
  advertising campaign, you would be wise to bring in your PR people and
  ask: Is there anything in this that can stretch it beyond our media

Major newspapers report on the day-to-day business of advertising. More than 3,000 news items about advertising campaigns were found in four major metropolitan daily newspapers over a three-year span, and The New York Times publishes about 700 advertising stories per year--almost two items per day on average (Pasadeos, Phelps, and Lamme, 2000). The mere launch of a new campaign is often deemed newsworthy, and the topic of product advertising was frequently covered in the news media.

Publicity about an advertising campaign can be an effective marketing communication tool; that is, the publicity stimulates excitement, builds expectations, and heightens awareness for the advertisement (Marketing, 1995). Ries and Ries (1999, p. 42) describe the value of this kind of marketing activity as a situation where "publicity is the nail, advertising is the hammer," Harris (1998) labels it "value-added public relations," and Yates (1995) found that a common element among successful brands is that their advertising campaigns hit the headlines.

While there have been anecdotal cases showing the effects of publicity about advertising campaigns (Harris, 1998; Ries and Ries, 1999; Rothenberg, 1990; Yates, 1995), and experimental studies provide empirical evidence of publicity effects (Jin 2003; Jin, Suh, and Donavan, in press), we speculate the publicity effects exist beyond controlled experiments. There are many recommendations about how to break through crowded advertising clutter and make a brand's advertising more memorable. This study extends previous research to examine whether and, if so, how much long-term memory may be strengthened for brands that have benefited from publicity in a field study.

The strength of controlled experiments in laboratory settings is their ability to establish causality. The trade-off, however, is a lower degree of generalizability. Accordingly, researchers argue that field studies are critical in substantiating the findings of controlled experiments (Lutz, 1996; Wells, 1993). In this article, we used a naturalistic field study to investigate the effects of publicity messages related to the commercials aired during three Super Bowl games.


Exploring how pre-exposure to publicity about advertisements affects consumers' information processing and memory of subsequent advertisements using a Super Bowl context, Jin (2003) tested a dual processing model: (a) a repetition process and (b) a motivation process. The repetition process posits that message repetition strengthens message encoding and provides greater opportunities to process the message. The claim that repetition improves memory performance is a well-established and common-sense phenomenon. More importantly, repetition by publicity and the advertisement is conceptually different from the simple repetition of the same advertisement; that is, a repetition variation. Encoding variability effects (Melton, 1970)--repeated messages in different contexts are better remembered--account for improved advertised brand memory with extra publicity because brand information is observed in two different contexts: the publicity and the advertisement itself.

The motivation process infers that pre-exposure of publicity about incoming advertising campaigns increases subsequent advertising memory through active involvement in the advertising event. Those who were exposed to the publicity had significantly greater interest in the upcoming Super Bowl advertisements than those who were not. Jin's (2003) findings show causal relationships in advertising memory through the motivational process: interest in the advertisements [right arrow] attention to the advertisements [right arrow] subsequent memory of the advertisements.

Jin's (2003) dual processing model focuses on the facilitative role of publicity. Pre-exposure to publicity about advertisements increases the recall of the subsequently advertised brands. In a follow-up study, Jin, Suh, and Donavan (in press) replicated and extended the earlier study. From the list-strength model (Ratcliff, Clark, and Shiffrin, 1990; Shiffrin, Ratcliff, and Clark, 1990)--strengthening memory of some items inhibits memory for other nonstrengthened items--they predicted there would be potential inhibitive effects that repeated exposure via extra publicity can suppress retrieval of other nonpublicized brands that would otherwise have been retrieved.

Findings indicated that people who were pre-exposed to publicity about some brands' advertisements recalled the publicized brands more than those who were not exposed to the publicity. It is a facilitative effect of publicity. More importantly, those who were pre-exposed to publicity recalled brands that were not in the publicity significantly less than those who were not exposed to publicity. It is an inhibitive effect. This result suggests that the recall of salient items with more "memory strength" occurs at the expense of less salient items. Whereas the inhibitive effect of publicity was observed in recall, interestingly, they did not find the inhibitive effect in recognition memory. In summary, the publicized brands take advantage of two beneficial forces in recall: the facilitation of publicized brands and the inhibition of non-publicized brands. Recognition memory would only have the facilitative effect. Therefore, the effects of publicity about advertisements would be stronger in free recall than in recognition.


In past decades, increasingly more researchers of advertising and consumer behavior (McQuarrie, 1998; Wells, 1993; Zhao, 2002) have expressed concern about prevailing practices of the field for not paying enough attention to external validity (i.e., generalizability). Zhao (1997, 2002) used a simple model to discuss the key elements of external validity for a study of advertising or communication effects. To achieve complete generalizability, we would hope that all three elements--messages, people to be affected, and exposure environments--are representative of the populations to which the researchers hope to generalize.

Such perfect validity would be extremely difficult to achieve in reality. It would be challenging to produce a sample of messages that are randomly selected and large enough to be representative of the population of all advertising or publicity messages. It might be even more difficult to sample viewing environments for a study of television advertisements. Short of randomly selecting each of the three elements, however, there are some ways that researchers can achieve better generalizability. For instance, whenever possible, use real advertisements or publicity messages rather than specifically developed stimuli considered suitable only for the experiments; use randomly selected people rather than just students; and place the people in a realistic viewing environment, ideally without their knowing that they are being studied.

The naturalistic design of the current study will meet each requirement. The substantial media coverage of Super Bowl advertisements and real Super Bowl commercials provided an opportunity to test the publicity effects in a completely natural field setting, which means that no forced exposures of publicity and advertisements were involved. Also, viewing environments were natural. Super Bowl advertising as a research instrument is nothing new. As a visible advertising event, academic researchers have used the Super Bowl as a research context or instrument, suggesting that the event is useful for testing general theories and extension of previous research findings (e.g., Pavelchak, Antil, and Munch, 1988; Yelkur, Tomkovick, and Traczyk, 2004; Youn, Sun, Wells, and Zhao, 2001; Zhao, 1997).

A trade-off for the higher generalizability in a single field study is a lower degree of internal validity. Most field studies are correlational. A causal inference is not always certain, although various precautions can be taken to guard against some methodological threats. The established causal effects of publicity from previous experimental studies (Jin, 2003; Jin, Suh, and Donavan, in press) allow the validating "effect" found in this field study to be interpreted as causal to some extent.

The benefits of adding a naturalistic study may not be limited by methodology. It may not only increase the validity of answers this study is producing, but also enhance the quality of questions this study is asking. Zhao (2002) observed that controlled experimental designs force most researchers to ask the qualitative whether questions even though the methodology is quantitative (e.g., whether publicity has an effect, whether the effect is positive or negative, etc.). Such questions lead to dichotomous yes-or-no answers. The quantitative how much question (e.g., how much does publicity affect memory?) is difficult to answer within a controlled experiment. To answer such questions meaningfully, the range of the independent variable under manipulation would need to match, at least roughly, the range of its counterpart in natural settings.

Ideally, there would also need to be some other naturally varying independent variables against which to make comparisons. For most studies of advertising, several naturally varying independent variables would be, if possible, too expensive to reproduce in a laboratory. Consequently, most experimental researchers do not even ask the how much question. In this study, we ask and answer the question by comparing the effect size of publicity with several naturally varying variables such as advertising frequency and advertising liking.

Another notable difference in this study is the unit of analysis: each brand advertised during the Super Bowl games. The dependent variables (brand recall and recognition) were measured via telephone interviews and then aggregated across respondents. Although such a design is rarely seen in advertising research, it has been used often in other disciplines. For example, political scientists and communication researchers who study agenda-setting effects (the media set the agenda of issues in public) typically aggregate survey responses to measure dependent variables, analyze the content of media coverage to measure independent variables, and use issue as the unit of analysis (McCombs and Shaw, 1972).

We have no information about which individual respondents in this study were and were not exposed to publicity messages. That would have been a problem had we, like most researchers doing controlled experiments, used person as the unit of analysis. In that case, we would have had no independent measure, hence no study. With brand as the unit of analysis, our independent, dependent, and control variables are all at the aggregate level and, naturally, our analysis will also be at the aggregate level. For example, we will test the effects of publicity by gauging the correlation between the amount of publicity for each brand and the recall and recognition of each brand. We do not need exposure information at the individual level in such analyses.



The design of the current study is an ex post facto design. Neither an experimental manipulation nor a control group was involved. The dependent variables--advertised brand recall and recognition--were measured through telephone interviews after three Super Bowl games (1998, 2002, and 2003) and then aggregated across respondents. The independent variable (publicity) was measured by analyzing published news stories on the Super Bowl advertisements. The unit of analysis in this study is each brand advertised during the three games. Thus, every advertised brand aired during the games had its own numeric values of publicity, recall, recognition, and other control variables that are discussed in a later section.

Telephone interviews

Telephone interviews were conducted on the Monday following three Super Bowl games. Graduate and undergraduate students enrolled in research classes at a major university conducted telephone interviews with local residents. Random digit dialing was used to include unlisted numbers. Interviewers asked to speak with the adult person in the household who had the next birthday. If a call yielded no answer, that number was redialed at least three times before being discarded. Interviewers asked each respondent whether he or she had watched the Super Bowl game and, if so, which parts (first quarter, second quarter, etc.).

Recall and recognition measures of advertised brands followed. The analysis we report in this article is based only on the responses from those who watched at least one part of the broadcast. The sample includes 903 adults and the average response rate was 62.3 percent. None of the respondents knew beforehand that we would be conducting the interviews, allowing the viewing situation to be completely natural.


Dependent measures. For the recall measure, those who watched any part of the game were asked to list all advertisements they remembered seeing. Free recall rates for each brand were then calculated as:

Recall rate of a brand = (R / W) x 100

where R is the number of respondents who recalled the brand, and W is the number of respondents who watched the segment(s) (first quarter, second quarter, etc.) of the program in which the brand was advertised. For the recognition measure, respondents were given a list of brand names and asked if they remembered seeing the commercials for those brands during the game. In this case, false alarms exist (i.e., respondents may claim recognition of an advertisement when they did not actually see it). To address this concern, interviewers emphasized that the brands listed may or may not have been advertised during the game. The purpose of the instruction was for respondents to use a higher criterion placement in a recognition test, which reduces false alarm rates (Hirshman, 1995).

With regard to a false alarm test, we included nine brands that were not advertised during the game on the list of brands for the recognition test. We calculated the recognition rates for each brand weighted by each respondent's correction rate from the false alarm test. For example, a respondent who got 75 percent right in the false alarm test was given 75 percent weight for the correctly recognized brands. The correlation between weighted and unweighted scores was .91. These results suggest that false alarms occurred rather randomly and did not significantly affect the recognition measure. The weighted recognition rates of each brand were calculated as:

Recognition rate of a brand = (G / W) x 100

where G is the weighted number of respondents who recognized the brand, and W is the number of respondents who watched the segments of the program in which the brand was advertised.

Independent measure. To assess news coverage of the advertisements, we conducted a content analysis. Included in the sample were three local newspapers, three national newspapers, three network television news programs, one cable news network, and seven national magazines. The three local newspapers were Chapel Hill News, News & Observer of Raleigh, and Durham Herald-Sun. The three national newspapers sampled were USA Today, The New York Times, and The Wall Street Journal. The three television networks were ABC, CBS, and NBC, and the one cable news network was CNN. Based on the possibility of news coverage of Super Bowl advertisements, the following magazines were included: Reader's Digest, TV Guide, Time, Newsweek, U.S. News & World Report, People, and Sports Illustrated.

The time period of the content analysis was seven days before each year's game. For national newspapers and magazines, the search was conducted using the Lexis/Nexis database and hard copies available at the university library. For local newspapers, we searched CD-ROMs produced by the local papers. For television news, we used the "Television News Archive" database (

Two research assistants collected all relevant news articles from the same data sources and analyzed news articles independently. In each story, they examined how many times each brand's name was mentioned. We examined intercoder reliability by the correlation of the two independent data sets. The correlation between the two coders was .98. Overall coverage of publicity depends on circulations or ratings; greater readership or higher ratings reach more readers and a larger audience. To make the coverage of the national and local media comparable, a vehicle's circulation or television rating in the national media was converted to the proportion of local households to the total U.S. households, which was .02 percent. Thus, all brand level media impressions are based on local households where the telephone surveys were conducted. We calculated media impressions as:

Media impressions = [SIGMA] (NB x circulation or TV rating x weight)

where NB refers to the number of times a brand's name is mentioned in each news story. When a news story was from national or local media, two weights were given: .02 for national and 1.0 for local, respectively.

Control variables. Survey respondents might have liked some advertisements more than others, increasing the chance of recalling liked advertisements. Favorable feelings at the time of exposure influence how the information is organized in memory. Cues--the highlighting of specific features--actively used for acquiring material play a role in what is stored, how it is stored, and how the material is later retrieved. Positive affect may be used by viewers in encoding, storing, and retrieving (Lingle and Ostrom, 1981; Lutz, 1985; Tulving and Thomson, 1973).

Thus, attitude toward an advertisement was assessed by measuring overall degree of liking. Respondents were given a list of the advertisements that had been advertised during the Super Bowl broadcast. Those who remembered seeing an advertisement were asked how much they liked it. Liking (variable, advertising liking) was rated on a scale of 1 to 10. The limited length of a telephone interview and the relatively large number of brands involved in this study did not allow us to use a multiple-item measure, thus reliability could not be calculated. However, reliability should not be a major issue because liking was aggregated across respondents. The large number of respondents should minimize the impact of any individual errors.

The number of advertisements a brand aired during a Super Bowl broadcast could be a significant factor (variable, advertising frequency). In addition to the fact that higher frequency should be associated with better memory, repeated exposure to advertisements is likely to yield higher ratings on attitude scales than less repeated advertisements, as suggested by the mere exposure theory (Zajonc, 1968). Furthermore, some brands that frequently appear in the Super Bowl, year after year, may be more easily remembered because of the association of the brands to the Super Bowl. We counted the number of times each brand appeared in the Super Bowl over the last five years (variable, previous appearance).

When we pooled the three years of data, there was a chance that differences between the years could confound results. Multiple years involved different lists of advertisements, different interviewers, different samples of local residents, and varying amounts of publicity. Two dummy variables were created, and the brands from 1998 served as a comparison group (variables, year 2002 and year 2003).

Product category (variable, product category) posed yet another concern requiring statistical control. Brands in certain product categories might be more easily remembered than others (Youn, Sun, Wells, and Zhao, 2001), and some of those brands may be related to brand familiarity and other control variables. Therefore, seven dummy variables were created to represent seven product categories that included services, shoes/clothes, health/beauty, household, food/beverage, public service announcements, and entertainment (e.g., movie trailers). The eighth category--auto-related brands--served as a baseline product category.


The final sample consisted of 227 advertised brands. Two separate multiple regressions were run for recall and recognition. The analysis procedure was the same for each dependent variable. The years variable was entered into the first control block, product category into the second, three advertising-related variables (advertising liking, advertising frequency, and previous appearance) into the third, and the independent variable (media impressions) into the last block. Descriptive statistics, including the intercorrelations of variables, are reported in Table 1. Media impressions and recall had unique patterns. Some brands had substantial news coverage, while about 60 percent of the brands garnered little or no publicity. About 57 percent of the brands were never recalled. By nature, deviation from normality is inevitable for media impressions and recall. Recognition had a reasonably bell-shaped distribution. We conducted various data transformations of the questionable variables and found that there were no significant differences regarding regression coefficients and significance tests. The results reported here were based on the sample without any data transformations.

Effects of control variables

For the years variable, neither recall nor recognition differed significantly from one year to the next (see Table 2). The first control block explains no variance in recall and 1.2 percent variance in recognition; neither is statistically significant. The effects of product category appear to be greater. The second control block explains more than 15 percent of the variance in recall and 22 percent in recognition. In the third block, three advertising-related variables significantly impact both recall and recognition. The more frequently brands appeared in previous years' Super Bowl advertising spots, the more likely people were to recall those brands. However, previous appearance did not play a significant role in recognition. Advertising frequency during the game was positively related to both dependent measures, while advertising liking was significant only for recognition. Those three variables together explained more than 40 percent of the variance in recall and 21 percent in recognition.

Effects of publicity

Media impressions had significantly positive effects on recall (B = 1.06; t = 7.31; p < .001) and recognition (B = 1.66; t = 3.33; p < .01). Advertised brands that had more publicity tended to generate higher brand memory. Standardized coefficients (beta) are often used for comparing the effects of different independent and control variables. The beta coefficients show that the media impressions variable (ssrecall = .49 and ssrecog = .29) is the best predictor, competing with advertising frequency (ssrecall = .15 and ssrecog = .15), advertising liking (ssrecall = .08 and ssrecog = .15), and previous appearance (ssrecall = .21 and ssrecog = .10). In addition, incremental [R.sup.2] statistics indicate that results for both recall (8.7 percent) and recognition (2.8 percent) had the predictive power of the media impressions variable; both are statistically significant. As a summary, we found that publicity had a positive impact on the memory of subsequent advertisements for both recall and recognition. Moreover, publicity effects were more significant in recall than in recognition.


Summary of the study

Because previous experimental studies established the causal effects of publicity about advertising campaigns on subsequent advertising memory (Jin 2003; Jin, Suh, and Donavan, in press), the main purpose of this study is to examine the effects of publicity in a field study using actual publicity messages about the commercials aired during Super Bowl games and the advertisements themselves. Neither was experimental manipulation of publicity and advertisements involved nor were processing variables measured. In this sense, the current study is empirical, exploring whether and how much long-term memory will be strengthened for brands that are benefited from publicity. Are there publicity effects? The answer is affirmative. The results of this study indicate substantial positive effects of publicity on both advertised brand recall and recognition.

Although our data are correlational, our findings were not completely random or a product of chance. The causality of publicity effects cannot be completely invalidated. First, the publicity that was content analyzed preceded the advertisements and memory tests. This time sequence is an important necessary condition to secure causality in a research design. However, we cannot exclude the possibility that more memorable advertisements might have received greater prepublicity. It is then difficult to disentangle the effects of publicity from the effects of more memorable advertisements. For the observed correlation between publicity and memory to be completely masked, none of the survey respondents could be exposed to any publicity about the advertisements. This is an unrealistic scenario. In addition, control variables such as advertising frequency and advertising liking in the analysis might have lessened this issue to some degree because the regression coefficient of the independent variable (media impressions, publicity) represents its unique contribution, blocking out shared common variances between the independent variable and other control variables.

Second, unknown confounding variables might have played some role, but total variances explained by the variables (recall: 67 percent; recognition: 47 percent) are high enough to reasonably argue that the variables in this study capture significant amounts of the variances. Third, the unit of analysis in this study would tend to lessen bias from individual differences. The dependent measures for each brand--aggregated across respondents--involve different genders, ages, levels of education, and the like, including those who read/watched the brand advertisements in various media and those who did not. Thus, individual differences should be partialled out when the unit of analysis is brands.

Unlike the previous experimental studies in which publicity was experimentally manipulated as a dichotomous variable, the media impressions (publicity) variable that was continuous in the current study allowed us to answer the how much question, estimating differential memory performance based upon the amount of publicity. Results indicate that publicity has positive linear effects on both recall and recognition. The more publicity, the higher recall and recognition are. Theoretically, message repetition should have a ceiling effect in memory. The relationship between media impressions and memory should have a curvilinear rather than linear pattern. A plausible explanation is the difference between availability and actual exposure to publicity. The media impressions variable in this study referred to how much publicity about each brand was available. Although a brand may have had heavy publicity, not all respondents were exposed to all available publicity. Some of the respondents might not have been exposed; some of them perhaps once or twice, etc. In a typical controlled experiment, when repetition is manipulated, there should be a ceiling effect. In a field study in which brands are the unit of analysis and no manipulations are executed, a linear pattern would be reasonable.

Monetary value of publicity

From a practical point of view, one might ask how much monetary value publicity has. According to the regression coefficients in Table 2, an additional 10,000 media impressions (local household based) may be associated with an increase in recall by 1.06 points and in recognition by 1.66 points. Additional advertising buying (advertising frequency) is related to a 1.40 point increase in recall and a 2.82 point increase in recognition. When we convert the local household based media impressions to the national based ones, media impressions of 500,000 may be equivalent to an 80 percent (60 percent) effect of one additional advertisement in recall (recognition). When a Super Bowl advertisement costs about $2 million, the value of 500,000 impressions would be $1.6 million for recall and $1.2 million for recognition.

Differential effects of publicity on recall versus recognition

The issue of recall and recognition in measuring advertising effectiveness is an old debate. Although there are various ways to measure recall and recognition in advertising, the fundamental difference between the two measures in this study is that for recall, survey respondents were to retrieve the brands advertised during the Super Bowl games without aid; for recognition, respondents had to identify the brands they had seen advertised when a list of brands were given.

Our findings indicate that publicity effects were much stronger in recall than in recognition. Theoretical accounts of differential publicity effects on recall versus recognition are worthwhile. Compared to recall, recognition requires a lower threshold, strength, or encoding, which might result in weaker impacts of publicity. A related account for the weaker effect in recognition can be a null inhibitive effect in recognition. As suggested by Jin, Suh, and Donavan (in press), the pre-exposure of publicity involves two distinct but related effects in memory: facilitation (benefit) for the publicized brands and inhibition (harm) for the nonpublicized brands. Thus, the recall measure might have reflected the inhibition that provided the publicized brands with an additional benefit to the facilitation. It is unlikely there would be an inhibitive effect in recognition although publicity has a facilitative role.

The contrast of recall versus recognition is related to differential retrieval processes. Free recall involves memory search with a context cue associated with a situation where consumers learn target information or an item that is already recalled. On the other hand, the memory probe cues in recognition consist of a context cue and/or test-item cues because test items are given in a recognition test (Gillund and Shiffrin, 1984). In a free recall test where test items are not available, strengthened (publicized) brands will harm the recall of other nonstrengthened (non-publicized) brands due to the sampling-with-replacement process (Rundus, 1973). The process suggests that when strengthened publicized brands are recalled, the brands are more likely to be sampled again. As a consequence, the probability and ability to recall remaining nonpublicized brands decrease. In the list-strength paradigm, strengthened items are recalled earlier than nonstrengthened items. However, such an inhibitive process is less likely to occur in recognition due to the presence of test items (Ratcliff, Clark, and Shiffrin, 1990; Shiffrin, Ratcliff, and Clark, 1990). Thus, publicity can be a more effective catalyst to increase recall-based memory of advertised brands.

One measure is more or less useful than the other depending upon the expected utility of information. When consumers engage in memory-based processing or decision making outside a store, a retrieval set--brands recalled at a particular point in time--is particularly important because consumers must recall the brand before it is considered for purchase (Bettman, 1979). We, in the current study, do not infer that recall of brands advertised is related to a retrieval set of brands (recall of brands). Because recalling a brand in a purchase situation is presumably stimulated by a need evoked in the consumer's mind (e.g., replenish home inventory, delight the kids, break up the monotony of regular habits, etc.), brands better connected to the need will be evoked more readily. All other factors being equal, however, it does seem reasonable that advertising-driven brand memory--if the advertisement communicates the consumer's underlying need and he or she remembers the brand--could make the brand itself salient in the consumer's memory.

In essence, the role of publicity about a brand advertisement is to make the brand or advertisement salient from crowded advertising clutter. Within advertising contexts, there are common ways to make a brand salient in memory such as advertising repetition (high frequency) and creative executions (e.g., copy, endorser, graphic, jingle, etc.). This study suggests that publicity about advertising campaigns can be another way to obtain the goal. When a brand is salient in consumers' memory due to extra publicity, an inhibition of other competing but nonpublicized brands in recall is a rather automatic process because the recall of salient brands with more "memory strength" occurs at the expense of less salient brands. Using a pre-announcement to kick off an advertising campaign can enhance the efficiency and effectiveness of the campaign without extra costs. Active efforts by marketers utilizing public relations counterparts to pre-announce campaigns can be beneficial.

Publicity: A marketing communication channel for intentional learning

As Belch and Belch (2004), Duncan and Caywood (1996), Lutz (1996), Schultz, Tannenbaum, and Lauterborn (1992), Shimp (2003), and Thorson and Moore (1996) suggest, the key point underlying the potential synergy of integrated marketing communications (IMC) is the recognition that different marketing communication tools have different strengths. One of the important strengths of publicity related to memory is that publicity creates an environment where intentional learning occurs because the process of encoding news usually requires active participation by the audience. Janiszewski, Noel, and Sawyer (2003) argue that a combination of intentional (elaborated) processing during one presentation and incidental (unelaborated) processing in another results in better memory. In this sense, publicity has a unique value that traditional advertising does not have.

Furthermore, information processing of publicity (news) in general requires a relatively high (elaborated) cognitive work, which helps to retain learned information. For example, Jin (2003) found that subjects who were exposed to publicity about Super Bowl advertisements but did not watch the Super Bowl game remembered the brands in the publicity as much as those who watched the game but were not exposed to the publicity. The findings indicate that brand information learned from publicity is stored in the long-term memory.

Brand-centered information processing

In highly competitive advertising environments, consumers often remember elements of an advertisement's execution, but cannot remember the brand or identify it correctly (Kent and Allen, 1994). To strengthen advertising effectiveness, Baker, Honea, and Russell (2004) suggest presenting the brand name at the beginning of an advertisement because knowing the brand earlier strengthens memory association between the brand and advertising content. Similarly, pre-exposure to publicity about advertising may produce this cognitive effect: prior knowledge of brand advertisements via publicity increases "brand-centered information processing" once consumers are exposed to subsequent advertisements. As a result, those exposed to publicity prior to advertising exposure have a more accurate advertised brand memory.

How to attract publicity

With findings highlighting the positive impact of publicity on consumers' memory, a critical question is how to attract publicity. Advertisers are looking to stretch their advertising dollars and publicity seems to be one creative way, but they have little control over publicity. Substantial amounts of advertising campaign publicity are based on public relations sources such as press releases and press conferences. News media use these sources as a primary way of gathering information about organizations' activities. Studies indicate that 40 to 70 percent of total news coverage can be attributed to public relations efforts (e.g., Lattimore et al., 1997). Accordingly, it would seem necessary--though not sufficient--for advertisers to make active efforts employing marketing public relations (MPR) to qualify for possible media coverage. Distributing print and/or video news release packages including teasers of advertisements and interviews with spokespersons would allow news media to produce their own segments and voice-overs (Harris, 1998).

Not every effort to get publicity will garner media coverage, so it is important for advertisers to identify what factors determine the likelihood of being publicized. News selection is a complex process reflecting a variety of goals, values, and developmental factors. Literature suggests that information must have some news value. In general, newsworthiness involves timeliness, proximity, prominence, conflict, oddity, and human interest (Shoemaker and Reese, 1996). Within an advertising context specifically, Pasadeos, Phelps, and Lamme (2000) found two major categories of information that impact the newsworthiness of an advertising campaign: "its importance" and "how interesting it is." Public relations efforts related to newsworthiness result in greater chances for campaigns to be publicized. The Super Bowl is arguably the most visible advertising event every year, increasing the news value of its advertising (e.g., stratospheric price tags, fun advertisements, etc.). Celebrity endorsers can also generate publicity for brands by capturing media attention (Agrawal and Kamakura, 1995; Erdogan, Baker, and Tagg, 2001). Other instances often covered by news media include new campaign launches, firing/ hiring of advertising agencies, size of advertising budgets, etc. (Pasadeos, Phelps, and Lamme, 2000).

Publicity specialists advise that forthcoming advertisements should have exploitable properties such as exceptionally creative work, unusual or innovative advertising approaches, evolving story lines, popular cast members, or musical interests (Marketing, 1995). To gain media attention, an advertising campaign should have news value in it, and the advertiser should be actively engaged in MPR efforts. Without PR, there will be little to no publicity.


There might be a method effect especially for the recognition measure. A recognition test is sensitive to how test items are presented (e.g., visual versus verbal, written versus auditory, etc.). The recognition test we did "over the phone" has limitations to capture not only a broader, deeper range of memory traces but emotional impacts of advertisements on memory. We cannot exclude the possibility that publicity effects on advertised brand recognition will be weaker than we observed in this study when the recognition is measured by showing actual advertisements. Counting only the amount of publicity about advertised brands, our study assumed that the content or quality of publicity was equal. However, different types of publicity might have differential effects. Our study did not capture them.

The Super Bowl is a highly visible event and a special case in the advertising environment. Therefore, we make no claims in this study about the generalizability of its findings. Contrary to the everyday advertising environment, people who are not even interested in the game will sometimes watch the Super Bowl because of the advertisements. Outside of the Super Bowl, most publicity about advertising campaigns does not provide consumers with information about when or where they can view the featured advertisements. However, the readers or viewers of news about Super Bowl advertisements know precisely when they will be seen. Our study does not show how publicity affects subsequent advertisements in a more typical low-involvement, inattentive exposure situation (Krugman, 1965). That is not to say publicity effects will be lessened in non-Super Bowl contexts. The "von Restorff effect" in memory literature suggests that unexpected information seems to capture one's attention, is processed more extensively, and is subsequently more likely to be recalled than information expected in a given context (Lynch and Srull, 1982). When consumers happen to see a commercial they previously read about, heard, or saw in the news, it may be distinctive to them, capturing their attention more than other spots in the same commercial pod. Considering the nature of the event, we may assume that attention to the Super Bowls' advertisements is parallel, thus reducing the von Restorff effect for brands that received publicity. In this case, it is possible that publicity effects might not significantly differ in non-Super Bowl environments.


Consumers' decision making on choice alternatives frequently relies on memory (Bettman, 1979; Lynch and Srull, 1982). Accordingly, efforts to make a brand salient in consumers' memory are an important marketing activity. There are many recommendations to make advertised brands more memorable in a cluttered processing environment. The present study suggests that publicity about an advertising campaign is an important marketing activity that simultaneously operates in favor of a marketer's brand and against competitors' brands. In a highly competitive advertising environment, publicity can become the nail to guide the hammer of advertising.

HYUN SEUNG JIN (Ph.D., University of North Carolina at Chapel Hill) is an assistant professor in the School of Journalism and Mass Communications at Kansas State University. His major research interests are in the areas of IMC, social marketing, consumer memory and attitudes, political advertising, and e-commerce. His work has been published in the Journal of Advertising, the Journal of Media Economics, and Journalism & Mass Communication Quarterly.

XINSHU ZHAO (Ph.D., University of Wisconsin at Madison) is a professor in the School of Journalism and Mass Communication at the University of North Carolina at Chapel Hill. His research has appeared in such journals as Communication Research, International Journal of Advertising, International Journal of Public Opinion Research, Journalism and Mass Communication Quarterly, and Public Opinion Quarterly.

SOONTAE AN (Ph.D., University of North Carolina at Chapel Hill) is an assistant professor in the School of Journalism and Mass Communications at Kansas State University. Her research interests are social marketing, political communication, media management, and advertising regulation and policy. She has published articles in Health Communication, the Journal of Health Communication, Journalism and Mass Communication Quarterly, the Journal of Media Economics, and Communication Law and Policy.


AGRAWAL, JAGDISH, and WAGNER A. KAMAKURA. "The Economic Worth of Celebrity Endorsers: An Event Study Analysis." Journal of Marketing 59, 3 (1995): 56-62.

BAKER, WILLIAM E., HEATHER HONEA, and CRISTEL A. RUSSELL. "Do Not Wait to Reveal the Brand Name: The Effect of Brand-Name Placement on Television Advertising Effectiveness." Journal of Advertising 33, 3 (2004): 77-85.

BELCH, GEORGE E., and MICHAEL A. BELCH. Advertising and Promotion: An Integrated Marketing Communications Perspective, 6th ed. New York: McGraw-Hill, 2004.

BETTMAN, JAMES R. "Memory Factors in Consumer Choice: A Review." Journal of Marketing 43, 2 (1979): 37-53.

DUNCAN, TOM, and CLARKE CAYWOOD. "The Concept, Process, and Evolution of Integrated Marketing Communication." In Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds. Mahwah, NJ: Lawrence Erlbaum, 1996.

ERDOGAN, B. ZAFER, MICHAEL J. BAKER, and STEPHEN TAGG. "Selecting Celebrity Endorsers: The Practitioner's Perspective." Journal of Advertising Research 41, 3 (2001): 39-48.

GILLUND, GARY, and RICHARD M. SHIFFRIN. "A Retrieval Model for Both Recognition and Recall." Psychological Review 91, 1 (1984): 1-67.

HARRIS, THOMAS L. Value-Added Public Relations: The Secret Weapon of Integrated Marketing, Lincolnwood, IL: NTC Business Books, 1998.

HIRSHMAN, ELLIOT. "Decision Processes in Recognition Memory: Criterion Shifts and the List-Strength Paradigm." Journal of Experimental Psychology: Learning, Memory, and Cognition 21, 2 (1995): 302-13.

JANISZEWSKI, CHRIS, HAYDEN NOEL, and ALAN G. SAWYER. "A Meta-Analysis of the Spacing Effect in Verbal Learning: Implications for Research on Advertising Repetition and Consumer Memory." Journal of Consumer Research 30, 1 (2003): 138-49.

JIN, HYUN SEUNG. "Compounding Consumer Interest: Effects of Advertising Campaign Publicity on the Ability to Recall Subsequent Advertisements." Journal of Advertising 32, 4 (2003): 29-41.

______, JAEBEOM SUH, and TODD DONAVAN. "Salience Effects of Publicity on Advertised Brand Recall and Recognition." Journal of Advertising (in press).

KENT, ROBERT J., and CHRIS T. ALLEN. "Competitive Interference Effects in Consumer Memory for Advertising: The Role of Brand Familiarity." Journal of Marketing 58, 3 (1994): 97-106.

KRUGMAN, HERBERT E. "The Impact of Television Advertising: Learning without Involvement." Public Opinion Quarterly 29, 3 (1965): 349-56.

LATTIMORE, DAN, OTIS BASKIN, SUZETTE T. HEIMAN, ELIZABETH L. TOTH, and JAMES K. VAN LEUVEN. Public Relations: The Profession and the Practice, 4th ed. New York: McGraw-Hill, 1997.

LINGLE, JOHN, and THOMAS M. OSTROM. "Principles of Memory and Cognition in Attitude Formation." In Cognitive Responses in Persuasion, Richard E. Petty, Thomas M. Ostrom, and Timothy C. Brock, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1981.

LUTZ, RICHARD J. "Affective and Cognitive Antecedents of Attitude toward the Ad: A Conceptual Framework." In Psychological Processes and Advertising Effects: Theory, Research, and Application, Linda F. Alwitt and Andrew A. Mitchell, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1985.

______. "Some General Observations about Research on Integrated Marketing Communications." In Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds. Mahwah, NJ: Lawrence Erlbaum, 1996.

LYNCH, JOHN S., JR., and THOMAS K. SRULL. "Memory and Attentional Factors in Consumer Choice: Concepts and Research Methods." Journal of Consumer Research 9, 1 (1982): 18-37.

MARKETING. "No PR, No Comment." January 26, 1995.

MCCOMBS, MAXWELL E., and DONALD L. SHAW. "The Agenda Setting Function of the Press." Public Opinion Quarterly 36, 2 (1972): 176-87.

MCQUARRIE, EDWARD F. "Have Laboratory Experiments Become Detached from Advertiser Goals? A Meta-Analysis." Journal of Advertising Research 38, 6 (1998): 15-25.

MELTON, ARTHUR W. "The Situation with Respect of Spacing of Repetition and Memory." Journal of Verbal Learning and Verbal Behavior 9, 5 (1970): 596-606.

PASADEOS, YORGO, JOE PHELPS, and OPDYCKE LAMME. "Newspaper Coverage of Advertising: The Unfolding Story." In Proceedings of 2000 Conference of the American Academy of Advertising, Mary Alice Shaver, ed. East Lansing, MI: Michigan State University, 2000.

PAVELCHAK, MARK A., JOHN H. ANTIL, and JAMES M. MUNCH. "The Super Bowl: An Investigation into the Relationship among Program Context, Emotional Experience, and Ad Recall." Journal of Consumer Research 15, 3 (1988): 360-67.

RATCLIFF, ROGER, STEVEN E. CLARK, and RICHARD M. SHIFFRIN. "List-Strength Effects I: Data and Discussion." Journal of Experimental Psychology: Learning, Memory, and Cognition 16, 2 (1990): 163-78.

RIES, AL, and LAURA RIES. "First Do Some Great Publicity: Building a Brand Takes More Than a Big Burst of Advertising Dollars." Advertising Age, February 8, 1999.

ROTHENBERG, RANDALL. "Commercials Become News and the Airtime is Free." The New York Times, January 8, 1990.

RUNDUS, DEWEY. "Negative Effects of Using List Items as Recall Cues." Journal of Verbal Learning and Verbal Behavior 12, 1 (1973): 43-50.

SCHULTZ, DON E., STANLEY I. TANNENBAUM, and ROBERT F. LAUTERBORN. Integrated Marketing Communications. Lincolnwood, IL: NTC Publishing Group, 1992.

SHIFFRIN, RICHARD M., ROGER RATCLIFF, and STEVEN E. CLARK. "List-Strength Effects II: Theoretical Mechanism." Journal of Experimental Psychology: Learning, Memory, and Cognition 16, 2 (1990): 179-95.

SHIMP, TERENCE A. Advertising, Promotion, and Supplemental Aspects of Integrated Marketing Communications. Mason, OH: Thomson, 2003.

SHOEMAKER, PAMELA J., and STEPHEN D. REESE. Mediating the Message. White Plains, NY: Longman, 1996.

THORSON, ESTHER, and JERI MOORE. "Introduction." In Integrated Communication: Synergy of Persuasive Voices, Esther Thorson and Jeri Moore, eds. Mahwah, NJ: Lawrence Erlbaum Associates, 1996.

TULVING, ENDEL, and DONALD M. THOMSON. "Encoding Specificity and Retrieval Processes in Episodic Memory." Psychological Review 80, 5 (1973): 352-73.

WELLS, WILLIAM D. "Discovery-Oriented Consumer Research." Journal of Consumer Research 19, 4 (1993): 489-504.

YATES, KAREN. "Making News: Press Coverage of Advertising Campaign." Campaign, May 26, 1995.

YELKUR, RAMA, CHUCK TOMKOVICK, and PATTY TRACZYK. "Super Bowl Advertising Effectiveness: Hollywood Finds the Games Golden." Journal of Advertising Research 44, 1 (2004): 143-59.

YOUN, SEOUNMI, TAO SUN, WILLIAM D. WELLS, and XINSHU ZHAO. "Commercial Liking and Memory: Moderating Effects of Product Categories." Journal of Advertising Research 41, 3 (2001): 7-13.

ZAJONC, ROBERT B. "Attitudinal Effects of Mere Exposure." Journal of Personality and Social Psychology Monographs 9, 2 (1968): 1-27.

ZHAO, XINSHU. "Clutter and Serial Order Redefined and Retested." Journal of Advertising Research 37, 5 (1997): 57-73.

______. "A Variable-Based Typology and a Review of Advertising-Related Persuasion Research in 1990s." In The Persuasion Handbook: Developments in Theory and Practice, James Price Dillard and Michael Pfau, eds. Thousand Oaks, CA: Sage, 2002.


Kansas State University


University of North Carolina at Chapel Hill


Kansas State University

The authors thank the co-editor Bob Woodard, Richard Lutz at the University of Florida, Surendra Singh at the University of Kansas, George Franke at the University of Alabama, and Bob Meeds at Kansas State University for their helpful comments on a previous version of the manuscript.
TABLE 1 Descriptive Statistics and the Correlations of Variables

Variables                           Mean   SD     Min.  Max.   A

Previous Super Bowl appearance (A)    .44   1.16  0      5     1.00
Advertising liking (B)               5.95    .93  3.57   8.56   .29
Advertising frequency (C)            1.48    .99  1      7      .50
Media impressions (D)                1.82   4.28  0     26.7    .63
Recognition (E)                     32.1   18.4   0     88.1    .42
Recall (F)                           3.04   9.01  0     65.9    .61

Variables                           B     C     D     E     F

Previous Super Bowl appearance (A)
Advertising liking (B)              1.00
Advertising frequency (C)            .29  1.00
Media impressions (D)                .36   .67  1.00
Recognition (E)                      .40   .42   .56  1.00
Recall (F)                           .37   .61   .78   .56  1.00

Note: All correlations are significant at p < .01 (two-tailed); the unit
of media impressions is 10,000; N = 227.

TABLE 2 News Coverage Effects on Advertised Brand Recall and Recognition

                        Recall                     Recognition
                   B (1)  Beta (2)  t          B (1)  Beta (2)  t

Constant           -5.71            -1.92      13.68            1.86
Control block
  1 (3)
  Year 2002        -1.19   -.06     -1.19       4.04    .11     1.64
  Year 2003        -1.48   -.08     -1.53       2.64    .07     1.10
Control block
  2 (4)
  Service           -.95   -.05      -.73       6.52    .16     2.02 (a)
  Shoes/clothes     -.19   -.01      -.10       6.26    .08     1.28
  Health            1.20    .02       .53      -5.09   -.05     -.90
  Household          .28    .01       .17        .12    .00      .03
  Food/beverage     1.37    .06       .99      11.35    .25     3.30 (b)
  Public service    -.83   -.02      -.38      20.72    .21     3.79 (c)
  Entertainment     -.50   -.02      -.36      16.17    .33     4.64 (c)
Control block 3
  Previous          1.68    .21      3.79 (c)   1.60    .10     1.47
  Advertising        .77    .08      1.60       3.00    .15     2.51 (c)
  Advertising       1.40    .15      2.50 (a)   2.82    .15     2.03 (a)
Independent         1.06    .49      7.31 (c)   1.66    .29     3.33 (b)
  variable Media
  impressions (5)
R (2) of control            .0                         1.2
  block 1 (%)
[DELTA]R (2) due          15.8 (c)                    22.3 (c)
  to control
  block 2 (%)
[DELTA]R (2) due          42.7 (c)                    21.5 (c)
  to control
  block 3 (%)
[DELTA]R (2) due           8.7 (c)                     2.8 (b)
  to media
  impressions (%)
Total R (2) (%)           67.1                        47.9

(1) B = unstandardized coefficients
(2) Beta = standardized coefficients
(3) The brands from 1998 served as a baseline for comparisons with the
brands from the other two years.
(4) Auto-related brands served as a baseline product category.
(5) One unit of media impressions variable is 10,000.
(a) p < .05
(b) p < .01
(c) p < .001
COPYRIGHT 2006 World Advertising Research Center Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Jin, Hyun Seung; Zhao, Xinshu; An, Soontae
Publication:Journal of Advertising Research
Geographic Code:1USA
Date:Jun 1, 2006
Previous Article:When synergy in marketing communication online enhances audience response: the effects of varying advertising and product publicity messages.
Next Article:How event sponsors are really identified: a (baseball) field analysis.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters