Chapter 15 Research in Advertising PDF

CHAPTER 15 RESEARCH IN ADVERTISING C H A P T E R OU T L I N E Copy Testing Summary Media Research...

CHAPTER 15 RESEARCH IN ADVERTISING C H A P T E R OU T L I N E Copy Testing Summary Media Research Key Terms Campaign Assessment Research Using the Internet Qualitative Techniques in Advertising Questions and Problems for Further Research Investigation Advertising Research and the Internet References and Suggested Readings 392 Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 393 For many years, research was not widely used however, to provide specific types of informa- in advertising and decisions were made on an tion that meet the needs of this industry. intuitive basis. However, with increased com- This chapter discusses the more common petition, mass markets, and mounting costs, areas of advertising research and the types of more and more advertisers have come to studies they entail. In describing these rely on research as a basic management tool. research studies, we aim to convey the facts Much of the research in advertising is the reader must know to understand the applied research, which attempts to solve a methods and to use them intelligently. specific problem and is not concerned with A significant portion of the research in theorizing or generalizing to other situations. these areas involves market studies con- Advertising researchers want to answer ques- ducted by commercial research firms; these tions such as whether a certain product studies form the basis for much of the more should be packaged in blue or red or whether specific research that follows in either the Cosmopolitan is a better advertising buy than academic sector or the private sector. The Vogue. Advertising research does not involve importance of market research notwith- any special techniques; the methods discussed standing, this chapter does not have suffi- earlier—laboratory, survey, field research, cient space to address this topic. Readers focus groups, and content analysis—are in who want additional information about common use. They have been adapted, market research techniques should consult A CLOSER LOOK Statistics Used by Advertising Researchers A recent study (Yoo, Joo, Choi, & Reid, 2012) Overall, basic statistics were used in 56% of examined the use of statistical techniques in pub- the articles, intermediate in 23%, and lished articles in four major advertising journals advanced in 21%. from 1980 to 2010. Articles were grouped into Content analyses were most likely to use 5-year intervals, and the statistics used in each basic statistics. Surveys and experiments article were classified as basic (descriptive statis- were more likely to use intermediate and tics, analysis of variance, Chi Square, t-test, advanced techniques. correlation); intermediate (regression, analysis One of the authors’ conclusions has special of covariance, partial correlation); or advanced relevance for those planning an academic career (factor analysis, path analysis, canonical corre- in advertising research: “To get ahead and enjoy lation, and several others). career success, future scholar/educators must Some of the major findings: have a substantial foundation not only in basic Over the years, the percentage of articles and intermediate statistics, but also in advanced using statistical analysis ranged from about techniques. Such knowledge is needed to (1) read 52% to 83%. Overall, about 69% of the and comprehend the published research literature published articles used statistical analysis. and (2) to ensure success in publishing their The percentage of articles using basic statis- research in the major advertising journals.” tics decreased from 1980 to 2010, while Those planning on a career in private-sector those using intermediate and advanced research would also benefit from a solid founda- statistics increased. tion in intermediate and advanced techniques. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  394 Part Four Research Applications McQuarrie (2011) and McDaniel and Gates In TV, a rough cut of an entire commer- (2011). cial might be produced. The rough cut is a The three functional research areas in film or video version of the ad in which ama- advertising are copy research, media research, teur actors are used, locations are simplified, and campaign assessment research. Each is and the editing and narration lack the discussed in this chapter, and the syndicated smoothness of broadcast (final cut) commer- research available in each case is described cials. In this way, variations in the ad can be when appropriate. Next, because qualitative tested without incurring great expense. research is popular with advertisers, the chap- The final phase of copy testing, which ter includes a discussion of three specific occurs after the finished commercials have qualitative approaches used to investigate appeared, serves to determine whether the advertising’s impact. Finally, advertisers are campaign is having the desired effects. Any shifting a significant portion of their budgets negative or unintended effects can be cor- to the Internet. Measuring the traditional con- rected before serious damage is done to a cepts in advertising research on the Internet company’s sales or reputation. This type of presents difficult challenges for both practi- copy testing requires precisely defined goals. tioners and investigators. This chapter closes For example, some campaigns are designed with an examination of some of the techniques to draw customers away from competitors; that researchers use to study online advertising others are conducted to retain a company’s audiences, strategies, and consequences. present customers. Still others are intended to enhance the image of a firm and may not be concerned with consumers’ purchase prefer- COPY TESTING ences. As we discuss later, this type of copy Copy testing refers to research that helps testing blends in with campaign assessment develop effective advertisements and then research. determines which of several advertisements There are several different ways to cate- is the most effective. Copy testing is done gorize copy testing methods. Perhaps the for ads in all media—print, audio, video, most useful, summarized by Leckenby and and digital. It takes place at every stage of Wedding (1982), suggests that there are the advertising process. Before a campaign appropriate copy testing methods for each starts, copy pretesting indicates what to of the three dimensions of impact in the per- stress and what to avoid. After the content suasion process. Although the model sug- of the ad is established, tests are performed gests a linear process starting with the to determine the most effective way to struc- cognitive dimension (knowing) and continu- ture the ideas. For example, in studying the ing through the affective dimension (feeling) illustration copy of a proposed magazine to the conative dimension (doing), it is not spread, a researcher might show an illustra- necessary for the steps to occur in this tion of the product photographed from dif- order—see Table 15.1. In any event, the ferent angles to three or more groups of model serves as a convenient guide for subjects. The headline can be evaluated by discussing copy research methods. having subjects rate the typefaces used in sev- eral versions of the ad, and copy can be The Cognitive Dimension tested for readability and recall. In all cases, In the cognitive dimension, the key depen- the aim is to determine whether the variable dent variables are attention, awareness, tested significantly affects the liking or the exposure, recognition, comprehension, and recall of the ad. recall. Studies that measure attention to Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 395 recognize the product, the headline, or the Table 15.1 Typology of Copy Testing brand name. Effects Ad comprehension is an important factor Dimension of Typical Dependent in advertising research. One study found Impact Variables that all 60 commercials used in a test were misunderstood by viewers (Jacoby & Hofer, Cognitive Attention 1982). To guard against results such as these, Exposure advertising researchers typically test new ads with focus groups (see Chapter 5) to make Awareness sure their message is getting across as Recognition intended. The T-scope is also used to see how long it takes subjects to comprehend Comprehension the theme of an ad—an important consider- Recall ation for outdoor advertising, where drivers may have only a second or two of exposure. Engagement Awareness, exposure, and recall are Affective Attitude change determined by several related methods. The print media use one measurement technique Liking/disliking that taps these variables: Subjects look at a Involvement copy of a newspaper or magazine and report which advertisements they remember seeing Conative Intention to buy or reading. The results are used to tabulate a Purchase behavior “reader traffic score” for each ad. This method is open to criticism because some respondents confuse the advertisements or the publications in which they saw the ads, advertising can use various methods. One and some try to please the interviewer by strategy involves a consumer jury, where a reporting that they saw more than they actu- group of 50–100 consumers looks at test ally did (prestige bias). To control this prob- ads and indicates which ad, if any, was best lem, researchers often make use of aided recall at catching their attention. A physiological techniques. For instance, they might also measurement technique, known as an eye- show the respondent a list of advertisers, tracking study (see Chapter 13), is also used some whose advertisements actually appeared to determine which parts of an ad are in the publication and some whose did not. noticed. A camera records the movement of For obvious reasons, this type of recall study the eye as it scans printed and graphic mate- is not entirely suitable for testing radio and rial. Analyzing the path the eye follows television commercials; a more commonly allows researchers to determine which parts used method in such cases is the telephone of the ad attracted initial attention. or online survey. Two variations of this A tachistoscope (or T-scope) is one way approach are sometimes used. In aided recall, to measure recognition of an ad. The T-scope the interviewer mentions a general class of is actually a slide projector with adjustable products and asks whether the respondent levels of illumination and with projection remembers an ad for a specific brand. A typi- speeds that can be adjusted down to a tiny cal question might be “Have you seen or fraction of a second. Ads are tested to deter- heard any ads for soft drinks lately?” In the mine how long it takes a consumer to unaided recall technique, researchers ask a Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  396 Part Four Research Applications general question such as “Have you seen any readership information for more than 80 ads that interested you lately?” Obviously, it years. Respondents look through an issue is harder for the consumer to respond to the of a magazine and indicate their readership second type of question. Only truly memora- of individual ads. Readers are classified into ble ads score high on this form of measure- four categories: ment. Some researchers suggest that the most 1. Nonreader (did not recall seeing the sensitive way to measure recall is to ask con- sumers whether they remember any recent advertisement) advertising for each particular brand whose 2. Noted reader (remembered seeing advertising is of interest. the advertisement) Several private sector research firms pro- 3. Associated reader (not only saw the vide posttesting services. Gallup & Robinson advertisement but also read some offer the InTeleTest method, which measures part of it that clearly indicated the the percentage of respondents who remember brand name) seeing a commercial and the percentage of 4. Read most reader (read more than those who can remember specific points. half the written material in the Additionally, they provide a score that indi- advertisement) cates the degree of favorable attitude toward the product, based on positive statements The Starch organization reports the find- made by the subjects during the interview. ings of its recall studies in a novel manner. Gallup & Robinson also conduct pretests Advertisers are given a copy of the magazine and posttests of magazine advertisements. in which readership scores printed on yellow Their Magazine Impact Research Service stickers have been attached to each (MIRS) measures the recall of advertisements advertisement. that appear in general-interest magazines. The newest measure in advertising and Copies of a particular issue containing the media research is called engagement, a mul- advertisement under study are mailed to tidimensional measure that attempts to deter- approximately 150 readers. (In the case of mine how involved a consumer is with an a pretest, the MIRS binds the proposed advertisement or media content. The method advertisement into each magazine.) The day combines cognitive measures with some of after delivery of the magazines, respondents the affective and conative measures discussed are telephoned and asked which advertise- in the next sections. Engagement can be ments they noticed in the magazine and measured in several ways. Researchers may what details they can remember about ask about recall of the ad, whether the con- them. These results are reported to the sumer liked the ad, and whether the con- advertiser. sumer talked about the content of the ad One of the best-known professional with others. There is evidence to suggest research firms is Gfk MRI, which conducts that engagement measures are positively posttest recall research. The company’s Issue correlated with intent to purchase a product Specific Readership Study uses online inter- (Kilger & Romer, 2007). views to report how many people saw a par- Peacock, Purvis, and Hazlett (2011) ticular ad in a specific issue of a magazine. examined 16 advertising campaigns that Data are reported by age, gender, and house- used both radio and television ads. They hold income level. found that engagement was positively related Starch Advertising Research (also a part to brand recall but that TV ads generated a of Gfk MRI) has provided advertising slightly higher negative emotional reaction. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 397 A CLOSER LOOK Cautions about Advertising Research The perception that many people have who are commercials that have words that must be outside the advertising community is that all read, are a complete waste for as much as businesses and their advertising agencies use 75% of the audience. In other words, a poly- the latest and most effective forms of advertis- chronic viewer obtains no information from a ing—that print and electronic advertising is on visual-only commercial. This is demonstrated eas- the “cutting edge” of knowledge of advertising ily in tests of TV commercials in which respon- (persuasion). With all due respect to advertis- dents are asked to rate commercials by looking ing agencies and others involved in the devel- away from the TV screen. When visual-only com- opment of advertising, this is not true. An mercials are tested and the respondents are example will help with this misperception. asked to rate the commercial or explain what When it comes to television viewing, it is com- information they obtained from it, the answer mon knowledge that about 75% of all TV viewers to both questions is always, “I don’t know.” do not watch the TV screen all the time. Instead of Therefore, it is clear from the countless number constantly watching the TV screen, these people of visual-only commercials on American televi- listen to TV while they participate in other activi- sion that advertising experts do not incorporate ties such as reading, eating, playing with chil- “cutting-edge” research into the development dren, and so on. As mentioned in Chapter 14, of commercials. Instead, advertising agencies the people who simultaneously participate in two create commercials for other reasons—to win or more activities, such as TV viewing and read- awards, to be considered “artistic,” or to mimic ing, belong to a category known as polychronic other advertising agencies. There is no concern behavior. In current terms, these people are mul- about the success of the ad in relation to its ability titasking. The people who do not participate simul- to communicate a message to the audience. A taneously in two or more activities belong to a significant amount of advertising money in the category known as monochronic behavior. United States is wasted on messages that have People exhibiting monochronic behavior watch no purpose or meaning. This is primarily the TV and do nothing else. fault of advertising agency employees who Because the majority of TV viewers are poly- don’t know much, if anything, about advertising chronic, it is scientifically logical that commer- (communication/persuasion), and of clients who cials should include both visual and audio unfortunately don’t know enough to question the information. Audio information does not mean advertisements presented to them by their agen- music. It means spoken words. TV commercials cies. For more information about this topic, see that include only visual information, including Wimmer (2000). Mersey, Malthouse, and Calder (2010) The Affective Dimension found two types of engagement for online The affective dimension usually involves media: personal engagement and social- research into whether consumers’ attitudes interactive engagement. Both types of toward a particular product have changed engagement were related to online ad because of exposure to an ad or a campaign. readership. The techniques used to study the affective Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  398 Part Four Research Applications dimension include projective tests, theater pupil dilation that occurs while the person is testing, physiological measures, semantic dif- looking at an ad. Changes in pupil diameter ferential scales, and rating scales. Projective are recorded because findings from psycho- tests provide an alternative to the straight- physiology suggest that people tend to forward “Do you like this ad?” approach. respond to appealing stimuli with dilation Instead, respondents are asked to draw a (enlargement) of their pupils. Conversely, picture or complete a story that involves when unappealing, disagreeable stimuli are the ad or the product mentioned in the ad. shown, the pupil narrows. The second test Analysis of these responses provides addi- measures galvanic skin response, or GSR, tional insight and depth into consumers’ feel- where changes in the electrical conductance ings. We discuss projective tests later in this of the surface of the skin are measured. A chapter. change in GSR while the subject is looking Theater tests involve bringing an audience at an ad is taken to signify emotional involve- to a special facility where they are shown a ment or arousal. The third technique, brain TV program that is embedded with test com- wave analysis, monitors brain activity during mercials. Respondents are given electronic exposure to a television commercial in order response indicators (ERIs—similar to hand- to measure the level of interest and involve- held calculators) that allow them to instan- ment by a viewer (Percy & Rossiter, 1997). taneously rate each commercial they see. As advances in neuroscience continue, it is There are a variety of ERI devices, but all likely that studies of the brain will become have buttons, a dial, or a sliding device so more important in advertising research. This that respondents can rate the commercials area has been given its own name, neuromar- (or other content) on a scale, such as “very keting (Marci, 2008). In 2011 the Advertising negative” to “very positive.” The respondents Research Foundation issued a report that record their perceptions while watching the recommended standards for neuromarketing commercial or other content. The miniaturi- (Stipp, Weber, & Varan, 2011). zation of handheld rating devices allows tests Hazlett (1999) describes a method to be conducted in focus room facilities or in whereby facial electromyography (fEMG—a specially equipped vans parked outside shop- technique that measures the electrical activity ping malls. These tests have been criticized of facial muscles) was used to track consu- because they require respondents to make mers’ responses to TV commercials. The too many responses and analyze content results suggested that fEMG was a better that may be too minute to be put into practi- measure of emotion than relying on viewers’ cal use, and they do not allow respondents self-reports. The research company Gallup & to change their answers because the answers Robinson offers a service called Continuous are recorded instantaneously in a computer. Emotional Response Analysis (CERA) that (Sometimes a researcher’s desire to use uses fEMG to track consumers’ emotional technology to impress clients overshadows arousal while watching an ad. The big the validity and reliability of a research advantage of fEMG is that it can differentiate approach. That is, the use of technology in negative arousal from positive arousal research can create a “leap backward” in because different facial muscles are involved data collection and analysis). in each response. Four physiological tests (see Chapter 9) Semantic differential scales and rating are used in this area of advertising research. scales (see Chapter 2) are used most often In the pupilometer test, a tiny camera focused to measure attitude change. For these mea- on the subject’s eye measures the amount of surements to be most useful, it is necessary Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 399 to (1) obtain a picture of consumers’ atti- evaluated three measures of affect: physio- tudes before exposure to the ad, (2) expose logical (GSR, heart rate, and fEMG), sym- consumers to the ad or ads under examina- bolic (a picture-sorting technique), and tion, and (3) remeasure their attitude after self-reports (verbal and ERI). They recom- exposure. To diminish the difficulties associ- mended that all measures be used in combi- ated with achieving all three goals in testing nation because less intense emotions are television ads, many researchers prefer a tapped by physiological measurements while forced-exposure method. In this technique, self-report measures capture conscious emo- respondents are invited to a theater for a spe- tional reactions. Symbolic measures provide cial screening of a TV program. Before view- a mental map of the brand. ing the program, they are asked to fill out questionnaires concerning their attitudes The Conative Dimension toward several different products, one of The conative dimension deals with actual which is of interest to the researchers. Next, consumer behavior; in many instances, it is the respondents watch the TV show, which the most important of all dependent vari- contains one or more commercials for the ables. The two main categories of behavior product under investigation in addition to usually measured are buying predisposition ads for other products. When the show is and actual purchasing behavior. In the first over, the respondents complete another ques- category, the usual design gathers precam- tionnaire concerning product attitudes. paign predisposition data and reinterviews Change in evaluation is the essential variable the subjects after the advertising has been in of interest. The same basic method can be place. Subjects are typically asked a question used in testing attitudes toward print ads such as “If you were going shopping tomor- except that the testing is done individually, row to buy breakfast cereal, which brand often at the respondent’s home. Typically, a would you buy?” This might be followed by consumer is interviewed about product atti- “Would you consider buying any other tudes, a copy of a magazine that includes the brands?” and “Are there any cereals you test ad (or ads) is left at the house, and the would definitely not buy?” (The last question respondent is asked to read or look through is included to determine whether the advertis- the publication before the next interview. A ing campaign had any negative effects.) Addi- short time later, the interviewer calls the tionally, some researchers (Haskins, 1976) respondent and asks whether the magazine suggest using a buying intention scale, has been read. If it has, product attitudes which instructs respondents to check the one are once again measured. position on the scale that best fits their inten- The importance of the affective dimen- tion. Such a scale might look like this: sion was emphasized by Walker and Dubitsky (1994), who noted that the degree I’ll definitely buy this cereal as of liking expressed by consumers toward soon as I can. a commercial was significantly related to I’ll probably buy this cereal awareness, recall, and greater persuasive sometime. impact. Indeed, several advertising research- I might buy this cereal, but I don’t ers have suggested that liking an ad is one of know when. the most important factors in determining its impact (Haley, 1994). I’ll probably never buy this cereal. Multidimensional measures of emotion I wouldn’t eat this cereal even if are useful. Micu and Plummer (2010) somebody gave it to me. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  400 Part Four Research Applications The scale allows advertisers to see how special store, which in most cases is a special consumers’ buying preferences change dur- trailer or field service conference room furn- ing and after the campaign. ished to look like a store. Subjects are then Perhaps the most reliable methods of post- shown a program containing some test com- testing are those that measure actual sales, mercials, given more chits, and allowed to direct response, and other easily quantifiable shop again. Changes in pre- and post- behavior. For newspaper or magazines, direct exposure choices are recorded. Symphony response might be measured by inserting a IRI (formerly Information Resources Inc.) coupon that readers can mail in for a free has taken this concept online. It offers a vir- sample or redeem online. Different forms of tual shopping experience where consumers an ad might be run in different publications can pick up, rotate, and buy products just and websites to determine which elicits the as they would in a real store. most inquiries. Another alternative suitable Actual sales data can be obtained in many for use in both print media advertising and ways. Consumers may be asked directly, electronic media advertising is to include a “Which brand of breakfast cereal did you toll-free telephone number or an Internet most recently purchase?” However, the find- address that consumers can contact for more ings from this survey would be subject to information or to order the product. error due to faulty recall, courtesy bias, and Three studies illustrate how researchers can so forth. For this reason, more direct methods examine behavioral response. Bates and Buck- are usually preferred. If enough time and ley (2000) examined the influence of exposure money are available, direct observation of to TV commercials that urged people to return people’s selections in the cereal aisles at a their 2000 census forms on the actual rate of sample of supermarkets can be a useful source returned forms. They found that exposure to of data. Store audits that list the total number advertising was related to knowing more of boxes sold at predetermined times are about the census but that there was no rela- another possibility. Last, and possibly most tionship between exposure to the ads and actu- expensive, is the household audit technique, ally returning a form. Burton, Lichtenstein, in which an interviewer visits the homes of a and Netemeyer (1999) discovered that expo- sample of consumers and actually inspects sure to an advertising sales flyer for retail their kitchen cupboards to see which brands supermarkets resulted in more than a 100% of cereals are there. In addition to the audit, a increase in the number of advertised products traditional questionnaire is used to gather that were purchased. More recently, Bronnen- further information about the respondents’ berg, Dube, and Mela (2010) studied whether feelings toward the commercials. DVR usage (specifically fast-forwarding through commercials) hurt the sales of advertised pro- MEDIA RESEARCH ducts. They monitored DVR behavior and household purchases over a one- and two- Two important terms in media research are year period. Interestingly, they could find no reach and frequency. Reach is the total statistical support for the assertion that DVRs number of households or persons that are affect the purchasing of advertised products. supposedly exposed to a message in a partic- Some research companies measure direct ular medium at least once over a certain time response by means of a laboratory store. period. Reach can be thought of as the cumu- Usually used in conjunction with theater test- lative audience, and it is usually expressed as ing, this technique involves giving people a percentage of the total universe of house- chits with which they can buy products in a holds that have been exposed to a message. Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 401 For example, if 25 of a possible 100 house- would probably be a factor in deciding holds are exposed to a message, then the which one is the better buy. reach is 25%. Frequency refers to the num- Media research falls into three general ber of exposures to the same message that categories: studies of the size and composi- each household or person receives. Of tion of an audience of a particular medium course, not every household or person in or media (reach studies), studies of the rela- the sample will receive exactly the same num- tive efficiency of advertising exposures pro- ber of messages. Consequently, advertisers vided by various combinations of media prefer to use the average frequency of expo- (reach and frequency studies), and studies sure, expressed by this formula: of the advertising activities of competitors. Total exposures Audience Size and Composition for all households=persons Average Analyses of audiences are probably the most ¼ Reach frquency commonly used advertising studies in print and electronic media research. Because adv- Thus, if the total number of exposures for a ertisers spend large amounts of money in the sample of households is 400 and the reach is print and electronic media, they have an 25, the average frequency is 16. In other understandable interest in the audiences for words, the average household was exposed those messages. In most cases, audience 16 times. Notice that if the reach were information is gathered using techniques 80%, the frequency would be 5. As reach that are compromises between the practical increases, average frequency drops. (Maxi- and the ideal. mizing reach and frequency is directly related As noted in Chapter 13, the audience size to the amount of money invested in an adver- of a newspaper or magazine is commonly tising campaign.) measured in terms of the number of copies A concept closely related to reach and fre- distributed per issue. This number, which is quency is gross rating points (GRPs), intro- called the publication’s circulation, includes duced in Chapter 14. GRPs are useful when both copies delivered to subscribers and it comes to deciding between two media alter- those bought at newsstands or from other natives. For example, suppose Program A has sellers. Because a publication’s advertising a reach of 30% and an average frequency of rate is determined directly by its circulation, 2.5, whereas Program B has a reach of 45% the print media have developed a standard- and a frequency of 1.25. Which program ized method of measuring circulation and offers a better reach–frequency relationship? have instituted an organization, the Audit First, determine the GRPs of each program Bureau of Circulations (ABC), to verify that using the following formula: a publication actually distributes the number GRPs ¼ Reach  Average frequency of copies per issue that it claims. (The specific procedures used by the ABC are discussed For A: later in this chapter.) GRPs ¼ 30  2:5 ¼ 75:00 Circulation figures are used to compute the cost per thousand (CPM) of various pub- For B: lications. For example, suppose Newspaper GRPs ¼ 45  1:25 ¼ 56:25 X charges $1,800 for an advertisement and has an ABC-verified circulation of 180,000, In this example, Program A scores better in whereas Newspaper Y has a circulation of the reach–frequency combination, and this 300,000 and charges $2,700 for a space of Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  402 Part Four Research Applications Table 15.2 Determining Advertising Efficiency from Ad Cost and Circulation Data Newspaper X Newspaper Y Ad cost $1,800 $2,700 Circulation 180,000 300,000 $1,800 $2,700 Cost per thousand ¼ $10.00 ¼ $9.00 180 300 circulated copies the same size. Assuming the newspapers are (as has been discussed) because of the diffi- the same in all respects except for advertising culty respondents often have in recalling spe- costs, Table 15.2 shows that Newspaper Y is cific content. a slightly more efficient advertising vehicle A second technique is aided recall, where because of a lower CPM. the interviewer names several publications Note that this method considers only the and asks whether the respondent has read number of circulated copies of a newspaper any of them lately. Each time the respondent or magazine. This information is useful, but claims to have read a publication, the inter- it does not necessarily indicate the total num- viewer asks whether he or she remembers ber of readers of the publication. To estimate seeing the most recent copy. The interviewer the total audience, the circulation figure must may jog a respondent’s memory by describ- be multiplied by the average number of read- ing the front page or the cover. Finally, the ers of each copy of an issue. This information respondent is asked to recall anything that is obtained by performing audience surveys. was seen or read in that particular issue. In A preliminary step in conducting such a variation on this process, masked recall, surveys is to operationally define the concept respondents are shown the front page or magazine reader or newspaper reader. There the cover of a publication with the name are many possible definitions, but the one blacked out and are asked whether they most commonly used is fairly liberal: A remember reading that particular issue. reader is a person who has read or at least Those who respond in the affirmative are looked through an issue in a certain time asked to recall any items they have seen or frame. read. Three techniques are used to measure The third technique, called the recogni- readership. The most rigorous is the unaided tion method, involves showing respondents recall method, in which respondents are the logo or cover of a publication. For each asked whether they have read any newspa- publication respondents have seen or read, pers or magazines in the past month (or the interviewer produces a copy and the other time period). If the answer is “yes,” respondents leaf through it to identify the subjects are asked to specify the magazines articles or stories they recognize. Respon- or newspapers they read. When a publication dents who definitely remember reading the is named, the interviewer attempts to verify publication are counted in its audience. To reading by asking questions about the con- check the accuracy of the respondent’s mem- tents of that publication. The reliability of ory, dummy articles may be inserted into the the unaided recall method is open to question interviewer’s copy of the publication; Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 403 Table 15.3 Determining Ad Efficiency from an Extended Database Newspaper X Newspaper Y Ad cost $1,800 $2,700 Circulation 180,000 300,000 CPM $10.00 $9.00 Number of people 630,000 540,000 who read the issue (3.5 readers per copy) (1.8 readers per copy) Revised CPM $2.86 $5.00 respondents who claim to have read the Unfortunately, a practical and reliable dummy items thus may be eliminated from method for measuring the number of expo- the sample or given less weight in the analy- sures per issue has not yet been developed. sis. Many advertising researchers consider Perhaps the most important gauge of the recognition technique to be the most advertising efficiency is the composition of accurate predictor of readership scores. the audience. It matters little if 100,000 When the total audience for each maga- people see an advertisement for farm equip- zine or newspaper is tabulated, the advertiser ment if only a few of them, or none, are in can determine which publication is the most the market for such products. To evaluate efficient buy. For example, returning to the the number of potential customers in the example of Table 15.2, let’s suppose that audience, an advertiser must first conduct Newspaper X and Newspaper Y have the research to determine the demographic audience figures listed in Table 15.3. Based characteristics of people who purchase a par- on these figures, Newspaper X, because it ticular product. For example, potential cus- has more readers, would be considered the tomers for beer might be described typically more efficient choice. as males 18–49; those for fast-food restau- Another variable to consider in determin- rants might be households in which the ing the advertising efficiency (or media effi- primary wage earner is 18–35 with at least ciency) of a newspaper or magazine is the two children under 12. The demographic number of times a person reads each issue. characteristics of typical consumers are then For example, imagine two newspapers or compared with the characteristics of a pub- magazines that have the same number of lication’s audience for the product. The cost readers per issue. Publication A consists pri- of reaching this audience is also expressed marily of pictures and contains little text; in CPM units, as shown in Table 15.4. The people tend to read it once and not look at numbers indicate that Newspaper X is it again. Publication B, on the other hand, slightly more efficient as a vehicle for reach- contains several lengthy and interesting arti- ing potential beer customers and much more cles; people pick it up several times. Publica- efficient in reaching fast-food restaurant tion B would seem to be a more efficient patrons. advertising vehicle because it provides sev- Because of the ephemeral nature of radio eral possible exposures to an advertisement and television broadcasts, determining audi- for the same cost as Publication A. ence size and composition in the electronic Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  404 Part Four Research Applications Table 15.4 Calculation of Ad Efficiency Incorporating Demographic Survey Results Newspaper X Newspaper Y Ad cost $1,800 $2,700 Circulation 180,000 300,000 CPM $10.00 $9.00 Number of people who read average issue 630,000 540,000 Number of potential beer drinkers 150,000 220,000 Number of potential fast-food customers 300,000 200,000 CPM (beer drinkers) $12.00 $12.27 CPM (fast-food customers) $6.00 $13.50 A CLOSER LOOK Target Audience Language Each mass medium has several target audiences, audience target is generally shortened to the usually one broad target and many others for words “target” or “demo,” so it is common to specific elements, such as a TV program, radio hear a question such as “What’s the demo for show, or newspaper insert. In programming, this program?” editorial, and advertising sales discussions, the media poses special problems for advertising to using only one medium. However, adver- researchers. One problem in particular tisers with a good budget often use several involves the use of the CPM measure for media simultaneously, which is known as media planning. The various measures of synergistic advertising or synergistic mar- program audience discussed in Chapter 14 keting. The task is to determine which may or may not reflect the number of people media combination will provide the greatest who actually watch a TV program. Industry reach and frequency for the product or ser- experts suggest that engagement measures vice. A substantial amount of media research (discussed previously in this chapter) are has been devoted to this question, much of it more useful for advertisers (Marich, 2008). concentrated on the development of mathe- matical models of advertising media and their audiences. The mathematical deriva- Frequency of Exposure tions of these models are beyond the scope in Media Schedules of this book. However, the following para- In some situations, advertisers with a small graphs describe in simplified form the advertising or promotion budget are limited concepts underlying two models: stepwise Copyright 2012 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.  Chapter 15 Research in Advertising 405 analysis and decision calculus. Readers who medium. In addition, a schedule’s objective wish to pursue these topics in more rigorous function value takes into account such vari- detail should consult Rust (1986) and the ables as the probability that the advertise- Internet (see “Using the Internet” at the end ment will be forgotten, the total cost of the of this chapter for suggestions to find more media schedule compared with the adverti- information about synergistic advertising/ ser’s budget, and the “media option source marketing). effect”—that is, the relative impact of expo- Stepwise analysis is called an iterative sure in a particular advertising vehicle. (For model because the same series of instructions example, an advertisement for men’s clothes to the computer is repeated over and over is likely to have more impact in Gentlemen’s with slight modifications until a predeter- Quarterly than in True Detective.) mined best or optimal solution is reached. In the last 20 years or so, many new The Young & Rubicam agency pioneered