Inquiry Response Rates. Cost and Revenue Per Inquiry of Repetitive Print Advertising

Arch G. Woodside, University of South Carolina
ABSTRACT - How many exposures are enough? Theoretical arguments by Krugman (1972) propose that three exposures may be optimal for remembering broadcast advertising messages. Evidence presented here suggests that two exposures may be best for producing the highest direct inquiry response rates and lowest cost per inquiry from print advertisements.
[ to cite ]:
Arch G. Woodside (1981) ,"Inquiry Response Rates. Cost and Revenue Per Inquiry of Repetitive Print Advertising", in NA - Advances in Consumer Research Volume 08, eds. Kent B. Monroe, Ann Abor, MI : Association for Consumer Research, Pages: 428-431.

Advances in Consumer Research Volume 8, 1981      Pages 428-431


Arch G. Woodside, University of South Carolina


How many exposures are enough? Theoretical arguments by Krugman (1972) propose that three exposures may be optimal for remembering broadcast advertising messages. Evidence presented here suggests that two exposures may be best for producing the highest direct inquiry response rates and lowest cost per inquiry from print advertisements.


How are costs and revenues affected by repetitive advertising aimed at producing direct response inquiries and sales? Unfortunately, published reports appear to be nonexistent on the costs, number of inquiries, and revenues related to each of repeated insertions of the same advertisement in the same magazine or newspaper.

Simon (1979) has pointed out that recall and attitude measures of advertising are simply proxies for sales, and ultimately they must be related to sales. His analysis of Pomerance and Zielske's (1958; Zielske, 1959) field laboratory study comparing two 13-week media schedules indicated that "the amount of advertising impact (as measured by recall-weeks) is much higher for the every-four-weeks schedule than for the weekly schedule" (Simon 1979, p. 418). Simon noted some "severe" qualifications and criticism of this study including the following two: 1) The data were generated by just one advertising medium, direct mail, and direct mail would not be a likely medium for a staple food product (the product advertised in the study); and 2) A sales measure might show a different sort of response pattern than the recall measure used. Also, a recall measure has a built-in saturation limit (100%) that is not relevant for actual sales purposes.

Ray and his associates (Ray and Sawyer 1971, Swinyard and Ray 1977) report increases in purchase intentions of later versus first and second insertions. In the Swinyard and Ray (1977) study, female household subjects were asked the following question after receiving four direct mailings of appeals for blood donors and Red Cross volunteer workers: "If you were going to volunteer work for one of these organizations, which one would be your first choice? Second choice?" Behavioral intentions were coded as the proportion of respondents rating the Red Cross as first choice among five name social health organizations. Intention to volunteer for the Red Cross increased from 5% to 17.4% to 22.7% from the first to the second to the fourth exposures for the subjects receiving the direct mail appeals.

Information on costs, number of inquiries, and estimated revenues from a print advertising campaign which included multiple insertions of the same advertisement in the sane magazines and newspaper is reported in the present article. The immediate objective of the campaign was to maximize the number of direct inquiries with the lowest possible costs. The ultimate objective of the campaign was to product the greatest possible sales given a fixed advertising budget.

The main purpose of the research was to measure the changes in costs per inquiry (CPI) of several insertions of the same ad in each of several magazines. The same ad was repeated 3 to 13 times in 7 magazines and 1 newspaper.

Assuming a constant cost per advertisement insertion over a campaign, decreases in CPI for the second and third insertions compared to the first insertion were hypothesized. Then CPI was expected to increase for later insertions. Thus, a "U" shaped relationship between CPI and the number of insertions was hypothesized.

What rationales can be provided for the expected increases in the number of inquiries for the second and third insertions compared to the first (and thus decreases in CPI given a constant cost per insertion)? Increases in recall (Swinyard and Ray 1977), and brand awareness (Stewart 1964) for second, third, and later advertisements versus an initial advertisement found in prior research provide one rationale for the hypothesis. The theoretical work of Krugman (1962; 1968; 1972) provides another rationale. Krugman develops a cognitive theory to indicate "why three exposures may be enough" to produce behavior and thus sales.


The media costs, inquiry response races, and estimated revenues (for repeated insertions in two magazines) were studied for 7 magazines and 1 newspaper which included three or more insertions of the same advertisement. The advertisement was designed to promote a direct response inquiry to receive a "free South Carolina Trip Kit." The advertisement is shown in the exhibit:


Each advertisement included a unique room number in the address on the coupon to identify the magazine and specific insertion. The number of inquiries was counted per insertion for each magazine. The cost of the advertisement for each insertion was divided by the number of inquiries received, resulting in a CPI.

CPI, similar to CPM (cost per 1,000's of circulation), is often used as a measure for comparing the performances of competing magazines. A recent ad for Sport magazine compares its CPM with the CPMs for Playboy, Popular Mechanics, Popular Science, and five other magazines (Advertising Age 1980, p. 11). Data on CPI are collected by individual advertisers and are often used to answer the following question: Did the magazine or newspaper ('book") offering a low CPM also provide a low CPI in comparison to other books? The implication is that a book promoted as a "good buy" for an ad placement because of its comparatively low CPM may be, in fact, a poor buy because it delivers a comparatively high CPI for the specific ad placed. Also, a book having a comparatively high CPM may be, in fact, a good buy because its CPI is also comparatively low (cf. Woodside and Reid 1974).

Revenue per inquiry (RPI) was estimated for each of multiple insertions for two magazines in the study. A mail questionnaire was sent to samples of inquirers for each insertion six months after nearly all the inquiries were received (in October). The samples were selected systematically from each "room" for each insertion, The questionnaire had 18 questions including questions on visiting South Carolina and expenditures in South Carolina. RPI was computed by first multiplying the total number of inquiries, e.g., 500, by the percent of those survey respondents who visited South Carolina, e.g., 20 percent: 500 x .50 = 100. Second, the estimated number of visitors is multiplied by the average expenditure per visitor, e.g., $30. The result is the estimated total revenue: 100 x $30 = $3,000. RPI is estimated by dividing the estimated total revenue by the total number of inquiries: $1,000/500 = $6.

The average rate of response of questionnaires sent was 50 percent for the eight insertions studied following a second and third mailing of the questionnaire to nonrespondents. (Percentages of visits to South Carolina were not statistically different between respondents from the three mailings for each magazine.)


The hypothesis was partially supported. The CPI declined from the first to the second insertions for 7 of the 8 books as shown in Figures 1 and 2. However, the relationship between CPI and the number of insertions was more "V" shaped than "U" shaped. The CPI was greater for the third insertion compared to the second and to the first insertions for 6 of the 7 books, exhibiting a "V" shaped relationship between CPI and the number of insertions.

The inquiry response rates as percents of circulation are listed in Table 1 for the three insertions in Popular Mechanics. The difference (.02%) in response rates between the January and February insertions is small but meaningful given the large circulation of the magazine. Thus, CPI drops $0.17 or 5% with a .02% increase in inquiry response rate.

Note that RPI changes in the same direction as CPI for three insertions in Popular Mechanics. The poor evaluation implied by the relatively high CPI of $4.27 of the March insertion is offset by the relatively high estimated RPI of $217 for the same insertion.







A total of 13 weekly insertions of the ad was placed in TV-Guide starting on January 6, 1979. The resulting CPI values are shown in Figure 2. Note that a "V" shaped relationship exists for CPI and the first three insertions. CPI increases peaking at $5.59 for the eighth insertion, and then declines to its lowest point ($2.29 for the last insertion). These findings suggest a substantial seasonal effect of insertions on CPI. The early January ad may have attracted many inquiries intending to visit South Carolina for the spring season (and famous coastal gardens). The March 31 insertion (the last insertion) may have attracted many inquiries planning summer vacation trips. (State traced data indicate that June, July, and August followed by April are the four months having the greatest percent on nonresident travelers in South Carolina.) Thus, insertions in late February may be expected to have the highest CPI's. Unfortunately, only the limited information shown in Figure 2 is available to support this hypothesis.

Details on the circulation, costs for each ad insertion, and the percent inquiry response rate per insertion are presented in Table 2 for seven books. The percent inquiry response rates for the second compared to the first or third insertions were higher for all the books except Popular Science. The response rate declined for Popular Science from .68% to .50% to .38%. Thus, CPI increased for the second and third insertions in Popular Science since the cost for each ad insertion was constant.



In Table 3, the response races, CPI, and available RPI are listed for the ad insertions in TV-Guide. Note that RPI continually declines substantially from $217 to $133 for the five insertions included in the survey research study This may suggest that early ad insertions in the campaign may produce greater RPI than later insertions. Unfortunately, the RPI for the last insertion was not estimated to learn if the substantial drop in CPI between insertion 12 and 13 was coupled with a substantial increase in RPI.




Krugman offers the following thesis as why three exposures may be enough, "campaign effects based on, say, 20 to 30 exposures are only multiples or combinations of what happens in the first few exposures." For television advertising, he believes that the cognitive task presented by the first exposure is contaminated by a "what is it?" response - an attempt to define or understand the ad stimulus. The first exposure is likely to cause high recognition scores for the second exposure. The second exposure is likely to be dominated by the evaluative response of "what of it?" where the consumer judges the personal relevance of the now somewhat familiar ad. The third exposure becomes a reminder but also the beginning of disengagement.

Assuming that print advertising requires more consumer involvement than television commercials (cf. Krugman, 1965; Robertson, 1976) might suggest that such learning with involvement requites fewer exposures for optimal effect. Readers have the opportunity to linger and ask what is it?, what of it?, and what should I do about it? In a review of the literature Sawyer and Ward (1979) concluded that "the existing evidence suggests that the audience-control-led input of print advertising may generate more immediate learning and cognitive responses than low involvement "broadcast exposures" (p. 289). Thus, higher involvement in print versus broadcast advertisements is one rationale for the finding that CPI is lowest for the second of three print exposures.

Clearly, more unobtrusive field research is needed than available currently to learn the behavioral effects of repetitive advertising exposures. The results of the reported study support an inverted V-curve relationship between inquiry response rate and exposures, and thus, a V-curve relationship between CPI and exposures given a constant cost per exposure. Further evidence with other products and services (and in tourism advertising) is needed to support or refute this prediction before the conclusion is reached that two exposures may be enough in print media vehicles.

Do the results from the present study really suggest that two exposures are enough? This implication may be inaccurate. Managers may discover that multiple insertions of related but different ads produce the greatest sales returns per advertising dollar. Varying the size of the ad and the use of color might increase the sales impact of multiple exposures. Unfortunately, the investment in research to learn the sales impact of repetitive advertising is no higher than the low levels reported by Stewart in 1964. Stewart's field experiment on repetitive newspaper advertising effects on awareness, attitude, and sales remains the best work available.


Advertising Age (1980), 51 (April 7).

Krugman, Herbert E. (1965), "The Impact of Television Advertising: Learning Without Involvement," Public Opinion Quarterly, 29, 349-356.

Krugman, Herbert E. (1968), "Processes Underlying Exposure to Advertising,'' American Psychologist, 23, 245-253.

Krugman, Herbert E. (1972), "Why Three Exposures May Be Enough," Journal of Advertising Research, 12 (December), 11-14.

Pomerance, Eugene and Zielske, Hubert (Hugh) (1958), "How Frequently Should You Advertise?" Media/Scope (September).

Ray, Michael L. and Sawyer, Alan G. (1971), "Repetition in Media Models: A Laboratory Technique," Journal of Marketing, 8 (February), 20-29.

Robertson, Thomas S. (1976), "Low-Commitment Consumer Behavior,'' Journal of Advertising Research, 16 (April), 19-24.

Sawyer, Alan and Ward, Scott (1979), "Carry-over Effects in Advertising Communication," in Research in Marketing, Volume 2, Greenwich, CT: JAI Press, 259-314.

Simon, Julian L. (1979), "What do Zielske's Real Data Really Show About Pulsing?" Journal of Marketing Research, 16 (August), 415-420.

Stewart, John B. (1964), Repetitive Advertising in Newspapers, Boston, Division of Research, Graduate School of Business Administration, Harvard University.

Swinyard, Willie R. and Ray, Michael L. (1977), "Advertising-Selling Interactions: An Attribution Theory Experiment,'' Journal of Marketing Research, 14 (November), 509-516.

Woodside, Arch O. and Raid, David M. (1976), "Tourism Profiles Versus Audience Profiles: Are Upscale Magazines Really Upscale?" Journal of Travel Research, 12 (Spring), 17-25.

Zielske, Hubert (Hugh) (1959), "The Remembering and Forgetting of Advertising," Journal of Marketing, 23 (January), 239-243.