Advertising
Advertising
I. Economic AspectsLester G. Telser
II. Advertising ResearchCharles K. Ramond
I ECONOMIC ASPECTS
As more resources are spent on advertising, there is increasing controversy about its consequences. Only the economic aspects of advertising, however, are within the scope of this article. First, we will give some facts about the use and importance of advertising as a means of promotion; the data relate mainly to the United States, but we will also cite some figures for the United Kingdom and Canada. Second, we will show why firms use advertising. We will then discuss how advertising affects price competition and the size of firms and why advertising is used in the Soviet Union. Finally, we will consider those communications media that depend on advertising as their major source of revenue and discuss the relation between advertising and consumer sovereignty.
Scope of advertising
In 1960 total advertising outlays in the United States were estimated to be about $12,000 million, which was 2.3 per cent of gross national product (GNP). Since in 1947 advertising was 1.8 per cent of GNP, there is a slight upward trend in the post-World War II period. However, too much should not be made of this evidence. It covers a short period of time, advertising figures prior to 1935 are unreliable, and the figures just preceding World War II show that the relation between advertising and GNP was about the same as in the late 1950s (Blank 1963).
In 1954 Canadian advertising outlays were 1.6 per cent of GNP, and comparable figures for the United Kingdom are of the same order. In both of these countries advertising outlays constitute a smaller percentage of either GNP or national income than they do in the United States. So far as can be judged from the limited data available, advertising is a less important means of promotion in most other countries than it is in the United States.
There are four major channels of advertising. The printed media, such as magazines and newspapers, contain both advertising and editorial material and are sold directly to the public. The audio–visual media, such as television and radio, are supported primarily by advertising receipts. These media do not, in many countries, collect fees directly from the public. Instead, they aim to attract an audience for advertisers by providing the public with free entertainment. Direct-mail and outdoor advertising attempt to attract the public’s attention and make direct sales appeals. In the United States in 1959, advertising in the printed media accounted for between 45 and 50 per cent of total estimated advertising expenditures. Television and radio received about 14 per cent and 6 per cent, respectively, of advertising expenditures. Direct mail accounted for 14 per cent of the advertising bill, and outdoor advertising was about 2 per cent. The remaining expenditures were on sale display, advertising departments, and the like.
Modern advertising began during the early part of the nineteenth century, after the cheap daily newspaper and the national magazine were introduced. Radio in the 1920s and television in the 1950s caused major changes in the composition of advertising outlays. As a result of these innovations, the cost per advertising message has sharply decreased and the number of advertising messages has risen more than would be implied by the increased dollar expenditures.
The ratio of advertising to sales differs considerably among products. Industrial products, which are purchased primarily by a relatively small number of firms, have a much lower ratio of advertising to dollar sales than do consumer products. Promotion of industrial products depends primarily on salesmen. The products advertised most heavily in relation to dollar sales in the United States in 1957 were toilet preparations (14.7 per cent of sales), drugs and medicines (10.3 per cent), and soaps (7.9 per cent).
Canadian figures that give advertising as a percentage of sales by product category are roughly the same as the U.S. figures. Moreover, a study of advertising relative to sales in 1940 and 1941 by the U.S. Federal Trade Commission (1944) shows that the ratios by product category in the United States are quite close to comparable figures for the year 1935 in the United Kingdom as shown by the Kaldor and Silverman study (1948). That the pattern of advertising outlay with respect to commodity is similar in all three countries is an important finding and deserves a fuller explanation than has so far appeared. Equally important would be an explanation for the differences among the advertising intensities of various products.
Why firms advertise
A correct but umlluminating explanation of advertising is that firms find it profitable. Advertising, however, is only one, and not even the most important, method of sales promotion. Although some products are promoted mainly by advertising, personal selling still accounts for the largest share of promotional outlays. Advertising is a much less labor-intensive method of promotion than are many alternatives for accomplishing the same end. Therefore, it is a technique well suited to sell those products that are, or can be, widely used and for which potential customers are not readily distinguished from the rest of the population. A firm deliberates over the same kinds of factors in allocating its funds to different advertising media as in allocating funds to other methods of sales promotion.
For an understanding of why firms advertise, it is helpful to see how advertising affects sales. Advertising affects sales in two stages. In the first stage a firm buys space in the press or time on television in order to convey advertising messages that create awareness in potential customers. In the second stage these advertising messages induce sales with varying degrees of effectiveness. In order to determine its advertising budget a firm must both decide how many advertising messages of each type it should buy and estimate their effectiveness.
The audience of an advertising medium approximates the number of advertising messages that can be received via that medium. For printed media, such as magazines or newspapers, the paid circulation gives a first approximation of the audience. However, the readership of a newspaper or magazine may exceed the circulation, and, conversely, the number of readers who take note of the advertising may fall short of the circulation. Thus, paid circulation is at best only an approximation to the number of advertising messages received via a printed medium. The audience of a television or radio program is harder to estimate, and special research techniques have been devised for this purpose. Audience size (along with composition) is one of the key determinants of advertising rates. Given this information about an advertising medium, a company can calculate the cost and estimate the effectiveness of advertising messages conveyed by that medium.
The number of advertising messages is the relevant physical measure of the quantity of advertising. This is why advertising per dollar of sales is not always a reliable measure of the quantity of advertising. The absolute advertising outlay and the number of messages transmitted may be very large although the ratio between advertising expenditure and sales is very low. For example, advertising as a percentage of sales is very low for automobiles. However, in 1957 U.S. consumers received nearly $400 million of automobile advertising messages, half from auto manufacturers and half from auto dealers.
Advertising messages can be effective indirectly. Thus, potential customers can learn about products, even though not directly exposed to advertising messages, by hearing about these products from others. By a chain reaction, advertising messages can stimulate transmission of a sizable volume of information (Katz & Lazarsfeld 1955; Ozga 1960).
The preceding analysis explains some dynamic aspects of advertising. There is typically a delayed response to advertising for several reasons. First, many of the advertised products are not of great importance to consumers; thus, a certain amount of repetition or redundancy is necessary to create awareness of the product. Advertising in smaller amounts is not likely to pass the threshold of awareness and, therefore, is likely to be ineffectual. Second, in addition to the problem of making consumers aware of the product, there is the further problem of overcoming their inertia. Although inertia increases the delay between advertising messages and sales, it also makes the effects of advertising persist beyond the time the messages are disseminated. Thus, advertising expenditure can be thought of as an investment to create an asset—sometimes called good will. This asset yields a return for some period of time, it depreciates like a capital good, and it requires maintenance. The marginal rate of return on advertising, as well as the rate of depreciation, can be calculated. Nerlove and Arrow (1962) provide theoretical analysis and Telser (1962) shows empirical results along these lines.
This approach to advertising has several implications. First, we can expect new products to be more heavily advertised than established products. Second, continuous advertising of established products is necessary because new consumers enter the market and others either leave it or forget about the product. Third, in an expanding economy there will be relatively more advertising than in one that is stable or declining.
Advertising versus price competition
One of the criticisms of advertising is that it is wasteful. If firms compete by offering to sell at low prices, then buyers benefit. Such competition among sellers results in products of given quality being sold at the lowest price. However, it is argued, if sellers compete for customers by advertising, then buyers do not benefit by obtaining a lower price; on the contrary, they pay a higher price to reimburse the sellers’ advertising expenses. In its crude form this argument has little merit; if some buyers do not wish to purchase advertised goods, there will generally be sellers who find it profitable to cater to their demand. Then some buyers will seek out the lowpriced sellers who do not advertise, whereas other buyers will choose to pay higher prices for well-known goods.
A more sophisticated version of the argument that advertising may increase prices assumes that by advertising, a company can reduce the elasticity of demand for its product [seeElasticity]. Advertising can change the character of the demand so that customers become less sensitive to price and the advertiser obtains a loyal clientele. This makes possible higher prices and larger profits. Whether advertising can as a matter of fact create brand loyalty is not known with certainty although it is a possibility. There is some evidence worth bringing to bear on this question. We noted above that toilet preparations are the most intensively advertised consumer articles. If advertising tends to create brand loyalty, market shares of cosmetics and similar products should be more stable than market shares of less advertised items, such as branded food products. However, a study of the four leading brands of each of a number of articles in these two product classes over a 13-year period showed that market shares of toilet preparations were markedly less stable than were shares of branded foods. There was, in addition, a substantial turnover of brands in the toilet-preparation class. Because new brands are advertised much more intensively than are established brands, the high ratio, among toilet preparations, of advertising to sales may reflect the short life and high turnover of brands rather than brand loyalty. If this is true, the high intensity of cosmetics advertising results from the lack of brand loyalty to such products as compared with branded food items.
It is by no means obvious that advertising necessarily reduces the price elasticity of demand. Increased advertising may bring a firm new customers whose preferences for the product are weaker than those of the old customers. The new customers are consequently more sensitive to price, and the increased advertising thus increases the price elasticity. Although advertising by a given firm may reduce the price elasticity for its goods by strengthening preferences, competitive advertising has the opposite effect. On balance, increased advertising may increase price elasticity.
Advertising can intensify competition among retailers, thereby reducing retail prices. Since shoppers can compare the various retail prices of a well-known brand at different stores more easily than they can those of a nonstandard item, competition is keener among retailers selling the well-known item. This forces the retail prices of advertised articles closer to invoice costs. Retailers’ advice influences consumer choice of nonstandard items more than the choice of advertised articles. Hence, advertised goods need less promotion at the point of purchase. As a result, the character of both retailing and wholesaling has changed markedly. Sales personnel can be less skilled, and discount houses and self-service stores have become feasible.
Advertising and the size of firms
Defenders of advertising often claim that advertising, by creating mass markets, makes it possible to produce goods at lower unit costs. It is doubtful that this claim is supported by the evidence available. First, many companies that rely heavily on advertising operate plants of different sizes; this is inconsistent with production economies of scale. Second, there are many industries in which a few firms are quite large but which use little advertising; examples of these are to be found primarily among producers of industrial goods. Third, as was noted above, there are some heavily advertised commodities manufactured primarily by small firms; the leading examples are toilet preparations.
There are, however, certain important producers of consumer goods who are heavily dependent on advertising and account for sizable fractions of their industry sales. Examples are producers of breakfast cereals, soaps, automobiles, cigarettes, razors and razor blades, soft drinks, canned soups, baby foods, and distilled liquor. To sum up, there is considerable evidence against the general proposition that advertising, by expanding the market, makes it possible to lower unit costs, and some evidence to support the idea that in some industries, high concentration of output is associated with considerable advertising.
An understanding of the association between advertising and concentration in the industries just cited begins with an examination of the advertising rate schedule. Advertising rates rise as audience increases, but less rapidly. This can give the national advertiser of a given product an advantage over the local advertiser, because the former has a lower promotional cost. The cost advantage to the large firm makes the growth of the small firm more difficult. In addition, sales may rise more rapidly in response to advertising expenditures over some range. If this is true, it reinforces the tendency brought into play by the structure of the advertising rates. For these reasons some firms will make and advertise a large number of consumer goods in order to obtain the savings of largescale advertising. Certain large firms in the food and drug industries owe their size in part to these economies.
Advertising in the Soviet Union
Many of the effects of advertising stand out more clearly in the light of the Soviet Union’s experience with forgoing the use of both trademarks and advertising. Because of adverse experience with this policy, the Soviet Union has come to encourage advertising and the use of brand names. Its reasons for abandoning the old policy and adopting the new are very instructive. First, anonymous producers had less incentive to maintain quality because shoddy goods were not so easily identified by consumers. The government can now shift some of the burden of quality control to factories that are forced to trademark their products and can lose customers if their goods prove unsatisfactory. Second, when advertising is encouraged, information about new goods is disseminated more rapidly, and innovations are stimulated. Because in the past it was not possible to use advertising to generate demand for new products, the incentive to contrive new products was discouraged. Third, a more efficient marketing system, which conserves scarce resources of the state as well as saving the consumer’s time and trouble, is made possible by the use of advertising. Thus, the Soviet Union can now use self-service stores and distribute a given volume of goods with a smaller amount of labor. Goldman (1960) contains a careful account of Soviet advertising experience.
Advertising and the communications industry
In the United States, the broadcasting industry receives virtually all of its revenue from advertisers. Newspapers and periodicals obtain some two-thirds of their revenue from advertising and the balance from sales and subscriptions. Clearly, a substantial part of the entertainment and news provided the public in the United States is paid for directly by advertisers and only indirectly by the public. This creates concern about the quality of the entertainment and the degree to which it reflects the public’s taste. There is, in addition, the question of whether advertising affects freedom of the press. Finally, because most television advertising in the United States is purchased by relatively few companies and only three national networks exist, there are some special problems of economic policy with regard to regulation of the industry.
The public registers its taste for drama and motion pictures directly, by the purchase of tickets to those it likes. Producers have a direct financial incentive to cater to public taste. Although viewers and listeners do not pay directly for radio and television programs, it cannot be doubted that similar considerations guide the producers of these programs. Since advertisers desire large audiences and are willing to pay more for them, producers of radio and television shows have strong incentives to provide what they expect will be popular. The reason the cost of radio and television entertainment is not collected directly from the audience is the economic fact that it is cheaper to collect this cost from advertisers. If certain experiments with closed-circuit television prove successful, direct collection from viewers will make subscription television a profitable enterprise. Yet it is by no means obvious that advertising would be absent from subscription television. People buy magazines and newspapers directly, and both these media contain advertising. Newspapers that did not include advertising have either changed their policy or failed. Aside from the absence of advertising revenue, this suggests that people demand certain kinds of advertising.
There is still more direct evidence that people want certain kinds of advertising. Newspaper advertising rates have always been lower for local advertisers than for national advertisers. Since the cost to a newspaper of local advertising is, if anything, higher than the cost of national advertising, and since national advertisers have several alternatives not available to local advertisers, this phenomenon has mystified students of advertising for a long time. Ferguson (1962) provides an explanation. Most of the local advertising gives newspaper readers information about goods available from local retailers and about the terms of sale. Newspaper readers are as eager to learn about these matters as about news. Hence the more advertising of this kind a newspaper contains, the larger its circulation. Therefore, local advertising rates are lower than those for national advertising, which generally does not have the same stimulus on circulation. Perhaps the communications media would freely carry items as news about consumer goods and services if there were no advertising, just as they now review books and motion pictures.
Partly for these reasons, the more sophisticated critics of advertising take the position that advertisers and television producers cater unduly to popular taste and do not experiment boldly enough with higher art forms. Because there are only three major television networks in the United States and few television channels per city, these critics maintain that the desire to reach a maximum audience leads television producers to undue similarity and slavish imitation of what has proved to be popular. Minority audiences are ignored, and the television fare is reduced to a low, vulgar common denominator. It is one thing to argue that intellectuals cannot find suitable fare on television because it never pays an advertiser to try to attract this audience, an extreme position without empirical support; it is another to argue that serious intellectual programs should be provided during the evening hours and on the days when the potential audience is largest. It would be profitable for advertisers to cater more to minority tastes than they now do if there were more commercial television stations per city and if subscription television became a profitable enterprise. Nevertheless, even if both of these developments occurred and increased variety became possible, it would be naive to expect too much. Since there are fashions in novels and in movies, we should not expect less conformity of television programs.
Both radio and television depend on a relatively small number of firms for a substantial part of their advertising revenue. The 20 largest network television advertisers account for more than 22 per cent of network advertising receipts, and the 20 largest spot television advertisers account for nearly 14 per cent. By way of comparison, the 20 largest newspaper advertisers contribute less than 6 per cent of newspaper advertising revenue. Since there are only three national networks, it is safe to conclude that advertising rates for the largest firms are not arrived at by a purely competitive process. The Federal Communications Commission licenses all television and radio stations and limits the number of stations any one company, or network, may own. The public would benefit more from a larger number of television stations per city than it would from regulation of advertising rates. The development of ultrahigh-frequency (UHF) television, for example, would, by increasing competition, lead to a greater variety of television fare, just as the advent of frequency modulation (FM) radio and the consequent opening up of more radio stations gave listeners a wider range of choice.
Defenders of advertising have claimed as one of its benefits that it makes possible a free press. There can be little doubt that if there could not be advertising in newspapers, the price of newspapers would rise considerably and the circulation would fall—not only for this reason, but also because an important kind of news, advertising, could not be published. Without advertising, broadcasting companies could not survive unless subscription radio and television became profitable. When there is a free press the public can hear a greater variety of views than when the press is controlled by the government. Truth is less easily concealed when conflicting views can be presented to the public. A government that supports and controls the news media, no matter how benevolent it is, cannot be trusted to give as much free expression as would unfettered competition among companies in the news industry, which depend on the public and advertisers for revenue.
Advertising and consumer sovereignty
A convenient assumption in textbooks is that consumers have given, and possibly sacrosanct, tastes and that firms cater to these. As a practical matter this is a myth. People are molded by the opinions and pressures of others from the moment of birth. Our role as consumers begins with what we learn from our parents and continues throughout our lives. Of course, advertisers serve their own interests when they try to get us to buy their wares. However, most people attempting to persuade others are often serving their own interests, and we must learn to be discriminating. A free society requires freedom of persuasion, including attempts to influence how people spend their money. Critics who claim that advertisers lie or induce people to buy things they do not need or to discard goods that are still serviceable overlook the fact that sales clerks can do all this just as effectively; these critics certainly exaggerate the power of advertising. Moreover, they ignore certain mundane but nevertheless important facts about advertising. For every advertiser who tries to persuade consumers to do a particular thing or to buy a particular item, there are others trying to persuade consumers to buy a rival product. In modern industrial society consumers can choose from an ever-increasing and ever-changing array of goods. Wise consumption is difficult under these conditions and demands a degree of skill not required of shoppers in a simpler economy. Efficient distribution in a society of rising real wages requires greater reliance on such labor-saving devices as the self-service store. Advertising is one of the promotional techniques that meets the requirements of efficient distribution. We can expect it to play an ever-increasing role in the developing economies of the world.
Lester G. Telser
[Other relevant material may be found inCommunication, mass; Market R>esearch.]
BIBLIOGRAPHY
Advertising Age: The National Newspaper of Marketing. ? Published weekly since 1930.
Blank, David M. 1963 A Note on the Golden Age of Advertising. Journal of Business 36:33–38.
Borden, Neil H. 1942 The Economic Effects of Advertising,. Chicago: Irwin.
Canada, Dominion Bureau of Statistics 1956 Advertising Expenditures in Canada, 1954. Reference Paper No. 67. Prepared by the Merchandising and Services Section, Industry and Merchandising Division, Department of Trade and Commerce. Ottawa: Department of Public Printing and Stationery.
Ferguson, James M. 1962 The Advertising Rate Structure in the Daily Newspaper Industry. Ph.D. dissertation, Univ. of Chicago.
Goldman, Marshall I. 1960 Product Differentiation and Advertising: Some Lessons From Soviet Experience. Journal of Political Economy 68:346–357.
Harris, Ralph; and seldon, Arthur 1962 Advertising and the Public. London: Institute of Economic Affairs.
Kaldor, Nicholas 1950 The Economic Aspects of Advertising. Review of Economic Studies 18:1–27.
Kaldor, Nicholas; and SILVERMAN, RODNEY 1948 A Statistical Analysis of Advertising Expenditure and of the Revenue of the Press. Cambridge Univ. Press.
Katz, Elihu; and Lazarsfeld, Paul F. 1955 Personal Influence: The Part Played by People in the Flow of Mass Communications. Glencoe, III.: Free Press.
Nerlove, Marc; and Arrow, Kenneth J. 1962 Optimal Advertising Policy Under Dynamic Conditions. Economica New Series 29:129–142.
Ozga, S. A. 1960 Imperfect Markets Through Lack of Knowledge. Quarterly Journal of Economics 74:29–52.
Printers’ Ink: The Weekly Magazine of Advertising, Selling, and Marketing. ? First published in 1888. Since 1954 one issue each year is a guide to marketing.
Stigler, George J. 1961 The Economics of Information. Journal of Political Economy 69:213–225.
Telser, Lester G. 1962 Advertising and Cigarettes. Journal of Political Economy 70:471–499.
U.S. Federal Trade Commission 1944 Advertising as a Factor in Distribution. Part 5 of Report on Distribution Methods and Costs. Washington: Government Printing Office.
II ADVERTISING RESEARCH
Advertising research is here defined as the study of that part of human behavior attributable to overtly paid-for communications. It is thus distinguished not only from the broader discipline of marketing research but also from kindred studies of aggregate purchasing behavior, economic concomitants of advertising, and communications not always or not overtly paid for. [For discussion of these fields, seeCommunication, mass; Consumers, article onconsumer behavior; Market research; Propaganda.]
As subject matter for the social sciences, advertising remains curiously neglected. Advertising has long been of importance and today accounts for at least one per cent of the national income of most highly industrialized countries (International Advertiser 1965), but it had no important place in economic theory until the 1930s. Advertising accounts for a major share of the promotional expenses of most firms but has not been extensively treated in any but the most recent microeconomic theories of the firm (Ramond 1964c, pp. 662–675). Advertising accounts for an undeniably larger and larger visible and audible part of world culture but as a sociological phenomenon has been described mainly by journalists. Advertising is a much-discussed influence on individual behavior but has interested psychologists only insofar as it can be studied in the experimental laboratory or used to illustrate the clinician’s theories of unconscious motivation. It is hardly surprising that there is no accepted theory of advertising.
The little known about how advertising works comes largely from research done since World War II by advertisers and advertising agencies in the United States, western Europe, and Japan. Most early studies remained unpublished. Until the establishment in 1960 of the Journal of Advertising Research (JAR), there was no means for exchange of ideas among all gatherers or users of this work. Since then some fifty or sixty reports of advertising research are published each year, mainly in the JAR.
Not all this research is beyond criticism. Much of it is done by practitioners whose training in the social sciences and ancillary disciplines has been acquired on the job. In the face of time and cost limitations, methodological standards relax, When this happens, advertising research often merits the charge (Forrester 1958) that it is itself nothing more than advertising.
The primary basis for these political considerations is the organizational structure of the advertising industry. Nowhere in business, perhaps, has accurate evaluation been so inhibited by the organizational structure of the process to be evaluated. Advertising is bought for a manufacturer by an agent, and this agency has usually been responsible for the evaluation of advertising effectiveness. Someone has remarked that this is like having the fuel salesman evaluate the furnace. The fact remains that it has not always been in the best interest of the advertising agency to institute adequate research designed to evaluate advertising effectiveness. The agency has little to gain and much to lose. Instead of such evaluative research, it typically performs only those studies which are necessary to aid its own current decision making and that of the client. Such research is done only when needed and can rarely be accumulated for future guidance. Thus, the ultimate promise of any type of research —its self-liquidation through accumulation of permanent knowledge—is almost by definition denied in the advertising realm.
As most advertising research is designed to aid some decision, it is possible to classify the types of research according to the decision each is designed to help make. For example, decisions about what to say come under the heading of motivation research; decisions about how to say it, under copy research; where, when, and how often to say it, media research; and how much to spend, sales research.
Motivation research
The postwar transition from a war production economy to a consumer economy heightened advertisers’ interest in the motives of their target audiences. By the early 1950s the business community had learned from clinical psychologists that consumers were governed in part by unconscious motives, or at least by motives they had trouble expressing to ordinary interviewers. Depth interviews, projective tests, and the other paraphernalia of the clinical psychologist became popular tools in advertising research. By 1955 the public became interested in whether it was in fact being sold by advertising against its will. Popular books exploited this fear on the part of the book-buying public and ultimately received a quiet rejoinder from more realistic students of human behavior. John Dollard, at the second annual conference of the Advertising Research Foundation, in a frequently anthologized paper entitled “Fear of Advertising,” said: Nor do I fear research into unconscious motives, sometimes called motivation research, as an important factor in mass subversion by advertising. My reasons are as follows: many unconscious motives are stark and ugly and cannot be used in advertising appeals. Furthermore, unconscious motives are tricky; they are likely to come in conflicting pairs of desire and disgust, and one cannot evoke the desirable member of the pair without also evoking its linked opposite. It should be noted also that people are not immediately prone to carry out the unconscious motives which they do have. There are strong forces built into the personality which operate against most unconscious motives. Looking at the matter from a quite different standpoint, it has yet to be proved that unconscious motives can be steadily identified or that, if this were done, they are of superior effect in devising advertising themes. At the moment, the notion of using hidden factors in motivation to influence behavior on a mass scale is still in the status of a bright idea or a horrible fantasy, however you prefer to look at it; but it is not a reliable and valid instrument available to advertisers. (Dollard 1956, p. 7)
As public and professional expectations grew more realistic, motivation research gradually ceased to be regarded as the search for exploitable unconscious motives. Instead it became redefined more modestly as the study of those psychological variables which might be related to the consumer’s purchase of products or services. Put another way, motivation research became the study of relationships between the psychological attributes of a brand or product (its “image”) and the psychological attributes of the consumer (his “personality”). If markets can be segmented, so the argument runs, according to psychological as well as demographic variables, then it should be possible to fashion advertising appeals which are unusually effective in causing sales among the appropriate population segments.
Perhaps the largest body of data on which these hypotheses have ever been tested became available at the J. Walter Thompson advertising agency in 1959, when the Edwards Personal Preference Test was administered to a sample of over three thousand households (Koponen 1960). The Edwards test provided scores on 15 personality traits for the male and female head of each household, while a purchase diary indicated amounts and brands of various products purchased. A prior experiment had shown that a subgroup selected for its scores on certain traits bought more of a mail-order product than did an unselected control group, in response to a direct-mail advertisement using appeals designed specifically to satisfy that particular subgroup’s psychological needs. Neither group bought very much, however, and the results were deemed statistically significant but practically unimportant.
So, in fact, were the relationships found between personality variables and purchase behavior using the J. Walter Thompson data on beer, coffee, tea, and toilet tissue. Although brand loyalty, store loyalty, and amount purchased do correlate significantly with certain demographic and personality traits of the 8,900 subjects in the sample, these correlations are too small to give practical guidance. It seems safe to conclude that until more discriminating personality scales are developed, perhaps for the specific purpose of predicting purchase behavior, psychological market segments will not be significantly more useful to the advertiser than demographic market segments.
Copy research
Having learned the needs of his prospects and chosen general appeals or themes by which to reach them, the advertiser must then determine how best to execute those themes. Which copy, headlines, illustrations, music, etc., will best communicate his message? Studies answering questions of this sort have traditionally made up the bulk of advertising research and are still called —even in the age of television—copy research.
Reviews and collections of recent copy research appear elsewhere (Lucas & Britt 1963; National Industrial Conference Board 1963; Twedt 1965). The common implication of many of these studies is that the respondent’s verbal testimony about exposure to advertising cannot be taken at face value. Commercial services report the proportion of a sample who on being shown an advertisement claim to have noted it (the recognition method) or who, on being shown a brand advertised in a magazine they have read, can “play back” enough of the advertisement to indicate convincingly that they have in fact seen it (the aided recall method). The simultaneous popularity and questionability of these techniques led to the largest purely methodological investigation ever conducted in copy research, the five-volume Printed Advertising Rating Methods Study, or PARM (Advertising Research Foundation 1956–1957). The PARM study found, among other things, that recognition scores did not decrease as the time since reading increased, as one would expect if memory loss were occurring. This suggests not that the respondent was “recognizing” the advertisement he had actually seen but the likelihood that this was the kind of advertisement he would have noted, given the opportunity. Other studies find that respondents claim to recognize control advertisements that they could not have seen. The extent of misclaiming was directly related to the respondent’s reports of past reading behavior, his interest in the product advertised, and other personal characteristics. It may be concluded that while these claims perhaps have some value as projective data describing the respondent, they are substantially useless as reports of prior exposure.
They may also be useless as predictors of future behavior. Haskins (1964), in a review of 28 studies, has shown that factual recall of advertising can change without corresponding changes in behavior or other attitudes. He concludes that what is retained by respondents may have nothing to do with their subsequent purchases. Ramond (1965) has shown that such failure of attitude change to predict or coincide with behavioral change may be an artifact of the methods commonly used to measure these changes.
As verbal behavior became increasingly suspect, nonverbal behavior became increasingly popular as a measure of copy effectiveness. Laboratory methods, themselves suspect for their artificiality, found increasing favor as they began to measure relatively involuntary responses to advertising. These included visual recognition, skin moisture, pupillary dilation, and even the rate at which someone would press a pedal to maintain a television picture and sound. No measure has been found completely free from conscious cognitive influences, but the trend of current copy research seems to indicate that such is the goal.
Media research
Having decided what to say and how to say it, the advertiser must then decide where, when, and how often to say it. Surveys and analyses that guide these decisions are known as media research and have as their aim the selection of the audiences for advertising placed in media. They should not be confused with studies that are intended to determine the audiences of only the media themselves.
A comprehensive bibliography of U.S. media research may be found elsewhere (Ramond 1964a), along with a comparison of U.S. and European approaches. For example, there is much variety and controversy in U.S. media research. In the United States no single method has been hammered out for use in an industry-wide study, whereas this has been done in France, Britain, and Sweden (Ramond 1964b). All of these studies have been influenced by the work of the Advertising Research Foundation’s (ARF) Audience Concepts Committee as expressed in its booklet, Toward Better Media Comparisons (Advertising Research Foundation . . . 1961). The committee maintains that to understand the transmission of advertising through media one must count or measure at six stages: vehicle distribution, vehicle exposure, advertising exposure, advertising perception, advertising communication, and consumer response— usually sales.
Vehicle distribution is a count of things, namely, the number of physical units through which advertising is distributed. In broadcast media these are receiving sets in use; in print media, number of copies sold.
Vehicle exposure is a count of people, those whose open eyes or ears were confronted by the vehicle: for broadcast media, by the ongoing program; for print media, by the open page.
Advertising exposure is a count, not of things or people, but of events. These events are also confrontations of open eyes or ears by a turned-on set or an open page, but only where the set or page is carrying a commercial or advertisement.
The fourth, fifth, and sixth stages of the model are perception, communication, and response. There is both a conceptual and an operational difference between these stages and the previous three.
Distribution and exposure measures can in principle be—and have in practice been—objectively defined to the satisfaction of many if not all. In dealing with perception and communication of advertising, however, we face the problem of separating the effect of the medium from the effect of the advertising message itself. So far no research supplier has come forward with any measure of these stages which is as objective, repeatable, and intuitively compelling as those of distribution and exposure.
Actually, perception and communication may be defined by an infinite number of responses. Though a perception may be defined as that which is seen, heard, or in some way received by a sense organ, our knowledge of whether it has been received must be derived from the response of the receiving person. He must somehow translate his private observation into a public gesture. Reports of recognition, recall, or attitudes are possible classes of responses which might define perception or communication of advertising. But note that whatever definition is used, it will require not only that the consumer see or hear the advertising but also that he remember something about it until he is asked to report it. Thus the number of possible definitions is further multiplied by the number of cues or aids to memory which could be used.
While the ARF Audience Concepts Committee made no recommendations concerning operational definitions of perception and communication, it did suggest a conceptual distinction. Perception is defined as an all-or-none phenomenon: it either occurs or does not occur. There are no degrees of perception.
But communication involves more than merely seeing an advertisement. For example, an advertisement can add to the consumer’s knowledge, change his attitude, or make him resolve to purchase the product advertised. It may change his beliefs, make some more prominent than others, or even evoke moods in which his judgments operate differently. Thus one must recognize degrees of communication. One may merely count perceptions, but one must measure communication in a more complex way. The goal in counting perceptions is to extract from each the effect of memory, but the goal in measuring communication is to learn the degree of remembering. Thus, although perception and communication differ conceptually, in practice they both require measures that isolate and distinguish the effect of advertisement-plusmedium from memory, and the effect of the advertisement from that of the medium.
We now come to the final stage of response to advertising, and from the advertiser’s point of view by far the most important. Response at this stage is necessary to justify any advertising at all. Other measures are satisfactory only insofar as they are related in some way to sales. However, sales are the outcome of a great many factors other than advertising. Not only are the personal factors, the attitudes, beliefs, perceptions, and so forth of each individual important, but external factors such as price, market conditions, changing tastes, etc., are also operative.
Theoretically, the way to estimate the sales response to an advertising unit is to arrange for tests with all factors that could possibly influence sales—except for advertising—held equal or randomized. In one area the advertising unit would be displayed, in the other it would not, and the difference in sales in the two areas would be measured. Practically, experiments of this kind turn out to be very difficult. In the first place, if we try to choose markets that are identical in all the relevant characteristics, we usually find some discrepancy between the markets on a variable which we feel sure will affect the outcome of the experiment decisively. Again, we may list what we think are the relevant variables in order to account for their effects, but we never know whether we have been able to identify them all. Even with an exhaustive list of market variables, we may find that the nature of our advertising unit is such that it may act differently in the two markets because of the different psychological characteristics of their inhabitants, or the differing media habits in the two markets. Perhaps the ultimate objection to research designed to attribute sales gains to single advertising units is the very small effect which any one advertising unit may have. Sales tests have been applied mainly to the evaluation of campaigns, not of individual messages.
Sales research
No amount of research into motivation, copy, or media will tell the advertiser what, if anything, his advertising has done to influence the sales of his product. For this information he must conduct sales research, wherein he isolates the contribution to sales of his advertising expenditures. Many advertisers have done so, in the United States and elsewhere, despite the difficulties enumerated in the previous section. The most comprehensive recent review and bibliography is Martin Mayer’s booklet for ARF, The Intelligent Man’s Guide to Sales Measures of Advertising (Mayer 1965). Others include Dominick (1960), the National Industrial Conference Board (1962), Palda (1963) and Ramond (1965).
Properly designed marketing experiments are used to avoid many of the above difficulties (Banks 1965). Proper design requires, among other things, that extraneous influences on sales be dealt with: controllable factors should be controlled; uncontrollable but measurable factors should be measured and accounted for statistically in the analysis; and enough experimental units should be used in each treatment to permit accurate estimates of their variability. From experiments published, we may conclude that when the following conditions are met the sales effects of advertising can be estimated with accuracy, speed, and economy:
(1) The product or the brand has no substitute now or in the foreseeable future. The number of competing products or brands is small, and it is unlikely to be made obsolete by technology during the period of experimentation.
(2) The buyers of the product or brand (a) can be unambiguously defined; (b) can be easily reached by advertising and interviewers; (c) are geographically concentrated; (d) are temporally concentrated—the shorter the selling season the better; (e) spend little time “in the market.”
(3) The lot size of the purchase is constant from purchase to purchase by the same buyer and the same from buyer to buyer.
(4) Price is constant over time, markets, amount purchased, etc.
(5) Channels of distribution are many. The more channels of distribution to the consumer the less likely he will be frustrated in an advertisinginduced attempt to buy.
(6) Levels of distribution are few. The more wholesalers, dealers, and distributors there are between producer and consumer, the more individuals who must decide before purchase can occur, and the more individuals who must be influenced by advertising.
(7) The influence of personal selling is constant over time and over markets.
(8) Technical services provided by competitors do not differ.
(9) The copy platform is constant and unambiguous. The fewer the copy points, the easier it is to tell if communication has occurred.
(10) Special promotions are not undertaken. (11) Packaging is distinctive and constant. (12) The producer is the only advertiser of the brand, i.e., there is no cooperative or local advertising.
(13) Competitors are slow to respond to changes in marketing strategy and maintain more or less the same marketing policies.
(14) Competitors’ advertising and marketing policies are relatively constant over markets.
(15) Potential sales can be accurately estimated for small geographical units, e.g., counties or census tracts, and during short time periods such as weeks or months. This follows from several of the previous desiderata.
(16) Government controls over product design, price, competition, and advertising are minimal or at least unchanging.
Clearly not all of these conditions can be met by most advertisers. Meeting them, moreover, does not guarantee a conclusive experiment but only the avoidance of certain common errors. Experimentation is increasingly popular, not because it always works, but because in many cases it is the only way to have a chance of getting unambiguous measures. In marketing as elsewhere in business, chance plays its inevitable role. Part of this role, however, can be made manifest by the experiment itself.
As published experiments accumulate, the conditions under which the sales effects of advertising can be accurately estimated will become clearer. Until then the prudent advertiser will determine for himself whether his own circumstances augur well or ill for this valuable form of marketing control.
Charles K. Ramond
[Directly related are the entriesConsumers; Market research. Other relevant material may be found inAttitudes, article onAttitude change; Communication, mass; Persuasion; Propaganda.]
BIBLIOGRAPHY
Advertising Research Foundation 1956–1957 A Study of Printed Advertising Rating Methods. 5 vols. New York: The Foundation.
Advertising Research Foundation, Audience Concepts Committee 1961 Toward Better Media Comparisons. New York: The Foundation.
Advertising Research Foundation 1965 Are There Consumer Types? New York: The Foundation.
Banks, Seymour 1965 Experimentation in Marketing. New York: McGraw-Hill.
Dollard, John 1956 Fear of Advertising. Pages 1-9 in Advertising Research Foundation, Proceedings of the Second Annual Conference. New York: The Foundation.
Dominick, Bennet A. 1960 Research in Retail Merchandising of Farm Products: Appraisal of Methods and Annotated Bibliography. Washington: U.S. Department of Agriculture, Market Development Research Division, Agricultural Marketing Service.
Forrester, Jay W. 1958 The Relationship of Advertising to Corporate Management. Pages 75–92 in Advertising Research Foundation, Proceedings of the Fourth Annual Conference. New York: The Foundation.
Haskests, Jack B. 1964 Factual Recall as a Measure of Advertising Effectiveness. Journal of Advertising Research 4, no. 1:2–8.
HRB-Singer, Inc., State College, Pennsylvania 1962 The Measurement and Control of the Visual Efficiency of Advertisements. New York: Advertising Research Foundation.
International Advertiser. ? See especially 1965, Volume 6, no. 10.
Koponen, Arthur 1960 Personality Characteristics of Purchasers. Journal of Advertising Research 1:6–12.
Lucas, Darrell B.; and Britt, Steuart H. 1963 Measuring Advertising Effectiveness. New York: McGraw-Hill.
Mayer, Martin 1961 The Intelligent Man’s Guide to Broadcast Ratings. New York: Advertising Research Foundation.
Mayer, Martin 1965 The Intelligent Man’s Guide to Sales Measures of Advertising. New York: Advertising Research Foundation.
National Industrial Conference Board 1962 Measuring Advertising Results. By Harry D. Wolfe et al. New York: The Board.
National Industrial Conference Board 1963 Pretesting Advertising. By Harry D. Wolfe et al. New York: The Board.
Palda, Kristian S. 1963 Sales Effects of Advertising: A Review of the Literature. Journal of Advertising Research 4, no. 3:12–16.
Ramond, Charles K. 1964a Operations Research in European Marketing. Journal of Marketing Research 1, no. 1:17–24.
Ramond, Charles K. 1964b Trends in U.S. Media Research. European Society for Opinion Surveys and Market Research, Commentary (Special Supplement): 35–43.
Ramond, Charles K. 1964c Marketing Science: Stepchild of Economics. Pages 662-675 in Stephen Greyser (editor), The Marketing Concept in Action. Chicago: American Marketing Association.
Ramond, Charles K. 1965 Must Advertising Communicate to Sell? Harvard Business Review 43, no. 5:148–158.
Twedt, Dik Warren 1965 Consumer Psychology. Annual Review of Psychology 16:265–294.